OverUnity Research

We are all familiar with an electromagnet see picture below.

If i place a hall effect transducer at one end, as current increases the magnetic poles at the ends of the iron nail gets stronger.

but what happens if the current is increased until the electromagnet coil starts glowing.

Would you expect the temperature of the coil to affect the magnetism strength at the ends where measurement is being made as the temperature of the wire increases to red hot.

Does it make any difference when the wire glows as opposed to not glowing?

Offcourse we already know what happens when we heat a permanent magnet to a highish temperature.

I expect the magnetic field to decrease as the current decreases since the rod-hot wire will have very high resistance.

Sounds reasonable.
I need to get some high temperature insulation to be sure, although i have spent an hour so far trying this, sometimes i do not see an increase in mag field after a certain point and sometimes i see a sudden jerk down in mag field, i was using stainless steel thin wire and will need to change this as it does not need to be many turns on an iron nail to have a high resistance and therefore the measured mag field is barley detectable using my UGN3503

I have an infra red room  heater consisting of a  two foot long coiled wire of Nichrome (I guess) inside a quartz tube of approximately 0.5" I.D..

This has always emitted a loud buzzing when first turned on and operates silently after the wire reaches an orange glow at about 11.07 Amps. This has always puzzled me but I never thought to research it until this thread was started.

The Amps drawn from the line does not change much, going from 11.50 Amps cold to 10.86  Amps fully heated, around 5.9% change.

I will try to measure the change in the emitted AC magnetic field, which theoretically should only drop by about 5-6%, reflecting the change in Amps drawn.

Thanks ION very interesting,i have also seen this in heating elements.
I will follow up this thread when i get some high temperature bits & pieces.

It was a bit of a crazy idea while thinking about the heater in the valve i am working on, and the other thing that made me think about this was the valve filament kick at power on, the difference here is that the kick is produced while cold while relaxation occurs when hot.

O0

Quote from: Peterae on 2012.08.05, 09:32:12
...
Would you expect the temperature of the coil to affect the magnetism strength at the ends where measurement is being made as the temperature of the wire increases to red hot.
...

The magnetic field depends only on the current in the coil. It doesn't depend on the wire temperature. That's what we expect. If you suspect a different behavior, only a careful experiment will be able to show it. The nail temperature will have also to be measured. The temperature dependancy of the magnetic permeability is well known and so influences the magnetism strength of the nail.

This morning I ran further tests on the infra-red heater.

My setup was scoped across a 0.01 Ohm shunt resistor to measure the current. Previously I used a Kill-A-Watt meter.

I measured the current and there are no spikes shown, the voltage and current waveforms are identical and in phase as you would expect from a resistive heater.

The strange thing is when I use a magnetic pickup coil into the scope, I get a very small 60 Hz. sine wave with a huge positive and negative going spike that I cannot account for. It dies out at around 11.07 amps leaving a small sine wave when the audible sound dies out.

The questions are:

Why does the extremely spiky waveform show up with the external magnetic pickup coil, but no such spike is shown when using the current shunt resistor.

Why does the spike diminish completely around 11.07 amps, right when the heater stops emitting the audible buzz and enters the glow phase?

I urge anyone with a heater to replicate this experiment.

I'm getting the same noise from my rectifier valve but only when it's turned on from stone cold and probably if i am slightly over driving it, it disappears after a couple of seconds, a low Hz vibrating noise, and i am also familiar with the noise you mention with the heater.
It looks like i need to add a sniffer coil to the valve  O0

I think the audible noise is from where the heating element coil is supported, but then the question would be why does the coil quiver while cold and not hot, specially when the hot metal would move a lot easier, maybe it's the ac magnetic field being attracted and repulsed relative to any other metal or iron inside the heater.
Still does not explain the kicks.

There might be a power control circuit that limits the current when you first turn the heater on.  That would help prevent tripping a circuit breaker.  That circuit might account for the noise and spikes that you are picking up with your magnetic probe.  There might be a line filter that would prevent detection with your 0.01 ohm line resistor.

I don't have one of these heaters, so this is just a guess.

No definitely not a power control circuit, i was brought up with large panel electric heaters  ;D
As a kid we had one 4 foot long in the early 70's and it had a grill along the top front, when you first switched it on you could see the heating element resonate and clunk against the supports down it's length, quiet a loud noise, once warmed it stops after a few seconds, the elements were nearly the 4 foot length and were air coiled.

Anyway i am hearing the same thing with my valve, it's as if the coil couples to something when first switched on, this maybe just because the current is much higher at this point and hence the mag field is much stronger allowing coupling.

but the interesting point ION brings are the spikes, the audible noise is nothing new to me.

We had a fan heater a few years ago that made the noise as well.

OK i have listened to my valve each time i power it up and i can hear all sorts of oscillations and vibrations from it fairly easily as it warms up.

Here's a recording of it warming up, sounds a bit like those double pulsing coil noises i was getting.

Quote from: forcefield on 2012.08.08, 16:04:29
There might be a power control circuit that limits the current when you first turn the heater on.  That would help prevent tripping a circuit breaker.  That circuit might account for the noise and spikes that you are picking up with your magnetic probe.  There might be a line filter that would prevent detection with your 0.01 ohm line resistor.

I don't have one of these heaters, so this is just a guess.

There is no power control circuit or line filter in this heater, just an on-off bi-metal thermostat. Anyway, if you read my first post you will note that there is no large inrush current, just around 5.9% higher current at first  turn on, not enough to cause concern.

Tungsten wound lamps have a huge inrush by comparison. I've worked in this field, designing soft start SCR circuits for such lamps to eliminate nuisance fuse blowing upon turn on. The effect I see here is something different.

I will attack this anomaly from a few different angles and report back as time permits.

This is a vertical infra-red coiled heater with a quartz or glass tube over the element. It is not sealed at the ends.

Coiled Nichrome is the standard for this type of heater.

The chart attached is in line with what I measure. Good data on this site:

http://www.wiretron.com/nicrdat.html

can you add time measurements to the table above?

how long after the initial turn-on does the noise stop?

Quote from: ION on 2012.08.08, 14:26:47
This morning I ran further tests on the infra-red heater.

My setup was scoped across a 0.01 Ohm shunt resistor to measure the current. Previously I used a Kill-A-Watt meter.

I measured the current and there are no spikes shown, the voltage and current waveforms are identical and in phase as you would expect from a resistive heater.

The strange thing is when I use a magnetic pickup coil into the scope, I get a very small 60 Hz. sine wave with a huge positive and negative going spike that I cannot account for. It dies out at around 11.07 amps leaving a small sine wave when the audible sound dies out.

The questions are:

Why does the extremely spiky waveform show up with the external magnetic pickup coil, but no such spike is shown when using the current shunt resistor.

Why does the spike diminish completely around 11.07 amps, right when the heater stops emitting the audible buzz and enters the glow phase?

I urge anyone with a heater to replicate this experiment.

I assume the heater wire in which the current flows is coiled. I guess, before the current flows the heater wires are in a room temperature, have a definite length to this temperature. When switched on, the current rises, and the wires expend because of rising temperature, and the opposing magnetic fields. As the sine wave goes on, current flow changing directions, the wires vibrate back and force with less and less amplitude, until they  reach the glowing temperature, where there will be no more movements. I guess the back vibration when the current falling, because of the fast heat dissipation, so related to the material temperature related length change.

I guess it's all happening because the wires need several AC cycle to reach the temperature, where the heat generated by the previous cycles, doesn't dissipate fast enough until the next current rise coming. When the glowing temperature reached, the will be no temperature related movements.

The pick up coil may pick up the field spikes generated by the current carrying vibrating wires. You cannot measure with a shunt resistor those spikes, because your measuring point should have in an outside reference frame, from where the movement of the wires (which are now moving magnetic field sources) could adding to the flowing current generated waveform.

I could be totally wrong.

QuoteThe strange thing is when I use a magnetic pickup coil into the scope, I get a very small 60 Hz. sine wave with a huge positive and negative going spike that I cannot account for. It dies out at around 11.07 amps leaving a small sine wave when the audible sound dies out.

Ion,  i've seen this same type of waveform from one of my toroid coils, and it was due to magnetic saturation.  As the material saturates magneticaly, the inductance drops faster and faster, and so the current increases rapidly, so the dI/dt is fast and creates a large dB/dt response which induces the voltage spike.

EM

Theres a problem with that EMdevices. Here's a quote from ION

QuoteI measured the current and there are no spikes shown, the voltage and current waveforms are identical and in phase as you would expect from a resistive heater.

Quote from: EMdevices on 2012.08.09, 00:12:06

Ion,  i've seen this same type of waveform from one of my toroid coils, and it was due to magnetic saturation.  As the material saturates magneticaly, the inductance drops faster and faster, and so the current increases rapidly, so the dI/dt is fast and creates a large dB/dt response which induces the voltage spike.

EM

EM:

There is no magnetic core to create any saturation effect.  Nichrome is very weakly magnetic and there is no closed magnetic structure that can saturate.

I know the effect you are talking about and have studied it carefully on my bench. This is not magnetic saturation effect.

I wish people would try the experiment with Nichrome wire, and carefully observe the effect on an external magnetic probe.

I'll try to post the some waveforms as time permits.

To All: please read my prior posts carefully before offering a "theory". Voltage and current waveforms are completely in phase showing a near pure resistive heater. Granted it is coiled but inductance will be very small at 60 Hz.

ION did you hear the sound recording i did of the valve rectifier heater warming up in post 10.
Not sure if you missed it or not sorry if not.

Peterae:

Yes, I listened and it sounds very interesting.

Attached is the sound of the Nichrome heater. Note that just before the pulse decays into the sine wave (around the middle of the file) there is an additional burst of sound like the beginning.

So there is a quick buildup to the first maximum amplitude, a slow decay, then two large buildups again before fading into silence.  The large amplitude waveforms occur as the spike is decaying back into the sine wave.

If you view this file you can see the changes. I did not fade the file, the sound dies out to a very low level.

Ion,


This magnetic kick is "continuous" until the wire heats up and the buzzing stops?

How far is the pickup from the wire?

Quote from: Grumpy on 2012.08.09, 13:30:19
Ion,


This magnetic kick is "continuous" until the wire heats up and the buzzing stops?

How far is the pickup from the wire?

Yes, the spike picked up with the mag probe is continuous until the sound dies out, but the sound goes through three major amplitude changes during the heatup period. View the Nichrome Heater.mp3 audio file with an wave viewer to get a clearer view.

The Magnatek magnetic probe is oriented vertically near one end of the coil, about 1.5" away.

The microphone was placed about the same distance and location. Attached is a capture of the wave file. Do not confuse the audio file with the spike I am referring to. This will be posted later.

Here are two camera shots of the scope traces. Forgive the fuzziness. Smooth sine curve is the heater current. Spike waveform is superimposed.


First shot (001) is spikes as picked up with mag probe full amplitude at initial time up to 5 seconds vs. current.

Second shot (002) shows spikes decaying after 5 seconds versus current.

Can you move the probe in-out to see if there are other peak points or is it anywhee within a zone arong the wire?  i.e. map the field perp to the wire.

@ION

Here is a possible explanation. Under its own magnetic field, a coil tends to collapse on itself, because each turn of wire acts as a magnet and they attract one another.
With AC current, this effect produces a mechanical vibration that is hearable. It also changes the inductance which becomes time dependent at the rythm of the signal, and can disturb it (for example by producing the peaks, depending on the mechanical displacements).
When the wire heats, its mechanical characteristics change, it becomes softer and consequently the mechanical resonance is decreasing.
It's just an idea. Further tests are needed.

Quote from: exnihiloest on 2012.08.09, 15:02:40
@ION

Here is a possible explanation. Under its own magnetic field, a coil tends to collapse on itself, because each turn of wire acts as a magnet and they attract one another.
With AC current, this effect produces a mechanical vibration that is hearable. It also changes the inductance which becomes time dependent at the rythm of the signal, and can disturb it (for example by producing the peaks, depending on the mechanical displacements).
When the wire heats, its mechanical characteristics change, it becomes softer and consequently the mechanical resonance is decreasing.
It's just an idea. Further tests are needed.

Possibly, but the inductance is way too small to account for this, a few uH at best. I will measure it to satisfy.

