I have started this thread in the hope that we can get some experiments done to prove an important point about magnetic energy. I have long held the view that in magnetized ferromagnetic material, be it soft or hard, there is an enormous quantity of magnetic energy stored in the inter-atomic space. This is not recognized by existing theory but IMO that theory is significantly flawed. It uses a charactristic called magnetization M that has dimensions of dipole-moment per cubic meter, it being the volume density of the aligned electron dipoles that are responsible for the effect. That should really be a number density, something like 1028 dipoles per cubic meter, but the manner in which the science has developed M is treated as a continuous attribute filling all space within the ferromagnet. That is as ridiculous as imagining that the mass of electrons and nuclei doesn't reside within those particles, but is spread around all space within matter. I started writing papers on this 14 years ago and the fourth version I published in 2005. It is appended here for your perusal.
When you charge an inductor that has a ferromagnetic core to store energy of value 1/2Li2 the actual energy stored in that inner space is Chi times that value, where Chi is the magnetic susceptibility, very close to the relative permeability muR (because muR=1+Chi). The additional hidden energy comes from the electron dipoles, not from the current source. My reason for bringing this up now is the possibility of releasing that large quantity of hidden energy simply by taking the core material above its Curie point. With relative permeabilities that can reach as high as 106 (Metglas) I think that sort of huge energy gain is worth striving for, even if the requirement for thermal driving means that the repetition rate will be very low. My vision is something that is driven alternately with thermal heating and cooling pulses thus alternating about the Curie temperature and releasing large value energy pulses at a low repetition rate. However the initial experiment can be a simple one-shot, just heat a charged inductor to above the Curie point and see what happens.
My reasoning on this assumes that a shorted coil around the core will, at the Curie switching point, hold the magnetic field at its pre-value even though the dipole alignments are destroyed. Because the inductance has dropped significantly to that of an air-cored one, the current through the shorted coil has to increase in value to take the place of the now defunct electron dipoles. You now have an air-cored coil that can discharge into a resistive load, and that energy is the original hidden energy within the ferromagnetic material. Is anyone willing to take this on? Of course the inductor must be wound in toroidal form on a ring core, and that core can't be enclosed within a plastic case (which rules out the Metglas mega-mu core), and attention must be paid to the magnet wire insulation that has to survive the heating.
Smudge
A creative and interesting idea, thanks for sharing it I will give it a try when I get out from under my domestic workload.
sounds good Smudge O0
OK to clarify and make a practical experiment, wind a primary coil with high temperature insulation, drive this with a AC current, also have a secondary coil that is shorted by a 0.1 Ohm resistor to monitor current, then heat core with blow torch, when really hot the current should increase in the secondary.
Although this does not sound correct to me because when hot it will be an air core and there will be no secondary current.
So maybe we drive the primary coil with DC 1 Amp, zero current flows through the shorted turn, we then heat the core above the curie point and all the stored core energy dissipates in the shorted turn, we can calculate the total stored energy by calculating the dissipated energy in our 0.1 resistor.
How evenly heated does the core need to be?
Quote from: Peterae on 2017.08.03, 18:09:00
sounds good Smudge O0
So maybe we drive the primary coil with DC 1 Amp, zero current flows through the shorted turn
The shorted turn is not shorted while the DC current builds up. It is shorted when the DC current is stable.
Quotewe then heat the core above the curie point and all the stored core energy dissipates in the shorted turn, we can calculate the total stored energy by calculating the dissipated energy in our 0.1 resistor
That's about it. The problem is how does the magnetization behave as you go through the Curie point because a slow change may not give you a measurable pulse. If we had a superconducting shorted coil then it would take on the current needed to keep the flux constant even over a long time frame, and then you could remove the short from across a load resistor to discharge the now air-cored coil. I think the experiment is worth doing with say your 0.1 ohm resistor just to see if there is a sudden pulse. I can't find any evidence that this has been done before. There is plenty of evidence for thermo-magnetic motors that really work but they are different to this. Interestingly this web site says that they can be OU. https://contest.techbriefs.com/2014/entries/sustainable-technologies/4921 (https://contest.techbriefs.com/2014/entries/sustainable-technologies/4921)
QuoteHow evenly heated does the core need to be?
