PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2021-10-26, 19:07:09
News: A feature is available which provides a place all members can chat, either publicly or privately.
There is also a "Shout" feature on each page. Only available to members.

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 [13] 14 15 16 17 18 19 20 21 22 23 24
Author Topic: Investigating "anomalies" in Bifilar coils  (Read 155818 times)

Group: Elite Experimentalist
Hero Member
*****

Posts: 3214
Repeating my post #258 here below as i now have tried 2 different ways to hook up (daisy chain) a 2th TBP coil.
Like TK reports, it causes the OU effect to go away.

Todays setup with 2 TBP coils was like depicted by Gyula in his post #171 bottom drawing, just one 51 Ohm resistor on the output of the 2th coil, and a single 1 Ohm
csr in the return line to the FG for measuring the current



Leads shortend where possible yields a slightly higher COP @ 1MHz (1.22 instead of 1.2).

Linking my both identical TBP coils together (both equiped with a 51 Ohm load resistor and 1 Ohm csr) drops the output of the first to half while input stays the same.
COP now way below 1 (@ 1MHz)

Input of 2th TBP coil was from top of 51 Ohm load resistor of 1st TBP coil (CH3 purple probe point), ground of 2th TBP coil was via its 1 Ohm csr to the common ground point.

I also tried Picowatts suggestion for an alternative Pin method, but that did not match up (using Vrms values). FG open circuit value 7.127V, when connected to circuit 5.082V.
Difference is  2.045V, divided by 50 gives 40.9mW while the measured input power was 147mW


Itsu


Itsu
   
Group: Tech Wizard
Hero Member
*****

Posts: 928
Hi Itsu,

Just wondering on your above test of 2 TBP coils chained in series: so far all the COP>1 results for a single TBP coil have been found between real terminating resistances (the FG's 50 Ohm at the input of a TBP and another 50 Ohm (or other real termination) across L2 i.e. the output. Now when two such TBPs are chained, a matching problem may happen between them which "consumes" useful power. We need to figure out a matching means, perhaps introducing an additional LC circuit in the right configuration at specific frequency to couple the output of the 1st TBP to the input of the 2nd TBP.

Thanks for doing the tests.

Gyula
   
Group: Guest
TK,

The schematic you provided showed four probes connected to the circuit, with two of those probes across R2b.  I thought you were just measuring the difference between those two probes for the R2b Vdrop.  Did you confirm that removing the other probes from the circuit had no effect on the R2b Vdrop or phase angle?  (Luc's measurements changed just from switching a probe's attenuation factor)
Yes, I was trying to do a differential measurement across R2b but could not get sufficient resolution on the scope for it to be meaningful. A small difference between two large signals, just couldn't be resolved accurately.

I'm not sure I know how to tell if removing the probes had an effect on phase angle....   :P

Quote

Was R2b positioned close to the FG's output or the coil?  What does the Vdrop across R2b look like with R2b positioned right at the FG output

 Do you mean moving R2b physically over to the function generator itself? I don't have probe cables long enough to reach and it is impossible to rearrange items in my space at the moment.

Quote

I would be very cautious about connecting a shield to various measurement points and just assuming it is having no effect elsewhere on this circuit.  Floated grounds should only be considered as isolating the signal and chassis/AC ground via a capacitance in parallel with a large resistance.  That capacitance and resistance can be measured.

I could not find a remove or delete button, apparently newbies don't have full driving privileges.

No worries, some kind soul cleaned up my mess...

PW

So now I have isolated the FG completely by using ION's suggestion of using a coupling transformer. Of course I just guessed at what might work, and it seems to have worked in spite of that!
I think I still have to conclude that the measured Vdrop across each R2 resistor is the same or very nearly so.
   
Group: Guest
TK,

This is very interesting!  Is this the same coil you started with that you previously posted the specs for?  Also having the dimensions we could calculate the MLT to determine the number of turns if you don't have that info.  I'm curious as to what your interpretation is of this peaked induction?

Pm

Yes, it is a "sandwich" of the TBF coil that I converted into the "PBF" coil or transformer, and the monofilar coil with the same turn count. Both are #27 magnet wire and both have 116 total turns, give or take one or two. They are wound on CD-ROM disks with the foil removed.
I am not sure about the frequency of the peaked induction, I was just trying to watch the DDS sweep frequency display at the same time watching the scope screen. It is possible that the peak actually happens at the monofilar coil's resonant frequency. I'll have to do finer sweeps to be able to tell more precisely. I do not have a Spectrum Analyzer unfortunately.

By the way, when the PBF coil is sitting at a particular "OU" frequency with phase shift in the -75 to -85 degree range, giving the COP>1 numbers, I can light up an LED load with the monofilar's output and I don't see any change in the 3 monitored signals on the scope.  I haven't tried heavier loads yet.
   