I will continue testing. Here are two additional DSO shots, showing the large spikes (from the Magnatek probe) at the zero crossing, and how they nearly completely die out  as the wire heats up.

They are superimposed on the current waveform, which is quite clean. There are smaller amplitude spikes besides the large ones at the zero crossing.

Additional note: Although I state that the spikes are occurring at the zero crossing, this is only with reference to the 60 Hz current waveform. In actuality the spikes appear to be at the peaks of the sine wave as picked up by the Magnatek, the phase of which changes with location.

Do note that the current and voltage waveforms of the input power are exactly in phase. I have not posted this shot yet.

Here are the current and voltage waveforms into the heater. They are in phase and the power factor measured with a Kill-A-Watt meter is 1.00 showing nearly no inductance.

No evidence of spiking in the current waveform.

If the big spieks are at the peaks, then the little ones corrspond to the zero crossings.

You say the "phase changes with location", what does this mean in the physical field?

Ion,

you are correct, the spikes are not from magnetic saturation, since they occur at the zero crossing.  thanks for the pictures.    What these are is most likely magnetic domain flipping, or barkhouser noise.    Nicrome wire is magnetic since it contains nickel, but not very strong.  As the wire warms up and the tempeture goes above the Curie tempeture the material is no longer magnetic so these effects die down.   

Have you guys seen the magnetic thermal "engine" or pendulum that swings and gets attracted to a magnet, but a flame is placed right there as well to heat up the pendulum mass (iron or steel) and as it heats up it looses its magnetism and falls back down, and swings and comes back.  It's very neat to watch!     

EM

Great work ION, it sure is a pretty picture seeing those kicks.

I am pretty sure this is what SM wanted us to find and this is why.

The shape of those kicks are exactly the same shape i see in the noise i had super imposed on my HT Supply, they were very small, the reason being that the magnetic kick or current kick as SM calls it  :) was being induced inside the valve onto my HT supply or very loosely coupled, this was a clue, when SM said he analysed everyting coming out he must of meant magnetic field as this is where it really shows up.
Heat really is a problem and now we know why, too hot and the dam kicks don't show up.

I am not sure how close i can get to this because of the valve construction, but i am going off to try and measure the mag field.


Happy days ahead.

O0

Quote from: EMdevices on 2012.08.09, 18:25:50
Ion,

you are correct, the spikes are not from magnetic saturation, since they occur at the zero crossing.  thanks for the pictures.   What these are is most likely magnetic domain flipping, or barkhouser noise.   Nicrome wire is magnetic since it contains nickel, but not very strong. As the wire warms up and the tempeture goes above the Curie tempeture the material is no longer magnetic so these effects die down.  

Have you guys seen the magnetic thermal "engine" or pendulum that swings and gets attracted to a magnet, but a flame is placed right there as well to heat up the pendulum mass (iron or steel) and as it heats up it looses its magnetism and falls back down, and swings and comes back.  It's very neat to watch!    

EM

EM

I've already considered all of what you say.  I've considered domain flipping and Barkhousen noise.  I'm not sure about either of these being the cause. Yes the Curie point may account for the eventual decay and the magnetic effects of NiChrome surely has something to do with this effect, but NiChrome "C" is extremely weakly magnetic. It is hard to pick up a small coil of it with a very powerful magnet.

However, NiChrome "C" has about 24% iron.

This site says NiChrome is non magnetic: http://hcrosscompany.com/metals/nichrome.htm

And this site : http://www.buzzle.com/articles/properties-of-nichrome-wire.html

has this to say:

QuoteNichrome is basically a name given to nickel-chromium resistance wires. It is a non-magnetic alloy which consists of 80 percent nickel and 20 percent chromium by weight, and is widely used in heating elements because of its relatively high resistivity. Out of the constituents of nichrome, nickel is an element with the chemical symbol Ni and atomic number 28, whereas, chromium is an element with the symbol Cr and atomic number 24. Nickel is generally a silvery-white lustrous metal and chromium a steely-gray and hard metal. Nichrome, the alloy, is silvery-gray in color, resistant towards corrosion and has a high melting point.

The latter two sites are probably talking about Nichrome "A"

More research is required before I form any theories.

Are the large spikes at the wave peaks or at the zero crossing?

Quote from: Grumpy on 2012.08.09, 19:52:55
Are the large spikes at the wave peaks or at the zero crossing?

When using the current waveform as a reference, the large spikes seem to occur at the zero crossing.

When referred to the sine on the mag probe, they are closer to the peaks.

Refer to the scope shots on the prior page.

this is a link to a Curie tempeture demonstration with nichrome wire.  

http://www.tarangscientificinstruments.com/manuals/Curie-Point.pdf




[edit:]

Can I interest anybody in a Curie Motor?     8)

we can probably design this with nichrome wire as well and place electrical contacts on the rim to heat up the sector that needs to be above the curie temperature.  The current flow could produce some electrodynamic forces, so that's why I'm using a laser beam, to show that it is strictly heat that renders that section of the wire non magnetic when the temperature is above the Curie point.

Quote from: EMdevices on 2012.08.09, 20:19:47
this is a link to a Curie tempeture demonstration with nichrome wire.   

http://www.tarangscientificinstruments.com/manuals/Curie-Point.pdf

We can agree that the Curie effect is responsible for the eventual decay of the pulses.

We need a good explanation for the cause of the pulses. Explanations thus far have been rather weak.

half wave rectify power to the heater (i.e. chop off the bottom half)

Then, do you still have the spikes and noise?  ( I d not think you will.)

since those spikes occur very close to the zero crossing, as the current changes direction and the magnetic field reverses, the Barkhausen noise is the best explanation I can came up with because that's where it occurs, right when directions change.    Why it is occuring in one solid burst is a bit puzzling, but it could be becasue we are dealing with a wire and not a core.  In a wire, if one domain flipps, it starts a chain reaction and it could trigger other domains in a avalanche mode, and the wire having a very small cross section than triggers everything at once, as opposed to a core.    That's my thinking at least, but research is needed to prove it.   I'll do some searches on the internet see what I can find.

@Grumpy,   correct, we won't see them if the reversal does not occur.

EM      

Ion is right,  we need to ponder this some more.  

I just studied the picture he posted more closely, and I don't agree that everything we see is Barkhausen effect.  

The zero crossing could very well be due to the Barkhausen effect, but the other spikes are CLEARLY due to a change in resistance which we can see in the current waveform as a discontinuity in the slope which I circled in red.    Remember,  a discontinuity in current, represents a large dI/dt, and the induced voltage in a coils is proportional to it, so a very slight change in the waveform produces large magnetic changes which produce by induction large voltage spikes in the pickup coil.

The question to answer is,  why is the resistance sudenly changing?   or is it?   what if it's the voltage source that is flaking out?   there is clearly a dicontinuity in the slope of the current waveform and that's what interests me at the moment.

EM

Quote from: EMdevices on 2012.08.09, 20:38:06

@Grumpy,   correct, we won't see them if the reversal does not occur.

EM      

Also, you will not see it if the AC wave is biased above zero as there is no reversal.

That looks a little strong for Barkhausen noise.

Can you measure the electric field near the magnetic probe? or place some sort of charge indicator there?

(http://www.overunityresearch.com/index.php?action=dlattach;topic=1505.0;attach=8313)

I am seeing other pulses as well just after the maximun pk is reached and then the sine changes direction, but interestingly these are half the height and interestingly these also seem to distort the sine slightly.

In the UK it's possible to buy small lower power quartz tube heating elements which are sold as replacements or spares.

A lower power version of ION's heater is needed so we can run from a variac, this way we maybe able to stop it getting to hot, it will also be possible to see the pulse amplitude with relation to the current drawn.

http://www.tlc-direct.co.uk/Products/DXPH08001.html?source=froogle

Although as ION states one can be wound with nichrome.
I have a reel of stainless steel wire which if i can work out how long i need to wind it for our mains i maybe able to wind one, stainless steel is not magnetic.

Peter,  I guess we're all posting at the same time.    See my picture in my previous post.


For all that are not familiar with induction, the basic formula for the voltage induced is

V = - N * dphi/dt,

where, 'phi' is the magnetic flux, 'N' the number of turns.   This says that the voltage induced in a coil of 'N' turns is proportional to the number of turns and the RATE OF CHANGE of the magnetic flux through the coil.    

So,  note what is important here, the RATE of change, and NOT the MAGNITUDE of the magnetic field.   As a consequence, we can have a very weak field that changes rapidly, and it will induce a higher voltage than a  stronger field that changes slowly.    (current produces the magnetic field, so it indirectly depends on the curent as well, and we should all know the basic inductor formula  V = L di/dt, which is just another form of Faraday's induction equation)

This is what SM was trying to explain, with his weak magnet moving very fast past the coils, analogy.

EM

Sorry EM  :)

Let's not forget the whole picture here, someone needs to do a spectrum plot on the audible noises from his coil, i am leaning towards harmonics of 60Hz in the US and up into the kHz as well.

yes, that's what I would expect as well, i.e. harmonics of the 60 Hz  (in the US) and 50 Hz in europe and elsewhere.

by the way,  a bread toaster produces the same noise when turned on, and transformers have the same noise, but it does not decay because the temperature does not exceed the Curie temperature.   The cause of this noise is the famous  magnetostriction properties of magnetic materials, i.e.,  changes in length due to magnetisation.

EM

EM said

QuoteIn a wire, if one domain flipps, it starts a chain reaction and it could trigger other domains in a avalanche mode,

This is more in line with what I've been thinking, and maybe what SM was trying to teach.

Consider also  that it might be a bulk electron effect, with unpaired electrons being aligned in a certain direction as the sine current increases, then instead of slowly relaxing to their former position as the sine proceeds towards zero, they flip all at once to zero by going "over center".

and:

QuoteThe cause of this noise is the famous  magnetostriction properties of magnetic materials, i.e.,  changes in length due to magnetisation.

I can totally agree with this........but the radiated spikes keep me guessing.

ummm

Isn't this effect measured outside the wire?

So, what's flipping?

@ Grumpy,   

what's flipping is the magnetic domains inside the wire.    The magnetic field is very strong the closer it is to the current, and the wire is slightly magnetic due to Nickel.

@ All,

looks like these guys beat me to it.  I never have any original ideas!  :(

http://www.oberlin.edu/physics/catalog/demonstrations/em/curiemotor.html


but maybe nobody has thought of this,  how about a solar Curie motor, that should be original!   ;D

EM

Quote from: EMdevices on 2012.08.09, 22:23:49
@ Grumpy,   

what's flipping is the magnetic domains inside the wire.    The magnetic field is very strong the closer it is to the current, and the wire is slightly magnetic due to Nickel.
but maybe nobody has thought of this,  how about a solar Curie motor, that should be original!   ;D

EM

and it makes a spike like that  1.5" from the wire

When nichrome is barely magnetic if at all

Ok guys I wound up 100 turns B&S #19 0.036" dia Nichrome on a 5/16 ceramic mandrel.

I don't know if this is type A, C or Tophet. The roll only said Nichrome, Nickel plated.

I cannot detect anything like the other scope shots with the mag probe. I don't hear any buzzing either.

Don't run off and buy any heaters until I figure this out.

My original heater is a Martel Quartz Glo Model Q15T.  ???

regarding the heater itself and probing it, can you somehow add a lot of dielectric to the probe?  If it is a coil, can you put plastic inside (like a rod) and outside (like a tube)?  Or put it in a plastic container? or other dielectric.

and see if it "changes the nature of the detected pulse".

Sounds crazy but I expect it will change the pulse "significantly".

@ Grumpy,

sure,  even 100 ft away from a single wire that carries a current of maybe 0.001 amps, if you interupt the current "instantly" you can generate an "infinite" voltage spike.    Like I said, it is not the magnitude that counts but the rate of change.

Let's think about a nuclear explosion and the EMP pulse,   most people think that it is some STRONG and POWERFUL magnetic field that causes the effect, but in actuallity, it is not so much the magnitude of the magnetic field, but the RAPIDITY AND EXPLOSIVENESS of the chain reaction that creates the CHANGE in the magnetic field which induces large voltages far away.

So,  rapid magnetic changes = higher induced voltage.   Faradays law of induction in a nut shell.