I would guess that the objective would be to take the whole core through the Curie temperature at the same time.
Smudge
QuoteThe shorted turn is not shorted while the DC current builds up. It is shorted when the DC current is stable.
Well that can be done with a switch, just switch the short on the turn after applying DC current to the primary. O0
QuoteI would guess that the objective would be to take the whole core through the Curie temperature at the same time.
So we need to make a miniature oven really with a temperature probe to monitor the temperature, we could drive the primary coil with AC run the temperature up while monitoring the secondary and when we loose our secondary output would give us an idea of the curie temperature.
I have some heating element wire so should be able to make a heater for a little brick furnace.
I could wrap the copper wire with kiln paper for insulation.
Any preference on core material, what is the best we can get for a reasonable price? or shall i just use one i have laying around.
@Peterae,
Use the one lying around that has the the lowest Curie temperature.
Smudge
Looks like my thermocouples i have are only capable of 250 Deg C maximum, hopefully that should be ok for most ferrites.
I have constructed a miniature kiln using 2 small baking dishes packed with high temperature fire cement that has up to 1250 Deg C resistance starts glassing at 800 Deg C, it will take a while to thoroughly dry.
I can then wind some turns on the core and sandwich between the top and bottom halves of the kiln, i wound the heater element into the bottom half of the kiln.
The space for the core and windings is 35mm diameter.
How well it will work i don't yet know, but hopefully next weekend i will try heating it up O0
The primary current carrying coil should i wind that all the way around the core or is just one side OK.?
1 turn for the secondary or would multiple turns be better.?
Thanks
Peter
Quote from: Peterae on 2017.08.06, 17:39:11
The primary current carrying coil should i wind that all the way around the core or is just one side OK.?
All the way round.
Quote
1 turn for the secondary or would multiple turns be better.?
One turn is not the best form of coupling so I would go for more turns and again wound all the way round. Use the largest dia wire you can fit so that any energy recovered is not lost in the coil.
Smudge
Biggest wire i will be able to use is 1mm because i have high temp sleeving for only 1mm.
Main problem with 2 wires wound all the way around is they may shield parts of the core from the heat with the fiberglass high temp sleeving, i will try to gap the winding as best as i can, but probably will not be able to get that many turns on a small core.
I will give it a go anyway O0
OK sorry for the delay.
Some specs
Core Type:4C65, OD=23mm, ID=18mm
Turns: 4.5 Bifilar copper enamelled wire in high temp sleeving, Wire is 18SWG.
Inductance of coil 1, 2.13uH
Inductance of coil 2, 2.16uH, Q @ 10KHz = 32.4, Resistance = 0.0042 Ohms, Impedance @ 10KHz 0.1361 Ohms.
Now i soldered a 0.1 Ohms Metal Film resistor across 1 coil and scoped this.
I ran the primary at 10KHz from a signal generator and scoped the second coil while gradually increasing the temperature inside the
kiln, tests started at 42.6 Deg C and over a period of about 15 minuets the temperature rose to 300 Deg C, i was looking for a change in amplitude of the scoped coil.
The average pk-pk reading started at 6.25mv with some noise, this value gradually rose to 6.41mv at 296 Deg C
I stopped at 300 Deg C and did not see a fall off of amplitude.
The K Type probe was positioned in the centre of the core.
EDIT
Just checked the data sheet for this core and it has the temperature profile, looks like the Curie temperature of this core is about 340 Deg C.
http://www.ferroxcube.com/FerroxcubeCorporateReception/datasheet/4c65.pdf
I have run out of time for now but will try again next weekend O0
It looks like the 4C65 was a bad choice of core, it was the only one that had the type printed on it for identification.
I have just ordered a 3E5 Core, this looks like a much sharper permeability fall off at 125 Deg C and falls from a u of 20000 to 0 in about 5 Deg C span.
