Author Topic: Confusing behavior in LCR/stepup transformer circuit - Help?  (Read 2987 times)

0 Members and 1 Guest are viewing this topic.

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
I have a stepup transformer driving a test load (46nF + 16 ohms), with a 1:10 turns ratio. The idea is that near resonance (~54kHz here) it will draw minimal current/power.

I've measured the transformer using an LCR meter (at the frequency range of interest) and with parasitics (leakage inductance), have the following circuit:


(I measured leakage by shorting one side of the transformer, and measuring inductance of the other, and vice versa; is that right?)

I'm sensing current using a small resistor on the ground side. (0.05ohm)

This all makes sense, the simulation shows a dip in current at around 54kHz. I would have expected the output to be flatter, but I guess the leakage inductance has a big effect. The result is here:



Anyway, when I build and measure it (source is an old power amp, not quite flat but OK), I get the following:



The gain looks reasonable, but why isn't current dipping near resonance? It's also trending up with frequency, rather than down.

Clearly something is wrong with my model, or my build, any ideas?  I've double-checked everything I could think to. Waveforms are clean and undistorted.
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #1 on: May 30, 2019, 04:48:48 pm »
It could be you have to add some parasitics into your simulation - your primary/secondary windings have got certainly a capacitance and the 47nF capacitor is rather an ideal one. Such a capacitor is usually modeled by several R, L, C.

You may get the Spice model for your capacitor as the output of this online app:
http://ksim.kemet.com/
« Last Edit: May 30, 2019, 05:00:06 pm by imo »
Readers discretion is advised..
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #2 on: May 30, 2019, 04:52:14 pm »
Thanks.  Interwinding capacitance is supposed to be less than 75pF.  The leakage inductance is (supposedly) included accurately.  The capacitors are film type, so they should be very clean.  What other parasitics should I be looking for?
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #3 on: May 30, 2019, 04:59:59 pm »
I measured the resistor's inductance as 2.5uH, and added that to the load (series).  I also added 75pF of capacitance between the primary and secondary (worst case).  The simulation had no noticeable change.

Maybe I'm getting my leakage wrong?  I'm pretty sure I did it right but I'm a little new to this.
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #4 on: May 30, 2019, 05:00:56 pm »
See above the capacitor model. Yours parasitic values may differ (depend on the capacitor type).
The coil's windings have the capacity in parallel to its L.
Do measure the self resonance freqs of the primary and secondary and you will get Cprimary and Csecondary.
« Last Edit: May 30, 2019, 05:12:19 pm by imo »
Readers discretion is advised..
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #5 on: May 30, 2019, 05:25:33 pm »
The capacitor is pure at this frequency range.  There is no notable inductance, and well under an ohm of ESR.  That's probably not the cause.

The transformer has a published max IWC of 75pF, so we're nowhere near self-resonance.  It's also designed for this frequency range (not 60Hz).

Maybe my leakage measurements were done incorrectly?
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #6 on: May 30, 2019, 06:28:11 pm »
You miss the output impedance of the amplifier (your AC source in simulation).
« Last Edit: May 30, 2019, 06:32:40 pm by imo »
Readers discretion is advised..
 
The following users thanked this post: me_engineer

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #7 on: May 30, 2019, 07:14:22 pm »
Ahh, that must be it.  I figured the transformer and LRC was independent of that, but it's not.  I don't know what the output impedance is, it's some inexpensive power audio amp.  I'll see if I can measure it.

Any tricks to measuring the output impedance directly? 
« Last Edit: May 30, 2019, 07:21:34 pm by me_engineer »
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #8 on: May 30, 2019, 07:31:28 pm »
Thinking further, I'm not sure this is the whole story unless there is a very unusual component in the amp.  What power amp has 10 ohms of output impedance?
 

Offline radiolistener

  • Super Contributor
  • ***
  • Posts: 4135
  • Country: 00
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #9 on: May 30, 2019, 07:36:03 pm »
Any tricks to measuring the output impedance directly?

You can put some sine signal from generator with fixed amplitude (for example 1 kHz) and measure output amplitude without load (open terminals). Next add different resistors as a load on amplifier output and measure amplitude with resistors. Find which resistor leads to exact half amplitude in compare with open terminals amplitude. This resistor will have the same value as your amplifier impedance.

You can evaluate it by measuring amplitude with two different load (for example R1=10 Ohm and R2=20 Ohm) and then by calculating:

Z = ( R1 - R1*(V1/V2) ) / ( V1/V2 - R1/R2 )
where
V1 - amplitude with load R1
V2 - amplitude with load R2

After that connect resistor with calculated value and check that amplitude is a half of value with open terminals.
Also you can use this way to evaluate output impedance if your amplifier is not intended to work with no load.
« Last Edit: May 30, 2019, 07:51:11 pm by radiolistener »
 
The following users thanked this post: me_engineer

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #10 on: May 30, 2019, 08:38:44 pm »
I did this measurement:
Open circuit output - 2.29Vrms
15 ohms output - 2.25Vrms
5 ohms output - 2.16Vrms
Z = ( R1 - R1*(V1/V2) ) / ( V1/V2 - R1/R2 )
(15-15*(2.25/2.16))/(2.25/2.16-15/5) = 0.32 ohms.