@ Ion,

the sound can be heard if the construction is condusive to it, i.e.  if it amplifies or couples well to the air, like a speaker, otherwise there is movement but poor acoustic coupling to the air.   Another thing to consider, now you have 100 turns, so your resistance is a lot higher, so now your power disipation is lower, and maybe your wire does not heat up to the same temperature, or the current is not so high to produce the effect.    The magnetic field is proportional to the current so if its not as high perhaps it won't work.  

keep stroking this problem, it's very interesting from multiple points of view ..

EM

EM said:

Quotethe sound can be heard if the construction is condusive to it, i.e.  if it amplifies or couples well to the air, like a speaker, otherwise there is movement but poor acoustic coupling to the air.   Another thing to consider, now you have 100 turns, so your resistance is a lot higher, so now your power disipation is lower, and maybe your wire does not heat up to the same temperature, or the current is not so high to produce the effect.    The bagnetic field is proportional to the current so if its not as high perhaps it won't work. 

Actually my resistance is less on the hand wound coil, 3.56 ohms versus 9.56 ohms for the larger unit. I run the smaller coil at it's limit, around 6 amps.

There might be something different about the wire used in the Martel heater, so I have to dig deeper.

I understand about the acoustic coupling to the frame.

G: I'd like to run more tests as time permits but it is getting too hot in my house :D

uh yeah and this is 60 Hz...kinda slow change

in lieu of encasing the probe in plastic, a thick plate will do - like a small cutting board.

if it causes a change in that pulse, it is because the plastic is altering the propagation velocity of the cause of the pulse (in simple terms)

humor me and try it

Quote from: ION on 2012.08.09, 23:03:53
EM said:

Actually my resistance is less on the hand wound coil, 3.56 ohms versus 9.56 ohms for the larger unit. I run the smaller coil at it's limit, around 6 amps.

what about the inductance of each?


get a cutting board - if it alters the pulse, then I know what it is, else it is magnetic and just a curiosity (I discovered this accidentally with a coil set on a coffee cup to raise it off the bench - and wooohooo!!! it lit me up)

Quoteuh yeah and this is 60 Hz...kinda slow change

I know, 60 Hz is very slow, and it's effect can bearly be seen in the spiky waveform, so the voltage spikes are not caused by this 60 Hz directly but indirectly, through the snapping action of the magnetic domains that is the underlying phenomena behind the Barkhausen noise.    It's like extending a ruber band untill it snaps, and the snapping is a lot faster than the rate of stretch that is occuring.

some good reading.

http://www.insightndt.com/papers/technical/t013.pdf

Quote from ION

Quotetwo foot long coiled wire of Nichrome (I guess)

If it's really old then there maybe a possibility it's Rhodium, I knew a scrap metal dealer who in the 70's especially looked out for old heating elements because of the use of Rhodium, i don't know to what extent it was used, i seem to remember him especially looking for old storage heaters but maybe it was more widely used in times when it was cheaper.

I also knew a scrap metal dealer who made millions by buying up all the surplus spark plugs from the 2ND world war effort, they used platinum for the electrodes  :o

@ION
What mag probe do you use? Do you know for what frequency range it is specified? If its inductance is low, the high frequency components of the signal can be reinforced (especially in the frequency range of its natural resonance) and the low frequencies reduced. So the ratio between the spikes and the sine wave could be false, and the least parasitic peak in the signal be hugely amplified.

Several times I have been surprised to see strange signals from the grid. This was due to the filtering of the circuit through which I observed the signal. If the circuit acts as a high pass filter, we can see even radio frequency signals. For example we can see the spikes due to a triac controlling a light elsewhere in the house.

Could you put a low resistance (<1 ohm) in series with the heater, and measure the voltage at its terminals, in order to estimate the real current? It is possible with an oscilloscope, with two probes in differential mode.

Quote from: exnihiloest on 2012.08.10, 11:06:22
@ION
What mag probe do you use? Do you know for what frequency range it is specified? If its inductance is low, the high frequency components of the signal can be reinforced (especially in the frequency range of its natural resonance) and the low frequencies reduced. So the ratio between the spikes and the sine wave could be false, and the least parasitic peak in the signal be hugely amplified.

Several times I have been surprised to see strange signals from the grid. This was due to the filtering of the circuit through which I observed the signal. If the circuit acts as a high pass filter, we can see even radio frequency signals. For example we can see the spikes due to a triac controlling a light elsewhere in the house.

Could you put a low resistance (<1 ohm) in series with the heater, and measure the voltage at its terminals, in order to estimate the real current? It is possible with an oscilloscope, with two probes in differential mode.

ex:
My main mag probe is a Magnetek 100mV / Gauss @ 60 Hz. (Magnetek Corp. Boulder Colorado)

I also use other pickup coils to verify that it is not an artefact of the probe, such as long telephone coil relay electromagnet around 3" long with many turns of fine #36 wire on an iron slug, and other coils with and without magnetic cores. All nearly the same  spiky waveform.

If you have not read the whole thread, you will note that I have used a 0.01 ohm resistor shunt to measure the current. This is a four wire "Kelvin" type resistor, 2 leads for current, 2 sense leads to the scope. No artefacts in the current waveform

I tried another similar heater, coiled Nichrome element in a quartz tube. same effect a little less intense.

I am guessing they are Nichrome. based on the hot to cold resistance ratio. determined initial reading and current drawn when hot, which can be worked back to find hot resistance.

I have shown this to another seasoned engineer like myself and he is also puzzled. Together we have over 80 years of experience designing and troubleshooting all types of power circuitry.

Barkhousen noise is not a good fit for what is happening here.

At this point I'm wasting too much time answering questions that were priorly answered and am going back to the bench to investigate further. Read the whole thread!!!

Cheers to all.....ION

Ok further tests on the bench this morning revealed some of the truth.

EM was right, this is a saturation effect of the type of Nichrome that has a high iron content.

I tore apart another jnfra-red quartz heater and laid the element on the bench so that I could slip sense coils over it.

This heater had weak but more noticeable attraction to a magnet.

The effect can now be seen sharply and in phase with the zero crossing. It is noticeable at much lower  current levels as I ran it off a variac. More pronounced at higher currents, but always sharply at the zero crossing, and a very large spike.

I cannot explain the many spikes seen in the earlier analog scope shots and will research it further.

I wonder if pure iron wire will have the same effect? Worthy of research.

Back to the bench!

 ;D  O0

Quote from: ION on 2012.08.10, 12:56:07
...
If you have not read the whole thread, you will note that I have used a 0.01 ohm resistor shunt to measure the current. This is a four wire "Kelvin" type resistor, 2 leads for current, 2 sense leads to the scope. No artefacts in the current waveform

I tried another similar heater, coiled Nichrome element in a quartz tube. same effect a little less intense.

I am guessing they are Nichrome. based on the hot to cold resistance ratio. determined initial reading and current drawn when hot, which can be worked back to find hot resistance.

I have shown this to another seasoned engineer like myself and he is also puzzled. Together we have over 80 years of experience designing and troubleshooting all types of power circuitry.

Thank you for having clarified and synthetized your results O0, and also your experience and level of expertise, this will avoid me now to propose the simplest explanations that qualified people like you have already discarded but that can constitute traps for the laymen.

Quote
At this point I'm wasting too much time answering questions that were priorly answered and am going back to the bench to investigate further. Read the whole thread!!!

Cheers to all.....ION

I understand. I confess that I had not carefully enough read the whole thread. I don't well master English and so, it's tedious for me to find the good information when it is scattered throughout the thread, and I miss some points. That's my poor excuse  :-[.

EX:

No problem, I consider you one of the brighter guys out there, you ask the right questions and are obviously well experienced in the art of electronics.

I also ran the heater from a high power amplifier, sweeping frequency to about 10K. The saturation spike gets higher and sharper with increasing frequency up to a point, then rounds of a bit.

I need to go back at some point and check that first "Martel" heater, the one that produced the myriad of pulses on the analog scope. I bypassed the bi-metal thermostat and the tilt shutoff switch, but the noisy pulses were still visible.

Tested also a coil of pure thermocouple grade iron wire that I had wound a few years ago. It does not exhibit anything like the Nichrome effect.

Worthy of research would be a test of many different types of pure wire and wire alloys to determine the magnetic effects. Wish I had the time and materials on hand.

Quote from: ION on 2012.08.10, 19:20:07
...
Worthy of research would be a test of many different types of pure wire and wire alloys to determine the magnetic effects. Wish I had the time and materials on hand.

This is a good way for the investigation.

The effect could be due to the temperature of the wire that oscillates at each half-period before reaching its nominal value, either around the Curie point (presuming NiCr weakly magnetic) or simply changing the permeability or less likely, inducing a phase change in the metal.
Therefore it would be depending on metals or alloys.
In the past I had modulated a filament bulb with AC audio around a CC current and was able to demodulate the light signal that I received with a LDR. There was a low-pass filter effect but the voice was perfectly understandable. This demonstrates that the filament temperature can change at a rapid rate up to Khz, contrarily to our intuition. According to your oscilloscope screen shots and the width of the spikes (around 1/20 of the 60Hz period), their frequency domain is about 1200 Hz.

I will try to get some nichrome wire, i should be able to drive this from my Class A.

I am running out of time for this month my Daughter is about to get married in a couple of weeks.  O0

Here is some additional information that adds a bit of mystery to this finding.

I wound two long thin coils like the ones in the Martel heater for testing.

One was made of thermocouple grade iron 14 guage. This is a pure soft iron. It should be easy to saturate.

It did NOT put out a saturation pulse anything like the Nichrome. In fact there was no sign of any saturation effect. The mag probe and other pickup loops showed a well rounded sine wave at several test currents.

The second coil was made of a thermocouple wire material called Constantan, 45% Nickel, 55% Copper.

This coil also showed no pulse effect like the Nichrome.

I have some Chromel on hand which is 90% Nickel, 10% Chromium. I will try to test this next.

Looking at my inventory, I have a large stock of Nichrome V made by Driver Harris Corp. I will have to research this further, but it is unlike the Nichrome in the heaters, very weakly magnetic.

Another set of tests would be to wind these wires into a given toroidal shape and use as an inductor core to determine saturation effects when current is not passed through the wire.

Too much research to do, too few hands.....anyone want to join the fun?

Quote from: Peterae on 2012.08.11, 12:44:14
I will try to get some nichrome wire, i should be able to drive this from my Class A.

I am running out of time for this month my Daughter is about to get married in a couple of weeks.  O0

Peterae:

Hope you have a good time at the wedding O0

Before you buy any Nichrome, we need to identify the exact type used in these heaters.

If you could use a heater around the house there are several vintage types on ebay, as well as replacement elements.

Looking at my inventory sheet reveals that the Nichrome I have on hand is 80% Nickel, 20% Chromium with temper "A" made by Driver Harris Corp. It is NOT what is used in the heaters.

The heater stuff has a very noticeable (not weak) attraction to a magnet, It must be the 24% iron type.

Wiretronic has a free downloadable wire calculator here:

http://www.wiretron.com/free.html

This site seems to indicate that Nichrome 60 Alloy 675 is predominately used in infra red heaters.

http://www.infraredheaters.com/nicrcoil.html

It has the following: Nickel 57 to 58% Chromium 16% Silicon 1.5% Remainder Iron (about 24%)

Thanks ION very handy wire calculator  O0

I just looked high and low  :) and located some very old replacement Halogen quartz heater spares, i knew they were around somewhere, i am not sure if they are any good, but worth a try, if i shake them i can hear the coiled element ratling, as they are sealed and halogen i guess i could break one to have a better look at the wire infact looks like 1 is broken anyway.
They are 400 Watt 240V 1.7 Amps, now i wish i had bought a bigger variac which is only rated for 0.75A Grrr, still it might warm it up enough to have a sniff at the field.

Exactly the same as these here
http://www.amazon.co.uk/Halogen-Heater-Replacement-Bulbs-400Watts/dp/B000R2CSDS/ref=pd_sim_sbs_kh_2

the heater element is only about 5-6mm OD spiral wound.

I don't think those quartz heaters will do it, they are usually tungsten wire, but I will test a similar item.