Group: Guest
Hi TinselKoala,

Would you mind to share some details on this?
What I am mainly curious about is
-phase shift at input goes away and retuning is needed to get the -80 to -82 degree phase shift when the resistor R1 removed from L2
-any other detail you have done on this test when you have a few minutes

Thanks
Gyula

I'll have to take another look at this testing before I can really say anything definite. The doubler does work to make a DC voltage output, that's about all I can say at the moment. I should have more results later this evening.
   
Group: Guest
I do not have data on reflections in this circuit, but the cos(phi) power factor can be used for the phase correction of the VRMS * IRMS product only for purely sinusoidal waveforms.  Any peak besides the first one indicates a departure from a pure sine wave, and the energy content of these peaks is not properly accounted for by the single cos(phi) power factor.

P.S.
I'm posting from a mobile phone now and I don't remember whether your spectrogram represented input current or input voltage.

This is an excellent point and one that I have also considered. Certainly my DDS FG does not make a very good sine wave especially at high output amplitude and frequency. I think the analog F43 is better, but it really gets bad near the top frequency range (which is under 4 MHz.) But in the 1-2 MHz range it is still pretty good I think. I don't have a Spectrum Analyser but the Rigol z-box can do a crude FFT display so might be able to say something about the spectral purity of the signal input I'm using. More work ahead!

Do we have some kind of idea just how far off the basic formula will be, for some degree of distortion in the sine wave? Do you think something like this might account for the fact that my scope's automatic measurement of average power is consistently about 20 percent higher than the manual calculation indicates?
   
Group: Guest
This is the sort of FFT that the z-box has. It's quite fiddly, takes some time to get an interpretable display.
   
Group: Guest
... And using this tool I can now see that I was wrong-- the DDS FG sine is actually somewhat cleaner than the analog F43 at this frequency ! I am surprised.

   
Group: Elite
Hero Member
******

Posts: 3541
It's turtles all the way down
TK

If you put a compression trimmer across the primary or secondary (preferable) of the wideband isolation toroid and a 50 ohm resistor or so from the primary back to the generator, you should be able to ring the transformer winding at the frequency of interest to a nice clean low distortion sine wave regardless of the generator output having some distortion products, this providing of course that you do not get near the saturation of the core. An air core would be more linear, hence lower distortion and lower distributed capacitance coupling than the ferrite.
But maybe your DDS is far cleaner than a ferrite core / cap in resonance...I don't know.

Just a thought.

BTW I have not done any upgrades to my scope...waiting for the dust to settle on revisions. Are we close to a good rev# at this point?

Kind Regards


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   

Group: Professor
Hero Member
*****

Posts: 2421
I don't have a Spectrum Analyser but the Rigol z-box can do a crude FFT display so might be able to say something about the spectral purity of the signal input I'm using. More work ahead!
Indeed it looks well according to that FFT.

Do we have some kind of idea just how far off the basic formula will be, for some degree of distortion in the sine wave?
Without quantifying the distortion it would be difficult. We'd have to do a cos(phi) correction for each harmonic separately - a tedious job at best.

But perhaps some rough limits can be established from the fact that e.g. inductive reactance to a square wave is approximately 19% smaller than the reactance to a pure sine wave of the same frequency, (due to the additional harmonic content in the square wave).
Of course we don't have a square wave anywhere, but Itsu's spectrogram showed even harmonics which are not present in a square wave.

The Ch1 * Ch2 math channel averaging method doesn't care about i&v harmonics (and their phase) so that might account for the difference with that VRMS * IRMS * cos(θ) method.
« Last Edit: 2017-05-04, 21:52:24 by verpies »
   
Group: Experimentalist
Hero Member
*****

Posts: 1365
All,

In setting up some comparison tests at 2MHz using coils with 3 or more windings, I have discovered that the voltage across the sense resistor R2 leads the voltage developed across the load resistor R1 by several degrees.  This can be seen in my post #90.  I do need to qualify this test by saying that the lower secondary connection to the junction of R1 and R2 is open.  Essentially I discovered in my 2 coil bifilar, there is no current in this winding so the secondary can be open ended.  The performance does not change so the induction is totally via displacement current in this configuration.  Those with the TBP need to confirm this is also the case with your coils.

These voltages should be in phase if all is right with no current entering the top of sense resistor R2.  I immediately suspected R1 to be somewhat inductive but resonance tests proved that is not the case.  Adjusting and rearranging the scope ground connections helped slightly but not completely.  Using an iso transformer between the generator and the circuit did lower the phase to <1 degree which is acceptable.

I have to say that I have designs with 3 coils that do have current exiting the secondary and entering the top of the sense resistor with different magnitudes and phase than the load current, so I didn't think there was anything unusual with traces being shown including my own.