OK i transferred my windings to the new core.
Inductance measures 269uH with 4.5 Turns primary and secondary.
I connected my 0.1Ohm resistor and drove the primary with a 10KHz signal, scoping the 0.1R i could verify this time that at 134Deg C the amplitude starting declining and vanished totally at 155 Deg C
Now i have 2 power supplies 1 is 5amp and my 2nd psu is 1 amp, i had to use the 5 amp for the supply to the heater which at 31.5 volts was drawing 5 amps, i used the 1 amp for the rising temperature test across the primary, and when the heater gets turned off for the test when the temperature falls i connect the 5amp to the primary to help see if it makes any difference.
I am sorry to say i could not see any difference in the current across the 0.1Ohm resistance at all having run upto 175 deg C and then cool to 100 Deg C many times.
The below scope shots are the same but at different timebase settings.
I'm confused: are you saying that you did see a temperature effect with the 10 kHz FG oscillations and didn't see it with DC?
Also... sanity check time.
First, for the second shot it doesn't seem that "Pulse Width" trigger mode is appropriate for the signal you are displaying, especially since your screen width is only 12 x 5 ns = 60 ns, but your pulse trigger is set to trigger on a 1 us pulse width. Are there any pulses in the scoped signal that actually fit that criterion? Can't tell from the scopeshot.
Second, the Hardware Frequency Counter is reporting 263 kHz, but the noisy waveform appears to show roughly 1 full cycle during 2 horizontal divisions, or 10 ns. My calculator says this is a frequency of 100 MHz. The HWFC operates on the entire memory buffer IIRC, so perhaps you have lower frequency pulses and you are simply zoomed in to a portion. Again, can't really tell from the scopeshot.
Third, at 1 mV/div vertically, and with 1x probe, your Rigol scope is very sensitive to its own internal noise and environmental noise coming from CFL and LED light fixtures, power supplies, and etc.
So I'm not sure your scopeshots are actually showing anything useful other than noise. Especially since the first shot is using the normal slope trigger and the second shot is using the Pulse Width trigger.
Can you repeat the same test with the scope set to, say, 1 or 5 microseconds/div and using the Pulse Width trigger at 1 us width as before?
QuoteI'm confused: are you saying that you did see a temperature effect with the 10 kHz FG oscillations and didn't see it with DC?
Yes that is correct, driven using a 10khz signal generator on the primary i could see the AC signal across the 0.1 disappear as the temperature hit that band stated above.
Regarding the frequency reading, there is not enough signal for it to lock on to, it was a garbage reading, yes i missed the trigger setting but have tried quiet a few timebase settings to see if any pulses could be seen up to a timebase setting of 500ms, i also tried different trigger settings to see if i was missing anything, no change was seen at all during the temperature rise with 1 amp and the temperature fall with 5 amps.
I am guessing that the temperature rise and fall is too slow, maybe if the curie temperature was instantaneously reached then maybe there would be a sharp current increase in the resistor, yes i zoomed down to 1mv/ div to see the noise because there were no signs of any pulse higher than that level.
https://en.wikipedia.org/wiki/Barkhausen_effect
The only strange thing i would say is that in the first scope shot, it can be been seen that the noise is DC offset to the negative side of the scopes zero line, maybe i need to check the scope calibration.
It may also be worth switching to a 1 ohm resistor instead of a 0.1R resistor to increase the signal level above the noise floor of the scope.
Only just got back from several days away for wife to attend grandson's passing out parade, not military but police.
Peter, can you please tell me what your power supplies are. Are they typical bench PSU's which can be voltage and current controlled? If they are, then when used as a current source I think they are still low impedance voltage sources, it is just that their voltage controller uses output current as the reference for voltage control. Is it possible to drive current through a high resistance into the primary and look for a voltage change as the temperature goes through the Curie point? However I suspect that you are right about the temperature rise and fall being too slow to yield anything meaningful. Are you able to quote a thermal time constant figure, i.e. an idea of maximum rate of change of temperature for the core in your experiment? If this scheme is to go anywhere it could probably require an array of micro-miniature cores each with appropriate heating/cooling channels to get the thermal time constants fast enough for it to work. Anyway thanks for doing this work.