Here is the stuff:

http://www.ebay.com/itm/Nichrome-resistance-wire-18-AWG-gage-30-feet-foam-cutting-sealing-heating-/221087194159?pt=LH_DefaultDomain_0&hash=item3379d2dc2f

This has the 24% iron.

You may be able to get a free sample from a wire house.

Too bad I have in stock over 6000 feet of the wrong stuff, the 80/20.

The pulse you will see with the right stuff is pronounced, very interesting, and defies all theories so far.

I thought (and agreed with EM) that it was simply saturation effect, but now am doubting this theory based on tests with the pure iron wire.

Something about that alloy Type 60 that is different. I wonder what would happen if we designed the alloy for the optimum pulse?

yep i have some on order  O0
I have a lot of things i want to test out, IE pulse 1 with 1000V pulses using my Marx avalanche pulser, resonate a tuned coil made of the stuff etc and yes as always i am getting ahead of myself  :) first i need to see the effect  O0

I am just limited by time for a few weeks so it will appear not much is happening  :-\

I will try one and see if i can get any results while waiting for the nichrome to arrive.

Quote from: ION on 2012.08.09, 14:37:12
Here are two camera shots of the scope traces. Forgive the fuzziness. Smooth sine curve is the heater current. Spike waveform is superimposed.


First shot (001) is spikes as picked up with mag probe full amplitude at initial time up to 5 seconds vs. current.

Second shot (002) shows spikes decaying after 5 seconds versus current.

Judging from the scope shots, it appears the spikes are produced both when the current quickly changes from increasing to decreasing, and when the current changes direction at the zero-crossings (more pronounced). Notice the polarity of the spike is the same for when the current goes from increasing to decreasing, and from positive to negative at the zero-crossing? The spike is the opposite polarity when changing from negative to positive current at the zero-crossing. This sounds very much like inductive kickback. A very small change in current seems to be causing this IK. Could this be? Could the heating/diminishing effect be caused by the added resistance largely damping out the spike's amplitude?

QuoteThe most common way for a toaster to create the infrared radiation is to use nichrome wire wrapped back and forth across a mica sheet, like this:

It looks like toasters use nichrome as well, infra red radiation  O0

Quote from: poynt99 on 2012.08.11, 14:58:09
Judging from the scope shots, it appears the spikes are produced both when the current quickly changes from increasing to decreasing, and when the current changes direction at the zero-crossings (more pronounced). Notice the polarity of the spike is the same for when the current goes from increasing to decreasing, and from positive to negative at the zero-crossing? The spike is the opposite polarity when changing from negative to positive current at the zero-crossing. This sounds very much like inductive kickback. A very small change in current seems to be causing this IK. Could this be? Could the heating/diminishing effect be caused by the added resistance largely damping out the spike's amplitude?

Hi Poynt

Interesting observation, yes it does seem to possibly be inductive kickback, but why is it nonexistent in pure iron wire? And why so sharply sensitive to change in current and direction?

The resistance only changes about 5 to 6% from cold to glowing orange, even less over the sharp point at which the decay occurs. (11.07 amps).

I think we have identified the Curie  point and fading ferromagnetic property of iron as the cause of the sudden decay (at around 11.07 amps on my heater). The time to decay is quicker if the element has not cooled all the way to room temp.

Quote from: ION on 2012.08.11, 15:27:22
Hi Poynt

Interesting observation, yes it does seem to possibly be inductive kickback, but why is it nonexistent in pure iron wire? And why so sharply sensitive to change in current and direction?

I'm speculating that is has something to do with the alloy mix. Pure iron as a magnetic core has a relatively poor high frequency response. It would be interesting to try your experiment with Permalloy tape, or something with a similar alloy mix.

Quote from: poynt99 on 2012.08.11, 16:32:20
I'm speculating that is has something to do with the alloy mix. Pure iron as a magnetic core has a relatively poor high frequency response. It would be interesting to try your experiment with Permalloy tape, or something with a similar alloy mix.

Yes the alloy is the key, but then the question is "what is the optimum alloy"?

I'm not sure if I have Permalloy tape on hand, I have a few unidentifiable tapes.

1st attempt with limited equipment and a foot high buried bench  C.C
My variac just wont handle the power of this tube, i am still running with a 100 Watt bulb in series with my variac to save my day.
I have wound a coil over the quartz this has a 1K resistor in parallel just to help look for a bit of power in the spikes.

I am scoping across the 400watt heater with Chan 2 and Chan1 is the sniffer coil.

All i am getting is dc pulses when i switch on and off.

I am not sure what my next move is, i could connect straight to the mains but it will burn hot real fast.
Maybe i could take the element out and measure the length/resistance and try to work out which wire it is and then maybe use a small amount of wire to hook up to my 15 Watt class A

Peterae:

I've already tested a quartz heater (sealed glass tube type) as used in fuser assembly of laser printer.

There was nothing unusual to find.

Tungsten filaments as used in this type of heater have a large positive temperature coefficient (PTC).

I would be very careful trying to drive this from your class A. The low cold resistance will cook your transistors fast.

Better to use your Variac and a step down transformer.

BTW, I built the JLH class A many years ago. It is a fine little amp. I was an avid reader of WW. JLH will be missed.

If you measure the ohms cold, then work backwards from the amps draw hot, knowing the applied voltage you can determine the hot resistance easily.

Hey that's great, it's a fantastic amp, i did the low ohms mod and drive a 1.2 ohm secondary of a mains transformer and use the other 1.2 ohm secondary to drive my heater, it has no trouble with low impedance drive.

I just cut a meter of wire from that halogen tube and it reads 5.4 Ohm doh which is useless without me knowing the diameter, looks like i am stuck until i get some nichrome through the post.
I just tried a neo magnet near it and i cannot detect any attraction, if it does attract its a very small amount.

EDIT warning welcomed , would hate to blow my amp, at some point i want to pipe some audio through it  :)
So yes indeed looks like i got Tungsten

Now who do i know who has a heater i can borrow and strip >:-)
I wonder if the Mrs hair dryer might do it  :) although i could be dicing with death if she See's that apart C.C

Rapid test with my toaster today. The heating wire is not wound but in straight lines. I don't know which alloy it is made of.
The oscilloscope strobe was connected to a pick-up coil put near one of the wire line.

No spikes but the trace shape is not really a sine. There are flat steps each side of the zero crossing and at some other places.

Quote from: exnihiloest on 2012.08.12, 10:03:54
Rapid test with my toaster today. The heating wire is not wound but in straight lines. I don't know which alloy it is made of.
The oscilloscope strobe was connected to a pick-up coil put near one of the wire line.

No spikes but the trace shape is not really a sine. There are flat steps each side of the zero crossing and at some other places.

What type of pickup coil did you use? Also, depending on how the toaster elements are wound ( WW vs coiled) you may find the field cancels. It is hard to get at toaster elements with a probe.

Recently I have been researching building a low cost wide band mag probe. Attached is a good paper on the subject, but I'm looking for something simpler.

Some folks use a cheap telephone audio pickup coil (1 mV per 20 mGauss) shown here: http://www.lessemf.com/kits.html

I am tempted to knock up a circuit i built some years ago.
2 back to back ugn3503 hall sensors each one feeding inverting and non inverting leg of an op amp, add gain, for fun i added a vco and it worked as an audible compass and was quiet sensitive ;)

PS i have on order 4 different sizes of Nichrome now.

Quote from: Peterae on 2012.08.12, 13:13:46
I am tempted to knock up a circuit i built some years ago.
2 back to back ugn3503 hall sensors each one feeding inverting and non inverting leg of an op amp, add gain, for fun i added a vco and it worked as an audible compass and was quiet sensitive ;)

PS i have on order 4 different sizes of Nichrome now.

Yes, I've considered linear Hall effect sensors, and should have a few of those in the bin.

Size of the wire may not be important, composition definitely is.

Another simple mag probe paper attached:

Quote from: ION on 2012.08.12, 12:40:32
What type of pickup coil did you use?

I don't know exactly. It was the coil of an excellent ferrite transformer used for audio signals in professional electronic cards from phone switching centers, dating from the analog time. I have plenty of these. I had removed the ferrite. Length and mean diameter are both about 1.5 cm.

Quote
Also, depending on how the toaster elements are wound ( WW vs coiled) you may find the field cancels. It is hard to get at toaster elements with a probe.

I agree. I had to set the oscilloscope for 2mV, with probe 1x. The signal was around 8 mV when the coil was positionned inside the toaster, near the wire.

Quote
Recently I have been researching building a low cost wide band mag probe. Attached is a good paper on the subject, but I'm looking for something simpler.

Some folks use a cheap telephone audio pickup coil (1 mV per 20 mGauss) shown here: http://www.lessemf.com/kits.html

My coil is not optimized for 50 hz (the mains here), unlike this at your link, but it should show the spikes if they existed. I presume that my wire is not of the right alloy.

@Ion
You have to separate the fields in order to understand exactly what is going on, Electromagnetic = Electric + Magnetic. I use a dual polarity single analog output hall effect sensor to measure only the magnet fields and a DIY dual polarity single analog output electrometer to measure only the electric fields. If you do this it won't take very long to understand what you have been missing as a tickler coil is basically like groping around in the dark, it is archaic.

Regards

omg i just looked up what they use for wirewound resistors and the wiki says nichrome  O0 not sure what grade yet though.

New test with a 2 KW heater. I have disconnected the fan that induced a signal in the coil. But now my tests are short otherwise the breaker interrupts the current.

There are spikes. The signal was weak and noisy, weaker than with the toaster. The heating wire is obviously wound in order to minimize the magnetic field.

So I have wedged a narrow ferrite rod in a curl of the heater wire and the other end inside the coil. Now the signal is strong, around 20 mV.
I confirm the spikes.
They represent about 20% of the mains signal, and seem weaker than in ION's screen shots. The difference could be due to the nature of the alloy. Here the heating wire is rather strongly attracted by neodymium magnets.

I also carried out a direct current measurement by connecting 2 probes to 2 points of the heater wire, close to each other, and by monitoring the voltage with the oscilloscope in differential mode: the current shows no spikes, it is a smooth sine shape.

There is really a non-linear magnetic effect. The magnetic field is not proportional to the current. This is puzzling.

I can also confirm the pulses, i found an old 1K 25 watt wire wound pot, and can detect the pulse using my hall, but it's only there for a second or so the pot heats up pretty fast, next task in to film it  O0

As usual i screwed up something showed up which looked exactly like the pulse, it disappeared after a couple of seconds when the pot warmed up, the pulse was in phase with the ac waveform across the pot, but then i realized i could pick the pulse up anywhere round my bench, so powered off the variac it goes away, switched back on and it's amplitude varies with the variac setting but the hall sensor is no where near, so i disconnect the pot, blow me it's still there again varies with variac setting now this means the variac has no load, what the hell, I'm getting some sort of earth loop between my hall psu and the variac which are tied together via the scope earths, even though my psu is not earthed and uses a transformer?
The strange thing about it is that it goes after a couple of seconds when the variac has a load, it's definitely some sort of earth loop though because i see a little spark when i connect my scope earth to one side of the variac.

Looks like i will need to sit this out until the wire arrives.

This is probably a total waste of time but Just in case this means something i have filmed what i am seeing, the hall is not placed near anything of interest while scoping except when i place the north or south pole of a horseshoe magnet near it.


Having thought about the hall, when the north is used it drives the hall to the positive rail and when south is used then it drives to the 0V rail, so that explains that i think, so it's noise on my rail or a ground current as i said earlier, i wonder why it disappears after a second when the variac has a load.

I have recorded with the sound card the signal from the coil probe.
MP3 and Adobe audition screen shots with the spectrum in the attached files.

(http://www.overunityresearch.com/index.php?action=dlattach;topic=1505.0;attach=8346;image)

Nice work ex thanks for posting that  O0 very interesting.

Ditto on that. I'm glad that you were able to get some degree of verification.

I took the grille cover off my heater and was able to better position the probe.

Now I can see extremely large sharp spikes at the zero cross both plus and minus transitions.

Soon hope to post a video of the scope traces and with audio of the heater.

Interesting work, folks  :)

Anyone considered Barkhausen jumps for the zero crossing transition and the approach to and recession from peak?