Anyway I'm not sure at this point if this affects the overall results but with these low level tests direct from the generator, it would be wise to isolate the signal generator ground from the test set as ION suggested.  I might add that for comparison, I've gone thru all the posted tests and TK's posts #98, 116, 127, 208, and 298 (compare to 297) appear to be valid measurements in this regard.  Some of the others have considerable phase lead.

Pm

Edit:  I will again stress the fact that complete cycles must be measured either in the scope viewing window or between cursors.  This will become imperative as the output wattage levels get into single and double digits.   
   

Group: Professor
Hero Member
*****

Posts: 2421
...it would be wise to isolate the signal generator ground from the test set as ION suggested. 
It was also my suggestion and I think it predated ION's

- Putting a 1:1 ferrite isolation transformer on the output of your FG...
   
Group: Elite
Hero Member
******

Posts: 3541
It's turtles all the way down
It was also my suggestion and I think it predated ION's

You may very well have predated the suggestion. To be honest, I don't recall reading it, but you are certainly experienced enough in these matters to have suggested it.

It was the kind of thing I did routinely on the bench 30 or 40 years ago when I needed isolated signals, without even thinking that much about it, second nature.

The other morning, upon waking I had the problems of measuring the circuit on my mind, when it occurred to me that what was needed was a way to measure current on the input side of the TBC. The only reasonable way to do it was complete low capacitance isolation of  the SG output, but using an isolation transformer on the power supply for the SG would still leave a rather high capacitance coupling in the circuit. Since the frequencies were quite high, the small ferrite on the output of the SG was the next logical choice. I was thinking wideband transformer like those found in TV coupler / distribution modules.

I often try to give credit to the originator of an idea or suggestion wherever possible, it is just good ethics. It created much loyalty when I was heading a group of engineers back in the day.

So the original credit for the suggestion can remain yours.

Kind Regards
« Last Edit: 2017-05-05, 01:32:24 by ION »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   

Group: Administrator
Hero Member
*****

Posts: 3172
It's not as complicated as it may seem...
Do we have some kind of idea just how far off the basic formula will be, for some degree of distortion in the sine wave?
This should be relatively easy to test, and I would suggest using "kinder" frequencies such as the audio range where your scope is sure to perform well. One way would be to put a 1Vp sine signal into the attached circuit with the diode(s) removed and measure the baseline distortion of your generator/op-amp. Then insert a single diode for even-harmonic (2nd) distortion, or add a second diode in reverse parallel with the one shown for odd-harmonic distortion, and vary the amount of distortion by varying the pot. You may have to go up to a 500k or 1M pot to get more subtle amounts of limiting and distortion. Perhaps you can now compare the two measurement methods by driving a small load with the op-amp.

Quote
Do you think something like this might account for the fact that my scope's automatic measurement of average power is consistently about 20 percent higher than the manual calculation indicates?
That is troubling.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3214

I started to use a 1:1 ferrite isolation transformer yesterday, (see picture) and it lowered the COP @ 1Mhz from 1.2 to 1.03.
No more COP > 1 on other frequencies.

A spectrogram with or without the 1:1 ferrite isolation transformer however, does not show any significant difference.

I will be waiting for my caddock non inductive resistors to arrive for doing further measurements, but its not looking good.


Itsu
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4409


Buy me some coffee
I started to use a 1:1 ferrite isolation transformer yesterday, (see picture) and it lowered the COP @ 1Mhz from 1.2 to 1.03.
No more COP > 1 on other frequencies.

A spectrogram with or without the 1:1 ferrite isolation transformer however, does not show any significant difference.

I will be waiting for my caddock non inductive resistors to arrive for doing further measurements, but its not looking good.


Itsu

Itsu

Are you measuring the P/in after the 1:1 transformer,or at the input of the 1:1 transformer.
In other words,is the power the transformer is consuming,being included in the P/in measurements ?

BTW--it is great to see you working on this  O0


Brad


---------------------------
Never let your schooling get in the way of your education.
   
Group: Guest
This should be relatively easy to test, and I would suggest using "kinder" frequencies such as the audio range where your scope is sure to perform well. One way would be to put a 1Vp sine signal into the attached circuit with the diode(s) removed and measure the baseline distortion of your generator/op-amp. Then insert a single diode for even-harmonic (2nd) distortion, or add a second diode in reverse parallel with the one shown for odd-harmonic distortion, and vary the amount of distortion by varying the pot. You may have to go up to a 500k or 1M pot to get more subtle amounts of limiting and distortion. Perhaps you can now compare the two measurement methods by driving a small load with the op-amp.
That is troubling.

I'll see what I can do, but it may take a couple of days for me to get all that together.  At the moment I'm taking the scope's measurements at face value (I have carefully cross-checked the automatic measurements of things like Vp-p and phase angle with manual cursor measurements and they are fine.) I attribute the higher Math Average measurements to greater accuracy doing the instantaneous multiplication, as opposed to the manual calculation that assumes sine waves and misses the power in higher harmonics. But even using the Math average I still have no trouble getting COP>1 measurements in most cases.
   