Smudge
Hi Smudge
The power supply can be constant voltage or constant current set, the 1amp supply is actually adjustable between 0-100V so as it is feeding a short the voltage is very small before it maxs out on the current.
I have 2 videos i made today, one showing the sine wave going through the curie point and the other with a 1 Amp constant current, trouble is i was unable to log into my youtube account to upload them, my mobile number changed as well, been ages since uploading anything, so i am currently in account recovery which can take days, i will upload as soon as is possible and post here.
As the core goes through the curie point the sine wave goes triangular.
The thermal time constant can be then calculated from the timestamp on the video.
First video of Sinewave driven core as it approaches and goes through the curie point.
https://youtu.be/wHDRECEzCgw
2nd video link but still has over 1 hour to upload yet
https://youtu.be/KdDS05MhB5w
PS i did try a few times to of different trigger settings after making the video just in case something was missed.
I will try your resistor idea when u get a chance
That's interesting, it goes from ferromagnetic to non-ferromagnetic over just a couple of seconds and over just a couple of degrees. I think there is the chance to do something more with that relatively fast change.
Smudge
Amazing Peter, thanks for showing.
Itsu
Thanks Itsu
Yes Smudge it was faster than i thought it would be, if more heat could be applied then the speed could probably be increased.
I maybe able to make a ring of fire clay for the core and place another heater in the other kiln half and sandwich the core top and bottom with heat.
PS i have added the 2nd video link but it says still over 1 hour to finish uploading, there is nothing to see in the 2nd video :( in that nothing happens.
The temperature reading in the video is probably inaccurate because the probe bead is buried under insulation of the wire.
Actually another way round this heating problem is to have as many turns of Nichrome wire wound round the core and use that as the heater with maybe 5 turns of copper enamelled wire wound round it as well.
The Nichrome wire would also be the DC bias winding and would need to be probably 10's of amps with such a short length of nichrome wire, this way we get even heating and a large DC bias current.
Using your inductance measurement of 269uH and the fact that this goes down to almost zero over 2 seconds we get a rate of change of inductance dL/dt=1.35x10-4 Henries per second. Then since V=-i*dL/dt with 1 amp you should see a 2 second voltage pulse of 135uV. Worth a try using your 100V supply fed through a 100 ohm resistor. Might need your scope set to DC though as its AC coupling might lose that 2 second pulse.
Smudge
As a matter of interest, if you had superconducting wire the only energy input to your 270uH inductor at 1 amp would be its 1/2Li2 of 135 micro Joules. If your coil held the flux constant while the inductance reduced, the current would increase by a factor of 2000 (the core mu). The 1/2Li2 energy now stored in the lower value (air cored) inductor would be 0.27 Joules. That's an apparent COP of 2000, but what is not known is what will be the energy effect on the heat input/output/storage.
Smudge
QuoteAs a matter of interest, if you had superconducting wire the only energy input to your 270uH inductor at 1 amp would be its 1/2Li2 of 135 micro Joules. If your coil held the flux constant while the inductance reduced, the current would increase by a factor of 2000 (the core mu). The 1/2Li2 energy now stored in the lower value (air cored) inductor would be 0.27 Joules. That's an apparent COP of 2000, but what is not known is what will be the energy effect on the heat input/output/storage.
Interesting thanks Smudge it's nice to see some calculations to go with the experiment.
Now i know the temperature is nice and low for the curie point i would say i dont need to worry about the heat proof insulation and can use enamelled on it's own, now what we need is a trade off between higher numbers of turns and wire thickness.
The 135uV presumably is calculated for the 1.1Ohm resistor, if so then that would mean the current is only 122uA, this means we could increase the resistance value to increase resolution and decrease the wire thickness a lot to get more turns?