This effect is seen in ferrous electrical conductors and tends to smooth out as the wire reaches a certain level of heat.

Edit>>

Know that I think about it.... most of the heating elements discussed will have a fair amount of nickel in the alloy. This means that magnetostriction may also be playing a part.

Hi WW

Barkhousen jumps are a very fine graininess to a changing magnetic field. I've never seen anything near as large as these spikes.

We have covered that in prior posts.

Read from the beginning page for some fun and puzzlement.

Ah...


Almost never having the time to read more than a page or two, I'll just refrain from posting and see what your conclusions come to.

Cheers  O0

Quote from: ION on 2012.08.14, 01:23:45
...
Barkhousen jumps are a very fine graininess to a changing magnetic field. I've never seen anything near as large as these spikes.
...

I agree. The Barhausen effect is even difficult to detect.
We must also keep in mind that here, the heating wire is not a coil with a ferromagnetic core, neither it constitutes the core of the coil probe.
The heating wire is itself the "core" of its own winding, but a rather thin core with a not colinear magnetic field, so its ferromagnetism should influence both the heating winding and the coil probe only slightly.

I would rather see a hysteresis effect. When the threshold points are attained, there is a strong dB/dt, especially if the material has a high coercitivity.
To be confirmed.

Today I have powered my heater with a variac. At 150v (instead of 230v) the disappearing of the 50 Hz even harmonics doesn't occur. The breaker interrupts the circuit before.

Ah my wire just arrived i have 4 sizes 20,24,26 & 32 SWG Gauges, i still have 1 on order from a different supplier yet to come.

The wire is Nickel (approx. 60%) and Chromium (approx. 15%)
Not sure what the plan is, as from my last test i need some mains isolation if i use my variac, i maybe better off trying to drive from the Class A or driving the hall from a battery.
So i have 15 watts Class A if i go for 10V pk-pk i am looking at 1.5Amps with a required resistance of 6.666 Ohms.

So
Gauge: 20SWG Resistance per Metre (Ohms): 1.707             390 cm Don't have enough of this 1
Gauge: 24SWG Resistance per Metre (Ohms): 4.563             146 cm
Gauge: 26SWG Resistance per Metre (Ohms): 6.828             97.6 cm Length required
Gauge: 32SWG Resistance per Metre (Ohms): 18.99             35.1 cm

Quote from: Peterae on 2012.08.14, 10:50:08

The wire is Nickel (approx. 60%) and Chromium (approx. 15%)
Not sure what the plan is, as from my last test i need some mains isolation if i use my variac, i maybe better off trying to drive from the Class A or driving the hall from a battery.
So i have 15 watts Class A if i go for 10V pk-pk i am looking at 1.5Amps with a required resistance of 6.666 Ohms.

I hope the remaining 24% is iron. It should also have about 1.5% Silicon.

My recommendation would be to use a step down transformer after your Variac so you can boost the current and lower the voltage. A 24 volt secondary or thereabouts should suffice.

The effect is noticeable at lower currents than the 11 amps I have been using off the mains. I was able to see it at 2 to 3 amps when driven from a Variac.

This 80 / 20 Nichrome I have on hand is hard to work with, almost like spring steel and does not produce the effect.

Hopefully the iron content in yours will make it easier to wind. Tightly wind on a 0.375 wooden dowel, then you can remove the dowel.

The number of turns is to your liking, with consideration to how you intend to drive it.

Best of luck.

EX:

You are correct and have obviously done the research. Barkhousen jumps are visible in any soft steel or transformer core, but as you say they are difficult to detect.

Try driving your heater from a step down transformer so you don't keep blowing the breaker.

Well that non existant iron quote did get me a little bothered but it must be standard available wire they supply and looking at all the other specs from other suppliers the remaining is normally iron with a little silicon. If i do not see the effect then i shall look a little harder in to this, it's cheap stuff and he was a fast supplier that's why i choose him.

Thanks i will see what transformers i have available.

EDIT just tried bending a piece, seems to bend fairly easily  O0

Quote from: ION on 2012.08.14, 11:45:17
...
EX:
...
Try driving your heater from a step down transformer so you don't keep blowing the breaker.

I did it (see my previous post).
I have also improved the measurement with a common mode rejection. The mains current is full of pulses, noise, HF... that has upset me for a while. Now the signal is cleaner and I get only one pulse at each half-period. The pulses are very strong in comparisson with the mains frequency, but I guess that it is due to a high-pass filter effect of the measurement circuit.

Here  is a frame capture from a video of the effect. The video was too large (100meg) to upload. As POYNT and EX have shown, the small blips line up with the descent from peak of the AC current waveform i.e. when the current is changing amplitude, while the large spikes line up with the zero crossings, an actual change of direction of current.

Also included is a wave file where I plugged the magnetic probe into the microphone input and recorded it.

I had to normalize the gain as the signal was in the mud. You can view it with any wave editor as if it was a scope log.

When you listen to the wave file it is rich with harmonics.

Just use zoom in to see the waveform.

I've been thinking ahead on how to use this spike.   It's nice to know what causes it, and to maximize it, but what can it do for us?   

Here's what I've been thinking.

We can bring a tuned coil and capacitor close to it so that it gets stimulated by the rapid rise time of the snaping magnetic domains.   Better yet:    create the primary of a TESLA coil out of this nichrome wire, and than tune the TESLA secondary to be resonant at a harmonic of it, than everytime the pulse comes it adds more energy to the tank circuit formed by the secondary, and it will be in phase with it and keep it going.

who's going to be first to achive this?   

Just think about it, no need for any electronics, just nichrome wire for the primary and wire and a hat for the secondary, how much simpler can it get?    O0

EM 

Well still no results so far

I tried 1 meter of 32 SWG wire and wound it round a pencil.
I found a 0-25 0-25 transformer so wired these in series, i have now bypassed the 100 watt series bulb to give full mains.
Using my hall effect on a x1 probe to give me 10mV resolution god this scope is really noisy down this low setting.
Anyway the coil starts glowing orange after a about 5 seconds of the variac being up full.

Here's 2 photo's
1 of the coil and 1 of the scope across the coil and of the hall output.

I have not really noticed any audible noise either.

I need to yet check the wire resistance and use a magnet to see how magnetic the wire is, i will append that info to this post as i get it  O0

EDIT just an after thought, do you see the effect with the heater orientation to earth changed  8)

QuoteUsing my hall effect on a x1 probe ...

that's one of the problems, you won't see these with a magnetic field probe like a hall effect sensor.  

these spikes are not spikes in the magnetic field, but spikes in voltage induced in a pickoff coil, due to a RAPID but SMALL magnetization jump.

EM  

OK well in that case whats the best construction, i want a very narrow wound coil with roughly how many turns?

The thing with the hall is that i can push it in between the turns of the iron heater coil to maximize pickup

How about a low volts relay coil with the core taken out, although this will only sit near the heater coil. not inbetween turns

I'm not sure what ION is using, but you should be able to see the spikes with a small coil, tucked between the turns of the nichrome coil, and with a larger pickoof coil placed at one end of the nichrome coil, all without amplification.  But you might have to have some amplification to see anything depending on how strong the signal is.     Do they even make small pickof coils, the size of a hall IC chip?   

I will reiterate for the sake of those that may have missed earlier posts, but best to go back to page one for full details:

You don't need to put the pickup in between the turns. Most of the scope shots I posted had the mag probe at a right angle to the end of the coil or parallel to the end of the coil, and up to 1.5 inches away.

It greatly diminishes as you approach the center of the coil.

This also showed up readily with an old long relay coil with a soft iron slug through the center as a pickup device.

I also placed the heater into a large 6" hoop with about 50 turns of fine wire, with no magnetic core the pulse is easily visible.

I'm sure the Magnetek probe I am using is just a small slug of iron or ferrite with maybe 1000 turns of fine wire and perhaps a loading resistor. It is weakly attracted to a magnet. The probe shows orientation arrows along it's length. It reads 1.52k ohms.

Once the coil heats up the pulse disappears. It may be difficult to see it again until it is back down in temperature.

I get the pulse up until the point where the coil glows orange then it fades into the noise as the buzzing stops.

FYI the pulse reads about 350mV with this probe about a half inch from the end of the heater coil. This is equivalent to around 3.5 Gauss (with this probe, 100mV / Gauss), not very much. e.g. A large, linear wall wart produces about the same level at that distance.

Peterae said:

QuoteEDIT just an after thought, do you see the effect with the heater orientation to earth changed

How about a low volts relay coil with the core taken out, although this will only sit near the heater coil. not in between turns?

Orientation to earth does not matter.

The relay coil will work with or without core.

You seem to be getting plenty of low level noise from you hall sensor, but it looks random, nothing like the spikes I'm seeing. Strange.

We need to calibrate our probes before we can make meaningful measurements. Can you bring the hall effect sensors near a known low level magnetic field to get a rough idea if your pickup is linear and good at low levels?

Ion,

Peter needs to use a pickup coil like you, not a hall sensor.    A pickup coil responds to the RATE of CHANGE of the magnetic field, and the more rapid, the larger the signal induced.    On the other hand, the hall effect sensor just resonds to the MAGNITUDE of the magnetic field.   These spikes that you are observing are due to small magnetic changes that occur below the resolution of his sensor, or in the noise, so he won't be able to observe them properly.    Assuming he had a stronger signal, the hall effect would show a step response in the magnitude of the field, so where that step occurs, that's where a pickup coil would show the voltage spike.   One waveform is basicaly the derivative of the other, that's the difference.


Question:   have you tried a low current setting,  so that the nichrome coil produces those spikes all the time (maintain the temperature below the Curie temp)?   If the spikes are still present with such a low current level, than that's great.

EM     

EM:

Yes I have the effect at lower current levels as I stated before, in the 2 to 3 amps range.

Also I doubt it would be very efficient as a means to perturb into resonance the secondary of a "Tesla coil" as the effect is weak compared to the expenditure of energy.

My main interest in this is to determine if the unique alloy produces an avalanche effect of domain flipping, and how can we determine an alloy to optimize the effect. Pure iron wire has no such effect. Maybe it has too much remanence due to it's purity and cannot change magnetization that quickly.

If Floyd Sweet was real and his VTA did indeed perform macroscopic domain flipping with a small excitation signal and only after conditioning the material, we may have an area ripe for more research. Just guessing.

Re 32 SWG 1 meter length wire test.

Heater element is 19 Ohms and weakly attracted to a magnet.
Using the wire calculator you linked earlier it calculates using my wire diameter that a ft of Nichrome C is  5.800562 Ohm/Ft i measure 5.79119 Ohm/Ft it's the closest mach on that calculator.

I have switched to a relay coil now, with an iron core i am using my other bench scope which goes down to 5mV/Div

It looks like it is trying to show me a pulse as i have a little wiggle, also the waves are not in phase?

Shall i try another pickup coil without iron core or try thicker heater wire.

EDIT

I managed to get the relay coil closer and better aligned, the little pulses do not change from cold to hot.

Quote from: ION on 2012.08.15, 01:38:13
...
My main interest in this is to determine if the unique alloy produces an avalanche effect of domain flipping,
...

I have an annoying problem. Even if there is an "avalanche effect of domain flipping", how could it be viewed by the probe? The alloy doesn't constitute the core of a coil, it is the wire itself, and so, it can be influenced only weakly by its own magnetic field which is not along the wire but in circles around the center of the conductor, inside and outside.
In my heater, the wire doesn't constitute a coil. It has the shape of a 50 branches star, each branch being formed by the wire going to and from the tip, in parallel. So the magnetic field should have a very low level inside the wire. Therefore I wonder how it can flip the domains. But if the domain flipping is the real explanation (and now we have only this one), considering the weak coupling of the magnetic field to the magnetic matter, considerable optimization of the process could be done.

Quote from: exnihiloest on 2012.08.15, 10:53:32
I have an annoying problem. Even if there is an "avalanche effect of domain flipping", how could it be viewed by the probe?

QuoteThe alloy doesn't constitute the core of a coil,

it is the wire itself, and so, it can be influenced only weakly by its own magnetic field which is not along the wire but in circles around the center of the conductor, inside and outside.
In my heater, the wire doesn't constitute a coil. It has the shape of a 50 branches star, each branch being formed by the wire going to and from the tip, in parallel. So the magnetic field should have a very low level inside the wire. Therefore I wonder how it can flip the domains. But if the domain flipping is the real explanation (and now we have only this one), considering the weak coupling of the magnetic field to the magnetic matter, considerable optimization of the process could be done.