Group: Guest
I've spent all night making a detailed video demonstrating the setup and the phenomenon and another phenomenon we haven't yet discussed here. The video is about 15 minutes long. It will take some time (hours) to render and upload.

I intend to post the link here privately at first (via PM to the experimenters and other interested members) , unless Partzman gives permission to post it publicly. I'm sure it will cause a storm of controversy.

By the way, that reposting of the Transmission Line video to Hackaday has caused an incredible inflation of the view count over the past couple of days. It went from about 500 views yesterday morning, which is about what I usually get for these things, to nearly 4,500 views by yesterday evening. It seem to have topped out for now at about 4800 views. And not even one single "thumbs down" ! That is extremely unusual, my trolls must be asleep.

Front title slide and schematic for the upcoming video:

   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4409


Buy me some coffee
I've spent all night making a detailed video demonstrating the setup and the phenomenon and another phenomenon we haven't yet discussed here. The video is about 15 minutes long. It will take some time (hours) to render and upload.

I intend to post the link here privately at first (via PM to the experimenters and other interested members) , unless Partzman gives permission to post it publicly. I'm sure it will cause a storm of controversy.

By the way, that reposting of the Transmission Line video to Hackaday has caused an incredible inflation of the view count over the past couple of days. It went from about 500 views yesterday morning, which is about what I usually get for these things, to nearly 4,500 views by yesterday evening. It seem to have topped out for now at about 4800 views. And not even one single "thumbs down" ! That is extremely unusual, my trolls must be asleep.

Front title slide and schematic for the upcoming video:

Watching my inbox  ;)

This phenomenon wouldnt have anything to do with !!odd!! temperature variations with the resistors-would it?.


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3214
Itsu

Are you measuring the P/in after the 1:1 transformer,or at the input of the 1:1 transformer.
In other words,is the power the transformer is consuming,being included in the P/in measurements ?

BTW--it is great to see you working on this  O0


Brad

Hi Brad,

its after the 1:1 xformer, like depicted by TK's diagram here above.

The caddocks are in, so time to fire up the solder iron.....


Itsu

   

Hero Member
*****

Posts: 738
Front title slide and schematic for the upcoming video:

Now with isolation transformer it would be resonable test setup with grounded secondary.
I think there is big chance that OU goes away with this setup.


---------------------------
The only way of discovering the limits of the possible is to venture a little way past them into the impossible.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4409


Buy me some coffee
Hi Brad,

its after the 1:1 xformer, like depicted by TK's diagram here above.

.....


Itsu

Quote
The caddocks are in, so time to fire up the solder iron

 O0

Well,it seems that the old BPC out-performs the new solenoid type i made.

Using Partzmans circuit,i can get constant COP>172-176%

Seems the P/in measurements should be correct,as the voltage and current are in phase,meaning a power factor of 1. RMS value's are now used to calculate P/in,and P/out is the voltage across the resistors-ohms law.

So,it would seem that at least 3 here are seeing constant COP> result's--but dose anyone know why yet?,or where the red herring is hiding?

It has been great research at the very least--who'd of ever thought that current can actually travel in both directions,at the same time,through a coil  :o

Seem to remember some one once saying that,and most of us laughed at him.
Maybe time !for me anyway! to go and eat some humble pie.


Brad


---------------------------
Never let your schooling get in the way of your education.
   
Group: Guest
Don't fool yourself... current is the flow of charge, from higher potential to lower potential. I know some people who "go both ways" at once,  but current doesn't.

   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4409


Buy me some coffee
Don't fool yourself... current is the flow of charge, from higher potential to lower potential. I know some people who "go both ways" at once,  but current doesn't.

Not sure if there was any type of message in the !!I know some people who "go both ways" at once!! ,but are you saying the phase inversion we see across R1 and R2 is not the current starting to flow in the opposite direction?

As far as i know,when two current flows are 180* out of phase with each other,then they are flowing in opposite directions--of have i missed something some where?


Brad


---------------------------
Never let your schooling get in the way of your education.
   
Group: Experimentalist
Hero Member
*****

Posts: 1365
Don't fool yourself... current is the flow of charge, from higher potential to lower potential. I know some people who "go both ways" at once,  but current doesn't.

TK,

I'm really curious to hear/see the apparent anomaly you've experienced.  I would normally agree with your statement above regarding current flow but unless I misinterpreted the results, I have measured current exiting or entering a coil simultaneously in certain MEI configurations.  We'll get around to that at some point.

Pm
   
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 [13] 14 15 16 17 18 19 20 21 22 23 24
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2021-10-26, 19:07:09