Scope was set to DC anyway.
Quote from: Peterae on 2017.09.06, 17:01:35
The 135uV presumably is calculated for the 1.1Ohm resistor, if so then that would mean the current is only 122uA, this means we could increase the resistance value to increase resolution and decrease the wire thickness a lot to get more turns?
No the voltage just came from V=-i*dL/dt. It assumes that i comes from a current source that is high impedance so I suggested using your 100V supply (set to 100V) and a 100 Ohm series resistor. (Bench supplies set to a current limit are not IMO true high impedance current sources, they are simply low impedance voltage sources that control the voltage against the chosen current reference set by the current limit control). You would look for the 135uV across the primary winding carrying that 1 amp current, or across the secondary that is open circuit. The idea is to get some definite data to start the ball rolling. IMO we won't get anything near the effect of a superconducting shorted turn and we have to move ahead in stages, the first stage is actually seeing something to work with.
Smudge
OK i can give it a go but the winding with the 1.1 ohm resistance is showing mv's of noise so i am not sure how i am going to be able to see uV of pulse, unless the primary does not have this noise, i suspect the noise is probably coming from the scope front end.
It might be coming from your PSU set to current limit at 1 amp, does the current limiter create noise? Maybe using 100V with no current limit and dropped via a 100 Ohm resistor and the noise will be less.
Smudge
Just to clarify for this new test i can ditch the 1.1Ohm load on the secondary and leave the secondary open circuit.
I will certainly give it a go anyway.
Would i be correct in saying that when the core is below the curie point the magnetic field is contained within the core material, but above the core curie temperature the field expands to the surrounding space as well.
If so would the expansion field speed be slow as the temperature increases through the curie point or would there be an instant where the expansion field speed could be large, i am wondering what the expansion field may look like and whether it would be possible to induce a current in a piece of wire aligned such to detect the expansion, if we could oscillate the temperature between the below and above the curie point using only small amounts of energy we maybe able to collect the resulting field expansion with many pieces of wire.
Quote from: Peterae on 2017.09.07, 19:03:26
Just to clarify for this new test i can ditch the 1.1Ohm load on the secondary and leave the secondary open circuit.
I will certainly give it a go anyway.
Would i be correct in saying that when the core is below the curie point the magnetic field is contained within the core material, but above the core curie temperature the field expands to the surrounding space as well.
If so would the expansion field speed be slow as the temperature increases through the curie point or would there be an instant where the expansion field speed could be large, i am wondering what the expansion field may look like and whether it would be possible to induce a current in a piece of wire aligned such to detect the expansion, if we could oscillate the temperature between the below and above the curie point using only small amounts of energy we maybe able to collect the resulting field expansion with many pieces of wire.
I was thinking about this some time back Peter.
Rather than trying to raise and lower core temperatures,would it not be easier and more efficient to move the core in and out of a secondary coil,where the primary coil remains fixed to the core.
This would create an AC in the secondary,while the primary is supplied with a DC.
The inductance value of both coils would change throughout each cycle.
Just a thought.
Brad
Quote from: TinMan on 2017.09.08, 04:40:02
I was thinking about this some time back Peter.
Rather than trying to raise and lower core temperatures,would it not be easier and more efficient to move the core in and out of a secondary coil,where the primary coil remains fixed to the core.
This would create an AC in the secondary,while the primary is supplied with a DC.
The inductance value of both coils would change throughout each cycle.
Just a thought.
Brad
Sorry, I'm not following discussion but why should you move core ? You can eliminate core by saturation. O0
Quote from: Peterae on 2017.09.07, 19:03:26
Just to clarify for this new test i can ditch the 1.1Ohm load on the secondary and leave the secondary open circuit.
I will certainly give it a go anyway.
Would i be correct in saying that when the core is below the curie point the magnetic field is contained within the core material, but above the core curie temperature the field expands to the surrounding space as well.