We can form the wire types into a toroidal  core or open  "I" section core, and wind over it with copper wire.

In the case of the toroidal core we can wind the wire on a small spool like a tape wound core. Then through the center of the spool with our copper wire. This may be a more meaningful way to test the materials magnetic properties.

For the open I section, we could cut the wire into 2-3 inch lengths and pack them into a small cardboard tube, then over wind the tube with copper.

Peterae: It is puzzling to me that you are not seeing the spikes I am seeing. Maybe thicker wire will help.

The relay coil should work with or without the core. It will be more sensitive with a core, but the magnetic core will add it's own signature though small.

I apologize if you feel you are going down a rabbit hole with this.

I have tested power wirewound resistances like this one:
(http://img.directindustry.com/images_di/photo-m2/wire-wound-power-resistors-14137-2404809.jpg)

I have near two tens of them with different values. They don't work. I only observe almost flat steps each side of the zero crossing. No pulse.
According to http://www.resistorguide.com/wirewound-resistor/, the wire is NiCr. My resistances date the 70's, maybe not the same metal as today resistances. The wire is not attracted by neodymium magnets.

It is really important to clarify which alloy allows the effect. This surely explains the not conclusive result from Peterae, as well as the negative result with my toaster but positive with my heater.

I am happy to keep trying different things and wire if need be  O0

certainly must be the alloy i think, i will run tests on the other sizes from this supplier first and then i am still waiting for some from a different supplier, we were working on the basis the correct alloy was Nickle 60% Chrome 40% with 20-25 Iron

I am just going back to the beginning of the thread to find out if ION has successfully wound one from new wire that works or wether it is a ready made heater only so far that works.

Maybe the wire needs to age and have a good coat of oxides.

EDIT you know when you compare ex's picture of the waveform and the one in my last video i can see little wisps above the pk of the waveform just like ex but smaller, yet mine do not go when hot but infact match his picture when his coil is hot, so maybe my coil heats too fast and they go before i get a chance to see them, if this is the case i need thicker wire and more mass to slow the heating process, i dont want to turn down the supply because my coil is only just picking anything up.

Peterae:

Quick recap:

I tested 3 infra red heaters from different manufacturers, they all had the effect.
All were an unidentified alloy loosely coiled in a quartz tube. All elements were mildly magnetic.

I wound a coil with 80% Nickel 20% Chromium and there was no effect.

Same with pure iron...no effect
Same with 45% Nickel 55% Copper...no effect

ex:

There is a special alloy of Nichrome used for resistance wire where the tempco is important, but I believe there is no iron content in it.  Manganin and Constantan are two that I know of. There are many names in the trade for slightly different alloys, but Nichrome heater wire has a large tempco by comparison to resistor wire where temperature stability is important.

Here's 24 SWG much thicker

I've got to say this wire is quiet malleable, it winds quiet nicely on a pencil much much softer than spring steel for instance, i still get no sound when current flows, the pick up coil is picking up much better so the current is a lot larger and as it's thicker it takes longer to glow but still no sign yet, scope is now 20mV/Div as opposed to 5mV/Div in the previous video.



EDIT
I Just tried the thickest wire i have at 20SWG no sign of spike when cold, eventually popped my primary fuse 800mA to protect my 750mA variac, no point in me trying this gauge any longer just in case my variac pops.

I have found out, most of the heaters using the cheapest 60% Ni, 16% Cr, and the rest of Fe heating wire. I'm trying to get from this.

Nickel ferromagnetic at room temperatures, become non magnetic  at 355 celsius

Chromium antiferromagnetic, become paramegnetic at 34,85 celsius

Iron ferromagnetic, become paramagnetic at 770 celsius

Nice mixture of metals. I think as it has something to do with the temperature, would be good to know at which temperature the wire reach, when the pulses disappear.

ION,

can you try to capture one of the voltage spikes at the zero crossing,  with the scope timebase set to the ns scale.   I'm thinking that these large voltage spikes are actually many many Barkhausen noise spikes that occur all about the same time, and at the time base setting you're looking at they all show up as one large spike.       

EM

Peterae:
Can you give the ebay ad links where you bought the wire? Something strange going on, I'd like to see if they may have made a a substitution.

EM:

Will do as time permits.

I contacted the people i bought the wire from to ask for a data sheet, see attached.
I think i can say i can rule this alloy out.

So we have
Nominal Composition
Nickel 60%
Chromium 15%
Manganese 1.5%
Silicon 1.5%
Iron Balance 22%
Aluminium --

This is a quote from ION

QuoteThe heater stuff has a very noticeable (not weak) attraction to a magnet, It must be the 24% iron type.

Ha ha
My wire has a very weak attraction(well i tested the thin stuff, must try the thicker)  to a magnet  O0 now that is weird considering it has 22%

OK just checked wiretron.com

Nichrome C does indeed have a faint attraction to a magnet so it looks to me that your wire is not Nichrome C or Nichrome A as this has no magnetic attraction.

We need a resistance per foot reading on your wire and diameter measurement, this should identify it.  O0

As it will be coiled we could estimate the wire length by measuring the circumference of each turn to get wire length per turn and then measure the total resistance and how many turns it has.

Removing the heater from the frame I obtained the following:

Dia of wire: 0.052"

Dia of wind: 0.339"

Turns per inch: 19

Overall length of all winding minus terminations: 20.15"

Length of wire: 33.98 ft. (calculated)

Measured resistance: 9.1 ohms.

This is a bit higher than 8.48 ohms for type C or 8.16 ohms for type A.

Visual inspection shows there has been some oxidation raising the resistance so it is impossible to tell the alloy.

Testing with the quartz sheet off shows a reasonably strong magnetic attraction.

That's great news ION thanks for doing these tests, i just tested my thick wire with a neo magnet and i can hardly detect an attraction, very weak.

So either this wire is not C but i did check the resistance and it's pretty dam close, and the wiretron site also says type C is very weak magnetic attraction, i think this is all saying your wire is definitely not type C

and because the resistance is slightly lower than it should be i think there's only one on wiretrons site that fits and that is
Alloy 875
22.5% chromium
5.5% Alu
0.5% silicon
0.1% Carbon
71.4% Iron

The only fly in the ointment is i bet there's load's more alloys made by other companies as well  :'(

QuoteAlloy 875
Also called Kanthal A1, Kanthal, Alloy 875, Resistohm 145, Aluchrom 0, Alchrome 875, MWS-875, Stablohm 875

Here's some resistance data on major alloys
http://www.mwswire.com/resist3.htm

So using this wire data you have approx wire gauge 16 and you have about 0.2679 ohms/foot

looking at the data the nearest i can see is MWS-675

then looking up the properties MWS-675 only has slight magnetic attraction so it's not this one either.

that then leads us to MWS-800 or MWS-875
and checking the properties of these MWS-800 has no attraction and MWS-875 has a strong attraction

going down the resistance value order the next one we hit with strong magnetic attraction we get Alloy 42 , but the resistance of this is nearly 45% less than our measured value.

Conclusion it has to be Alloy 875  O0

Alloy 875 is cheap to make
http://tankii.en.alibaba.com/product/343122400-209485507/Alloy_875_Resistance_Wire.html

QuoteDetailed Product Description

Alloy 875, Also called Kanthal A1, Kanthal, Alloy 875, Resistohm 145, Aluchrom 0, Alchrome 875, MWS-875, Stablohm 875

High-temperature Heating Alloy TK1



TK1 large-sized cold-drawn wire products can be used for high-temperature resistance furnace. Practice has proved that: the product process is stable, integrated performance is good. Has good high temperature oxidation resistance and longer service life; excellent winding properties at room temperature processing, ease of processing molding; little rebound resilience and so on. Processing performance is better than 0Cr27Al7Mo2, high temperature performance is better than 0Cr21Al6Nb; operating temperature can reach 1400 °C. It performs better than Kanthal A-1 in some places.



Main specifications and uses:

Conventional Product specifications: 0.5 ~ 10 mm



Uses: mainly used in powder metallurgy furnace, diffusion furnace, radiant tube heater and all kinds of high-temperature furnace heating body.



MAIN CHEMICAL ELEMENTS AND PROPERTIES

Properties \ Grade
   

TK1

Cr
   

Al
   

Re
   

Fe

25.0
   

6.0
   

Suitable
   

Balance

Max Continuous Service Temperature(ºC)
   

Diameter 1.0-3.0
   

Diameter>3.0,

1225-1350°C
   

1400°C

Resisivity 20ºC (Ωmm2/m)
   

1.45

Density(g/cm3)
   

7.1

Approximate Melting Point( ºC)
   

1500

Elongation (%)
   

16-33

Repeatedly Bend Frequency(F/R) 20°C
   

7-12

Continuous Service Time
   

>60/1350

Micrographic Structure
   

Ferrite

You are right Peter, the stronger attraction to a magnet puts it out of either A or C category. I had not considered Kanthal or Alloy 875, but it makes perfect sense since it is a cheap alloy. Also the large amount of oxidation can now be explained. Sorry if I led you down the wrong path

Once again, the question that must be asked is why does pure iron not produce the effect while 74% does?

I will again run  the pure iron test at the highest current possible.

Testing the heater again removed from the frame at low current, 2-3 amps my waveform looks more like yours. At the highest current 11.5 amps right off the line, the large spikes appear at the zero crossings.

I counted the turns as best I could but it is difficult to see. Could be 19  or 20 turns per inch and this would affect the final resistance. I used 19 for my calculations, but 20 per inch would yield 5% more resistance.

Here are a couple of pictures:

I make that 18  8)
I added some more to my previous post. O0

It also mentions it's use in radiant heating elements.

off to be now, i will try to locate a UK supplier tomorrow, good work ION  O0

EDIT just seen your quote on high current, i did calculate pk current on my tests and i am pretty sure i was way up near 6 amps or so as well.

Luckily i can buy a complete 1200watt heating element, i could use a crock clip to adjust where i inject power for adjustment and they are cheap  O0
This wire is used in a furnace, What did SM say about furnaces  ;D

http://www.ebay.co.uk/itm/1200W-22-8-220V-Kanthal-A1-Heating-Element-Coil-Heater-Wire-/290755909737?pt=UK_Home_Garden_Washing_Machines_Dryers_Parts_Accessories_ET&hash=item43b2671869

PS i just ordered one, this is Kanthal A1

Here's the composition, the only worrying thing is that it says it's corrosion resistant.

QuoteThe material is magnetic up to approximately 600°C (Curie point).

http://www.kanthal.com/products/material-datasheets/strip/kanthal-a-1/

I see that there is a big progress in the indentification of the alloy!  O0

On my side I have carried out a new test. I have wound 20 turns of a thin enamelled copper wire around the resistive wire. This is equivallent to have the resistive wire as a coil core. I connected the copper coil to the primary of a transformer to raise the voltage and monitored the signal at the secondary.
I have still the pulses.
Then I decided to check if the permeability of the resistive wire was changing with the pulses. I "magnetically polarized" the resistive wire with a near permanent magnet. Therefore if the permeability was modulated, the magnetic flux from the magnet also, and this should reinforced the pulses picked up by the copper coil.
Here is what I observed: the pulses disappear with the field from the magnet. Even a weak field has a strong effect. If the magnetic axis of the field is colinear with the resistive wire, the effect is the strongest, the pulses disappear completely. If the field is transverse, the pulses are only weakened.

I had doubts about the "quality" of the mains here, because the peaks of the sine are a bit flattened. If so, the pulses could result from "the derivative of the angle" when the signal departs from the sine shape. This new experiment discards any possible artefact in the previous measurements and confirms that the effect can only come from the magnetic properties of the resistive wire.

Quote from: GibbsHelmholtz on 2012.08.19, 14:10:11
An example of this would be ION able to detect the spike while others struggling. 

He has magic hand?  His mind manifested the effect? Scientific explanation materialized yet?   

Since at the outset we had not positively identified the material in the heater, other researchers may have failed by using the similar test setup with the wrong material under test.