If so would the expansion field speed be slow as the temperature increases through the curie point or would there be an instant where the expansion field speed could be large, i am wondering what the expansion field may look like and whether it would be possible to induce a current in a piece of wire aligned such to detect the expansion, if we could oscillate the temperature between the below and above the curie point using only small amounts of energy we maybe able to collect the resulting field expansion with many pieces of wire.
One way of looking at the induced voltage in the coil as the magnetic field collapses is to imagine that the field expands just as you describe. You can then account for the induced voltage by field lines passing over the conductors by the laws of motional induction
E=
vx
B. Then the imaginary velocity
v is determined by the rate-of-change of the field in the core. I say imaginary because this an imaginary concept, there is no evidence that such a moving magnetic field exists outside the core, and I am sure people will have looked for it, we do have magnetic field detectors. There can be a magnetic field associated with the changing electric field outside the core, but that is not the same thing. Imagine a sawtooth flux waveform in the core, the induced voltage is then a square wave. During the flat portions of the square wave the magnetic field in the core is changing but the voltage hence also the E field around the core that is driving that voltage is constant. A magnetic field detector will not see any external magnetic field, the expanding concept is therefore flawed.
Smudge
Quote from: TinMan on 2017.09.08, 04:40:02
I was thinking about this some time back Peter.
Rather than trying to raise and lower core temperatures,would it not be easier and more efficient to move the core in and out of a secondary coil,where the primary coil remains fixed to the core.
This would create an AC in the secondary,while the primary is supplied with a DC.
The inductance value of both coils would change throughout each cycle.
Just a thought.
Brad
Then you have a mechanically driven generator and that will not be OU. It's been done, but not with the toroids such as Peter is using. A bit diffucult pulling a ring core out of a winding O0
Smudge
Quote from: forest on 2017.09.08, 06:26:20
Sorry, I'm not following discussion but why should you move core ? You can eliminate core by saturation. O0
Yes, and if your saturating is done with a crossed field then the saturating drive does not directly couple to the coil so as to induce voltage. This uses the principle that if the material is saturated along the x axis it is also saturated along the y axis. I have done work on this in the past without getting any clear results. If you examine this theoretically you find that isotropic material will not yield OU. The case for grain orientated material is IMO unproved and could be the basis for further experimentation. I seem to remember Steorn exhibiting something using rmagnets on a rotor passing over toroidal coils and doing just that, and they claimed OU. Maybe that was the precursor to their solid state Orbo.
Smudge
Tried the 100 Ohm resistor, very hard i only had 10Watts and needed 100Watt so had to really limit time it was on, firstly i have 4mv pk-pk of high frequency noise with the scope probe connected and 1mv pk-pk without the probe even connected.
Plus i get random bursts of noise 40mv pk-pk.
We need to get the effect over the 4mv noise otherwise i don't think we will ever see it.
Quote from: Smudge on 2017.09.08, 10:53:39
Yes, and if your saturating is done with a crossed field then the saturating drive does not directly couple to the coil so as to induce voltage. This uses the principle that if the material is saturated along the x axis it is also saturated along the y axis. I have done work on this in the past without getting any clear results. If you examine this theoretically you find that isotropic material will not yield OU. The case for grain orientated material is IMO unproved and could be the basis for further experimentation. I seem to remember Steorn exhibiting something using rmagnets on a rotor passing over toroidal coils and doing just that, and they claimed OU. Maybe that was the precursor to their solid state Orbo.
Smudge
Yes, I experimented with that also but without clear results. Now, I think it's because I have too little experience with saturating magnetic core - it has to be precisely computed the amount of energy required to just switch core into saturation with capacitor discharge.
Then you take a core , for example ferrite toroid core, and wind two kinds of coils around it at 90 degrees to each other.First and second is for supplying the input low frequency input and output one. The third coil is the one which is 90 degrees to the input/output coils and connected to the capacitor in disruptive discharge mode. What you get is the output should be HF very large power modulated to low frequency. Input coil should be protected from HF. It should work imho because I believe that the output power depends only on magnetic field strength and frequency and the core cannot pass two magnetic fluxes at 90 degrees one to another due to exclusive one direction spin orientation of electrons in magnetic domains . I may be mistaken but most of the time I have the correct intuition.
Quote from: Peterae on 2017.09.08, 18:27:34
Tried the 100 Ohm resistor, very hard i only had 10Watts and needed 100Watt so had to really limit time it was on, firstly i have 4mv pk-pk of high frequency noise with the scope probe connected and 1mv pk-pk without the probe even connected.
Plus i get random bursts of noise 40mv pk-pk.
We need to get the effect over the 4mv noise otherwise i don't think we will ever see it.
Alternatively get the 4mV noise below the expected level. Since we expect something like a half sine wave at a frequency of 0.25 Hz and your noise is HF we could probably shunt the noise away with a capacitor, or several capacitors and resistors in a low pass filter orientation.
Smudge
Quote from: Smudge on 2017.09.08, 10:53:39
Yes, and if your saturating is done with a crossed field then the saturating drive does not directly couple to the coil so as to induce voltage.
Smudge
My past experimentation with crossed field windings in various core materials did show appreciable coupling during both the linear portions of flux change as well as into saturation. This led to core designs like the special E core in my "Vari-Perm" experiments which does a good job of isolating the saturating control flux winding from the source winding. However, the energy required to lower the permeability is such that the overall outcome is conservative. Still trying to work out a solution.
Pm
Some time bac,a post appeared on my facebook page,about energy gains when a crystal is vibrated at it's resonant frequency.
Apparently crystals give off an elecrical charge when placed under pressure,and this pressure was applied by winding a coil around the crystal,and pulsed at the crystals resonant frequency.
The crystal also glowed a blue/green color when the resonant frequency was hit-along with a high pitch ringing.
The electrical discharge from the crystal was captured within the inductive kickback from the coil.
Just what i read,and have no idea if there is any truth to it.
Brad
Quote from: Smudge on 2017.09.10, 13:32:58
Alternatively get the 4mV noise below the expected level. Since we expect something like a half sine wave at a frequency of 0.25 Hz and your noise is HF we could probably shunt the noise away with a capacitor, or several capacitors and resistors in a low pass filter orientation.
Smudge
Hi Smudge
Sounds good
What do you think about me trying a build of an active low pass filter with a gain of 100, set the top frequency at 10 Hz.
Quote from: Peterae on 2017.09.11, 17:11:57
Hi Smudge
Sounds good
What do you think about me trying a build of an active low pass filter with a gain of 100, set the top frequency at 10 Hz.
That should work OK, it will put the expected signal above the front-end noise level that your scope is seeing.
Smudge
That was a bit of a journey.
I have to Couple to the filter using a capacitor as i need to block any DC from the 100 Ohm resistor, and because when i use a capacitor to feed the filter it forms a high pass filter which would be no good, now assuming our minimum input impedance is 10K and hopefully more i had to use as large a coupling capacitor value as possible, i have selected 1000uf Bi-Polar electrolytic giving us at 10K impedance. giving a -3db Frequency =0.02 Hz, assuming the impedance is much higher than 10K then this 0.02 Hz will be much lower.
Now for the filter design.
We wanted a gain of 100 or 40db, i set the -3db at 5Hz and -40db at 15Hz.
Stage A 1st Order Low-Pass
Stage B 2nd Order Low-Pass Sallen Key
Stage C 2nd Order Low-Pass Sallen Key
Design was optimized for low noise, i will screen over the components with copper foil.
PCB has been designed and ordered.
I'm going to plug this PCB company, 3rd time i have used them, $2 a PCB although P&P was $10, 5 PCB's for $18 inc P&P, they even provide the cad software which is via a web app, https://easyeda.com/
Now just to add more confusion, but sticking to the subject heading of hidden ferromagnetic energy release, here is another paper I have just written. The partnered coils thread started by EMjunkie came to an abrupt halt when he went off in a huff so maybe it can be continued here, since it appears to me that there is something good to come from it.
Smudge