By more careful analysis, Peterae has come a long way in identifying the material as Kanthal A1. This material is not used in all infra red heaters.

I do not have a metallurgy lab and was not able to positively identify the material under test at the outset.

We are getting closer. Kanthal has a very high iron content, more than likely the cause of the spikes due to magnetic saturation.

Progress is being made.

No magic hand or mind effect, just inferior science on my part. mea culpa!

Inferior science? huh...
You are the one who noticed the effect for the first time.
"In the fields of observation chance favors only the prepared mind."
Louis Pasteur
;)

Quote from: exnihiloest on 2012.08.20, 10:12:58
Inferior science? huh...
You are the one who noticed the effect for the first time.
"In the fields of observation chance favors only the prepared mind."
Louis Pasteur
;)

Unfortunately I had the preconcieved notion that all infra red heaters were wound with a variation of Nichrome containing small amounts of iron, not enough to account for the "effect".

While I had heard of Kanthal A1, I had no good reason to believe it was being used in infra red heaters.

It was Peterae's discovery of the heater wire with high iron content that now dispels any chance that the effect is something special.

I formally apologize to anyone's time I may have wasted on this investigation.

On the flip side, I have learned a lot about many of the wire alloys and how they react under high currents, so it was not a waste of time for me.

I have also ordered a heater core made of Kanthal A1 to confirm my error.

This exercise also gave me the incentive to go through and inventory some of the many spools of wire I do have on my experimental shelf, Manganin, Karma, Alloy 11, Contantan, Chromel etc. Too bad I had no Kanthal A1 or this would have been put to bed many posts ago.

When I have some time, I would like to investigate some of these in a slightly different way, by winding the wires into a core, and overwinding with a fixed number of copper winds like a toroidal inductor in order to determine magnetic effects of the alloys.

I will continue to investigate the "effect", because like a child, I am in wonderment and who knows what can be learned or discovered.

For me it has been fruitful and educational.

Hey ION
No waste of time, there's never a waste of time with anything we enjoy experimenting with, as you say we learn all the time, even if this turns out to be a dead end, who knows that we wont learn something that is crucial to understand and use as information when we build one of our next devices, i have learned so much lately, transistor sine wave oscillator, class A amps, and now heating elements i am having a great time and i thankyou for the encouragement for this.
I always try to post all my work for a couple of reasons, so that people know and can learn from my mistakes, yes most of what i do ends up not being unfinished or not working properly but i look at each item as a stepping stone to the future.

Very few share or really join in on builds and i feel it a privilege to be able to work as long side you and some of the others that do help.

I Just hope i have the wire identified correctly now you have bought some wire also  O0

Quote from: ION on 2012.08.20, 13:04:59
Unfortunately I had the preconcieved notion that all infra red heaters were wound with a variation of Nichrome containing small amounts of iron, not enough to account for the "effect".

While I had heard of Kanthal A1, I had no good reason to believe it was being used in infra red heaters.

It was Peterae's discovery of the heater wire with high iron content that now dispels any chance that the effect is something special.

I formally apologize to anyone's time I may have wasted on this investigation.

On the flip side, I have learned a lot about many of the wire alloys and how they react under high currents, so it was not a waste of time for me.

I have also ordered a heater core made of Kanthal A1 to confirm my error.

This exercise also gave me the incentive to go through and inventory some of the many spools of wire I do have on my experimental shelf, Manganin, Karma, Alloy 11, Contantan, Chromel etc. Too bad I had no Kanthal A1 or this would have been put to bed many posts ago.

When I have some time, I would like to investigate some of these in a slightly different way, by winding the wires into a core, and overwinding with a fixed number of copper winds like a toroidal inductor in order to determine magnetic effects of the alloys.

I will continue to investigate the "effect", because like a child, I am in wonderment and who knows what can be learned or discovered.

For me it has been fruitful and educational.

Ion, you're doing a good job, basic science,  and "basic" in this context is meant to mean VALUABLE.   

but you are not done yet,  :)

I would like to know if the main voltage spikes, at the zero crossing,  are actually many fast Barkhausen spikes, which only appear to be one spike due to the time base setting of the oscilloscope.    So,   if you wanted to investigate this,   trigger off of the spiky wavform, with a ONE SHOT setting, and set the trigger level so only those spikes trip it,  and than set the time base to the nano second or micro second range and see if you can resolve the one spike into many shorter spikes.     If you can resolve it into many spikes, it is clear we are seeing Barkhausen noise, like I predicted.    And since I predicted it, it must be true!     LOL   ;D   :D

EM

We have a very interesting mix of metals here.

Aluminium forms a thick oxide coating around the exterior of the Kanthal wire. I think MOS

we also have 0.8% carbon this is enough carbon to make our iron into steel, causing grain boundary anomalies, at elevated temperatures the strength of a grain is stronger than the grain boundary, at lower temperatures the strength of a grain boundary is greater than the strength of a grain this is called "equi-cohesive temperature."

then we have manganese which is paramagnetic
A mixture of manganese and aluminium at low levels increases resistance against corrosion due to the formation of grains absorbing impurities which would lead to galvanic corrosion

And then we have Chromium, The strengthening effect of forming stable metal carbides at the grain boundaries and the strong increase in corrosion resistance made chromium an important alloying material for steel.

And again we now have the ability to form Silicon Carbide or as we know in the electronics business SIC

I personally would say we have so much going on with this wire anything could be happening  >:-)

I observe that the pulses appear as soon as the current flows. So the heat is not their cause. I'm convinced we would still have the effect if the wire was cooled, for example in water.

The idea of new possible effects of a "hot" coil, which was interesting and led to the observation of the pulses, becomes now misleading.

The effect looks rather like magnetic saturations due to the current through undefined or not well identified components of the metal, with sharp transitions inducing strong dB/dt. We can use the temperature to change the conditions but it is not indispensable.

What do you think?

QuoteI observe that the pulses appear as soon as the current flows. So the heat is not their cause

Maybe i misunderstand what you are saying, but we know the effect stops when hot and that the effect is always there when cold and therefore surely we already know the heat is not what causes the effect and quiet the opposite.

We know that once a certain temperature is reached the effect stops, and this temperature maybe quiet interesting to measure i personally would be expecting around 600 Degree's C, but if it's a lot lower then things will be really interesting because it maybe then something to do with a semiconducting layer.

There's something very interesting about zero crossing as well, i noticed a big red sign when i worked on the Romero motor that said investigate me(unfortunately i have only just realized this DOH)
Anyone remember that at zero crossing with the magnet on the rotor i was getting loads of spikes on my pick up coil, they were only just noticeable.I will try to dig the appropriate video out where did that noise come from i wonder.I will append the video here when i find it  O0


OK found it, i put this down to switching transients, but hang on, why would i be picking switching transients up on a rotor coil that is being driven by a magnet when the coil that drives is nothing to do with the coil i am scoping,maybe this is barkhausen noise

Agreed Peterae, from the level of the spikes shown in your video they indeed look like Barkhousen noise, assuming of course there were no other switching transients taking place in your Romero motor.

You also have brought up many good points regarding the metallurgical properties of the wire we have been testing.

When  you and I get the Kanthal A1 heaters in we will have some interesting tests to perform.

I might note that my Martel heater core was highly oxidized.

I noticed in my stock I have some copper clad steel wire. The copper has a thin oxide coating (dull brown).

This might be interesting to play with. Remember SM said he had good results with some "old" wire.

Also remember the patent of the guy (escapes me) who extolled the virtues of the oxide coating.

Was it Barbat?

Quote from: Grumpy on 2012.08.21, 14:31:49
Was it Barbat?

Yes it was G, thanks for that.

Right on ION, i have an infra red thermometer to point at it  O0

QuoteRemember SM said he had good results with some "old" wire.

wow forgot about that one, where did that quote come from, the reason i ask is that if we are even close to the first principle of the tpu's operation then there's a lot wrong with stuff we have been told, and in this case this statement was correct so it maybe worth looking up what else he said at the same time just in case there's more, as i suspect some stuff is BS but some is not.

QuoteI noticed in my stock I have some copper clad steel wire. The copper has a thin oxide coating (dull brown).

Indeed, years ago in the UK the whole telephone system relied on twisted wire connections, hard to believe really LOL, the insulation is first stripped off and then literally both wires were twisted together and a cap was crimped over the top to seal it, well every now and then you would start getting radio stations playing down the phone LOL, offcourse this is because of the poor old wire coating had oxidezed and had formed a semiconducting layer DOH.

Interestingly he did not state copper to be the old wire, and i remember some mention about iron wire, even the wire from an old heating element would count ;)

QuoteWhen  you and I get the Kanthal A1 heaters in we will have some interesting tests to perform.

I not counting my chickens yet LOL but yes if this is indeed the correct wire we can start having some fun indeed  O0

From SM:

QuoteBy the way, something strange is that, once I made a unit with some old corroded wire for the collector and it worked very well for it's size.

QuoteNow, electrons can travel only so fast along the surface of the wire because of magnetic flux. what if you disable the effects of the flux? Now the electrons float freely without anything holding them back. electrons at the sped of light are now a possibility

QuoteSpeed of the electrons across the surface of the wire is part of the key in my conversion process.

Some good clues when considered from the point of view of surface effects as in the Barbat patent.

Wow thanks ION

It all seems to fit but i guess we could probably say that for a lot of instances.

Quotedisable the effects of the flux

So we have AC current, and at our zero crossing point have we in effect disabled the effects of the flux?
One minuet we have positive flux then we have zero flux then we have negative flux, notice we have the pulse each time, even when we go from positive to negative flux and also negative to positive flux, i would say we get the pulse each time the flux is zero  ^-^

Now i am going into la la land  ;D

Imagine that all iron wire was to do this but in varying amounts and it just so happens that you stumbled on a wire type that even SM did not know of, but SM worked out that by injecting a harmonic of a certain frequency or even maybe 2 different harmonics that this made the pulse appear in other types of wire how amazing would that be.

I cannot wait to find out the effect of the speed at which we cross the zero point

I cannot but help think there's a way to make that thermionic rectifier valve exhibit the same zero crossing kick you have, maybe time will tell me how to do this.

Quote from: Peterae on 2012.08.21, 09:58:37
Maybe i misunderstand what you are saying, but we know the effect stops when hot and that the effect is always there when cold and therefore surely we already know the heat is not what causes the effect and quiet the opposite.
...

You have well understood, it is what I said. Regarding the subject of the thread, I just wanted clarify what we handle: magnetic effects in resistive wires, not specially in hot coils.

In my heater, the wire was brilliant. There was no apparent oxide before I run tests without the fan, and so the too hot temperature oxidized the wire which became gray. But I noticed the pulse effect before the wire was oxidized and no difference after. I don't believe in a surface effect.

Ah OK ex yes, i am sorry for the thread title we sort of veered offcourse when ION stumbled on the anomalous pulse using his heating wire.

Maybe i should re title the thread, if we can verify the correct wire type then maybe call the thread Kanthal anomalous pulse during AC current flow.  O0
OR
Even split the thread into 2

QuoteBut I noticed the pulse effect before the wire was oxidized and no difference after.

That's good data and another thing ruled out thanks ex.
can i ask does your wire exhibit a strong magnetic attraction.

Quote from: Peterae on 2012.08.22, 07:10:12
...
can i ask does your wire exhibit a strong magnetic attraction.

Yes it does. The wire is wound as star branches around an insulated disk. A branch is flexible, it bends strongly towards a neodymium magnet and sticks to it. I think the effect is less strong but not far from that of an iron wire.

Cheers ex  O0
Still waiting on the slow boat from Hong Kong  >:(

I have put a diode  in series with the heater (YG912S6 fast, low loss, 600v 10a). The goal was to see if the phenomenon still occurs when there is no current reverse. The result is strange. Here is the signal from my coil probe:

(http://www.overunityresearch.com/index.php?action=dlattach;topic=1505.0;attach=8407)

I can't explain this shape. Perhaps an effect of sharp transition in the diode? It doesn't depend on the voltage that I could change with a variac.
There is probably a high-pass filter effect because my coil impedance is not high enough.
An idea?

Exn,

Can you superimpose the current and pick up coil waveform like ION?  I did it by hand but not sure if it resembled. 

Ex

Yes it looks like the diode switch-off causes the ringing oscillatory pulse, similar to a flyback effect.

The ringing frequency will be a function of distributed capacitance of the wiring, diode capacitance, stored energy and impedance of the wire inductance.

You can expand the scope timebase to examine the ringing frequency.

This shows stored energy in the system from the content of iron in the wire which increases the value of inductance.

Also the slower changing descending waveform does seem to coincide with the positive half of the pulsating DC.

The current waveform would be helpful but will be very slightly lagged from the voltage input. Perhaps just use the input voltage as a phase reference.

It is a bit difficult to explain the minor wiggles on the slow descending waveform and the flat spot around the middle.

The flat spot could be the diode forward drop when current changes direction at the peak of the half cycle.

Quote from: GibbsHelmholtz on 2012.08.24, 12:23:41
...
Can you superimpose the current and pick up coil waveform like ION?  I did it by hand but not sure if it resembled. 

@gh @ion
I can't monitor the current and the coil probe signal in the same time, because I need the two probes of the oscilloscope to monitor the current in differential mode.
Nevertheless your surimposition is correct, replacing the trace of current by voltage that I have monitored. The voltage should be in phase with the current, because the heating wire is wound in order to minimize the inductance.
Maybe there is only a slight delay between the zero crossing point of the sine, and the following transient of the signal. Not clear on my trace.

Quote from: ION on 2012.08.24, 13:26:43
Ex

Yes it looks like the diode switch-off causes the ringing oscillatory pulse, similar to a flyback effect.

The ringing frequency will be a function of distributed capacitance of the wiring, diode capacitance, stored energy and impedance of the wire inductance.

You can expand the scope timebase to examine the ringing frequency.

This shows stored energy in the system from the content of iron in the wire which increases the value of inductance.

Also the slower changing descending waveform does seem to coincide with the positive half of the pulsating DC.

The current waveform would be helpful but will be very slightly lagged from the voltage input. Perhaps just use the input voltage as a phase reference.

It is a bit difficult to explain the minor wiggles on the slow descending waveform and the flat spot around the middle.

The flat spot could be the diode forward drop when current changes direction at the peak of the half cycle.

I agree. And my measurement circuit can also add ringing: not only this from the inductance/capacitance of the coil probe, but also from a transformer that I use to insulate the measurement side from the mains.
But why is there a so high rising transient just after the zero crossing, while the input is still not far from zero? It is not a possible flyback effect because at this instant, the diode becomes conducting.

exn said:

QuoteBut why is there a so high rising transient just after the zero crossing, while the input is still not far from zero? It is not a possible flyback effect because at this instant, the diode becomes conducting.

I would think the diode is conducting in the forward direction during the positive half cycle and turns off or blocks shortly after the zero cross unless there is considerable inductance. In the latter case, current, not voltage must go to zero for the diode to turn off.

If possible sketch your circuit , then we can see if your measuring transformer is also supplying some inductance.

Quote from: exnihiloest on 2012.08.24, 17:09:04
...
But why is there a so high rising transient just after the zero crossing, while the input is still not far from zero? It is not a possible flyback effect because at this instant, the diode becomes conducting.

I also like to see it as a domain avalanche effect.  The domain is like a mass spring system.  When it is in neutral position(no current), it moves fast.  This creates high dB/dt.  If there is no hysteresis losses, it would move back and forth and ringing forever like a fly wheel after input is removed.   

My Kantanal Heating element arrived today and i can confirm it is indeed the correct alloy to show the pulse.

I am using a relay coil with iron core as a pickup, towards the end i pull it away from the heating element because things are starting to heat up  O0

Here's another test, here i use the yellow trace to monitor the voltage across the kanthanal element.

Also there's something very strange happening, the pulse shape depends on either the position of where along the heating elements length i place my sniffer coil or the pulse shape changes between switch on and off.

If you look at this video i switch on and off then reposition the snifffer coil because the waveform did not look right, i then having moved the sniffer coil further along the element then switched on again, and then i have a totally different shaped waveform, i have now seen this a few times.

It may also just be how close and the orientation of the sniffer coil is relative to the kanthanal element

Good work Peterae, looks like Kanthal is the stuff. I should be getting a sample in the mail soon.

I get the largest spikey pulse when at the very end of the coil, either perpendicular or parallel.

For the record, when I test my mag probe against an external coil of copper wire excited at 60 Hz, I get a nice clean sine wave.

Right yes ION i need to work on the pick up coil, what i will say is i am only using a short piece of element so i am relatively low voltage and high current, and the difference also with this wire is i can hear audible noise from the element which i could not with nichrome.

I did try my hall, although the sine picked up was buried in noise and although it was sine shaped there was no sign of spikes that matched but there really was a lot of noise so i would say it was inconclusive.

I also believe i could maintain a set temperature by turning my variac down and being able to maintain the pulse indefinitely, but i do need to confirm this, i have the wedding tomorrow so am out of experimentation time for a couple of days.

During my tests the amount of wire i used was 5.1Ohm and my yellow scope Chan was 20V/Div so i am about 70V/pk-pk giving me a peak current of 13.7 Amps

I am feeding power via the green and yellow crock

God my bench needs cleaning up  C.C

Anyway here's another test, i straightened a length of Kanthanal as best as i could and powered it up enough so i had the pulse but not enough for it to get too hot that the pulse disappeared, i then used my sniffer coil to run down it's length to trace out the wave patterns that were picked up, the spiked pulse appear to be nodal if that makes any sense, how do you work out the nodal spacing in a piece of wire for 50Hz?

Is anyone able to explain why the pulse changes down the length of wire.

Peter.

I would like to suggest you do the same test again, but this time firmly mount the wire on a piece of wood or something. This way it will be stretched out, not curved, and not moving. Taken a step further, I would then attach your sniffer coil to a rail mounted above the wire so that the transition along the coil can be smooth and consistently the same distance away.

I think then you would be in a better position to determine if there are nodes or maybe even regular nodes along the wire. You'll also get a better idea how the wave form changes as you go along, and perhaps if the wave form repeats itself.

Start there, then we have something to analyse.

Thanks for the advice Darren, yes i have just been trying to also improve this test, i think i may need to track down some uncoiled wire though, because this heating element is virtually impossible to straighten, but that's a great idea to try stretching it from 2 pillars.
I can then use a tube of something that does not melt and thread the wire through it and glue my coil into that, i could also wind axially a coil over the wire as well and run that down the length to see what gets picked up.

Peter,

I was thinking more along the lines of stapling the wire to a length of wood, perhaps every 4 inches or so, but whatever works for you is ok with me.

QuoteIs anyone able to explain why the pulse changes down the length of wire.

I think POYNT implies it is because the wire is not completely straightened and has undulations. I tend to agree, pending more tests.

Good work Peterae.

Indeed it does not make sense to me if there is a variation down it's length either, but will try to rule it out just to make sure.

I don't have staples but i think i have some horseshoe tacks somewhere but like staples they will be iron and surely they would affect the field as well.

If you enjoy exotic test fixtures, a 2 foot length of threaded rod driven from a toy gearmotor will smoothly and uniformly move your pickup coil along the length of a 2 foot wire that is tightly stretched between two posts.

A small block of insulating material with a hole drilled into it is glued to the pole of the pickup coil. The wire is threaded through the hole. This maintains precise distance from the wire as the threaded rod moves the pickup coil.

There are many simple variations on this.

Enjoy the wedding!  O0 Fixtures can wait.

I could go that way ION  O0 but i do not have a warehouse of bit's like yourself LOL so would need to buy bits.

So let's remove one variable MOVEMENT

So the plan is to make stationary tests that as i progress down the wire, with the wire stretched tight and the sensor fixed distance from wire by a hole and piece of wooden block and see how that goes.

Looks like i am out of this until i can find a higher amperage variac, i spent a good half hour last night trying to sniff the nodes and the patterns but what i did not realize was my variac inside my box was starting to cook along with the isolation transformer, when i realized i let it all cool down but today it looks like i fried both.

What i will say though is there are  some strange goings on, i had one cycle sine peak with a spike that had a sharp left shoulder then i would move about 4cm down the wire and find the same sine peak and spike with a sharp right shoulder and then when i moved another4cm i had the same sine with no spike, i then moved another 4cm and the pattern appeared to cycle over again, it took me a while to trace out the pattern, if i moved my sniffer coil up passed the heater wire i would have my sine one way, if i moved it down so i was the other side, my sine flipped which is what i would have expected, in between this flip was a null spot where just the spike could be observed, and it was this process i gradually worked down the wire length.

Any magnets nearby would greatly enhance the spike or pulse and they would not need to be really close to affect the waveform.

Sorry no pictures and no data, this was going to follow today, i can give you a picture of where i marked the wood that the spikes appeared
See the up & down arrows i drew on the wood, up arrows had sharp left transition and down arrows had sharp right shoulders to the spike transition, you can also see the gap in between where there was a sine cycle but it had no spike.

Maybe if i place 2 x 100Watt bulbs in series with the heater i can drop enough mains to run the tests without a variac or isolation transformer.

Also i had the wire bolted at each end and stretched it as much as i could, it made a nice note like a guitar string  :) but i could not get it totally straight for the tests as can be seen.

EDIT also you will notice the gap in the middle is not equal to 2 pulse widths it's less and i cannot explain that either.

Quote from: Peterae on 2012.09.06, 17:34:57
...
What i will say though is there are  some strange goings on, i had one cycle sine peak with a spike that had a sharp left shoulder then i would move about 4cm down the wire and find the same sine peak and spike with a sharp right shoulder and then when i moved another4cm i had the same sine with no spike, i then moved another 4cm and the pattern appeared to cycle over again
...

A possible explanation is that each sections of wire is source of magnetic fields that constructively add or cancel at the probe position, depending on its location.

Quote
Any magnets nearby would greatly enhance the spike or pulse and they would not need to be really close to affect the waveform.

In my experiment (see Reply #121), a magnet made the spikes to disappear, and I also observed that it had not to be very near the wire. I suppose that the effect depends on the magnetic properties of the wire.
In any case, interesting that you have confirmed the effect of a weak static field, Peterae.

one thing worth mentioning though is my pickup sniffer coil has an iron core so that could also be affected by a magnet being in close proximity, just as a quick test i sniffed my crocodile clip lead which feeds power and i do get the same sine waveform but without the spikes.

Peterae:

What is the resistance per foot of your wire? Seems to me it should be pretty low, in which case you can get away with a much higher turns ratio on your stepdown transformer e.g. if it is now 24 volts you can use a 6 or 12 volt secondary to unburden your Variac and transformer.

Variacs can be repaired if not too far gone. They need to be taken apart and varnished on the inner part of the core. I have brought quite a few back to life.

By the way, when unrolling a coil best to put a dowel through it and unwrap the winds perpendicular to the dowel, this way the wire is free of curls. I put one end of the wire in a vise and pull the dowel away from the vise.

Neat results with the sniffer coil, I'll have to check that out when my heater wire comes in.

The variac's wire is embedded in a thick glass epoxy, apart from where the wiper touches at the bottom.

The total cold resistance of my test wire right now is about 5 Ohms

Good thinking ION i have a microwave transformer i rewound to use as a valve HT 320-0V but i also put on a very thick 6V secondary and i could add turns if need be, looks like i could be back in business shortly.  O0

Quote from: Peterae on 2012.09.07, 07:19:43
The variac's wire is embedded in a thick glass epoxy, apart from where the wiper touches at the bottom.

The total cold resistance of my test wire right now is about 5 Ohms

Good thinking ION i have a microwave transformer i rewound to use as a valve HT 320-0V but i also put on a very thick 6V secondary and i could add turns if need be, looks like i could be back in business shortly.  O0

When I need high current low voltage, I use an old variac core and wind whatever turns through the center are necessary to get a good match to my load.

The MOT will do just fine if you have the room on the core.

Very interesting Peter.    The repeating paterns is indicative of standing waves.    Magnetostriction plays a part, which is the coupling mechanism between acoustic strain energy propagating in the wire and the magnetic field. 

Here's something to try:   Clamp a metal wrench or pliers to the wire at various distances and see how that affects the standing wave patern.  You might find there are nulls around the place you clamped. 

EM