This is nowhere near enough impedance to explain the discrepancy. 

Perhaps there is a large reactance of some sort as well in the amplifier output?  It's Class H, which I don't think has any kind of output transformer or anything?  But I don't know the internals.

Maybe someone could verify that I'm measuring my transformer correctly?

Thanks for all the help!
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #11 on: May 30, 2019, 08:59:00 pm »
Does an inexpensive audio amplifier work till 50-70kHz flat?
Readers discretion is advised..
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #12 on: May 30, 2019, 09:08:30 pm »
Not really that flat, but flat enough for these purposes.

The two features that are really confusing is the lack of improved efficiency (a dip in current) at resonance, which is absent in the physical version, and the current trend vs. frequency is opposite of what the simulation predicts.
 

Offline radiolistener

  • Super Contributor
  • ***
  • Posts: 4135
  • Country: 00
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #13 on: May 30, 2019, 11:11:48 pm »
your transformer has 1:100 impedance ratio transformation (1:10 turns ratio).
What do you expect from this circuit on the output?
R3 is your output?

Did you check whats going on for entire 0...300 kHz range?

 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #14 on: May 30, 2019, 11:34:15 pm »
Output impedance should be factored out due to measuring the source voltage and calculating gain from that point.

What is this component?  Is there any way for us to check that your LCR measurements are correct?

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #15 on: May 31, 2019, 01:10:30 pm »
your transformer has 1:100 impedance ratio transformation (1:10 turns ratio).
What do you expect from this circuit on the output?
R3 is your output?

Did you check whats going on for entire 0...300 kHz range?

My expectation was that the model would match the physical circuit.  Those two features (lack of current dip at resonance and current trending upward, rather than downward, with frequency) mean my model isn't correct.  I'm wanting to find out why so I can model this kind of thing correctly.

The combination of R3 and C2 is the load, but really I want to know why my model is not at all matching my measurements, that's the important part.
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #16 on: May 31, 2019, 01:16:50 pm »
Output impedance should be factored out due to measuring the source voltage and calculating gain from that point.

What is this component?  Is there any way for us to check that your LCR measurements are correct?

Tim

The stepup gain seems to be correct.  It's the other mismatch between measurements and model that I'm wondering about.

The transformer is something we've had around for a similar project years back, and was intended for this frequency range. I measured it with a decent (if older) HP LCR meter, at the frequency of interest, as:

Primary inductance (and R) - measured L+R series value across primary with secondary open, provides L4 and R6 values directly
Secondary inductance (and R)- measured L+R series value across secondary with primary open, provides L5 and R7 directly
Primary leakage - shorted secondary, measured inductance across primary, provides L8 directly
Secondary leakage - shorted primary, measured inductance across secondary, provides L7 directly

This seem the right way to characterize the transformer?  The numbers match the manufacturer's figures pretty well (though they weren't explicit about leakage, just stated "maximum").

If so, the discrepancy is probably from some substantial output impedance (reactive) in the amplifier that's tough to measure, since the output resistance is very small.  Maybe a huge inductance?  I'm not sure what else to check.
 

Offline The Electrician

  • Frequent Contributor
  • **
  • Posts: 747
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #17 on: May 31, 2019, 01:35:44 pm »
Can you post the transformer's spec sheet (or a link to it)?  Is it spec'd as an audio transformer?

Measure and plot the impedance at the primary with the load connected over the frequency range 10 kHz to several hundred kHz.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #18 on: May 31, 2019, 04:21:57 pm »
Primary inductance (and R) - measured L+R series value across primary with secondary open, provides L4 and R6 values directly
Secondary inductance (and R)- measured L+R series value across secondary with primary open, provides L5 and R7 directly
Primary leakage - shorted secondary, measured inductance across primary, provides L8 directly
Secondary leakage - shorted primary, measured inductance across secondary, provides L7 directly

This seem the right way to characterize the transformer?  The numbers match the manufacturer's figures pretty well (though they weren't explicit about leakage, just stated "maximum").

Ah. Hm. Well the leakage is wrong -- you can't measure these independently, best you can do is hand-wave and split it in half.

The numbers are also inconsistent with the inductance ratio, which I also wonder if that's actually the turns ratio or not.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #19 on: May 31, 2019, 04:39:06 pm »
Hmmm... what you have drawn is a series resonant LCR network, hence you should see a maximum in current at the resonant frequency.

Also, the preferred way to represent leakage inductance in LTSpice is by reducing the coupling coefficient, K, from 1 (leakage fraction is then equal to 1-K^2). The simulation will sometimes run faster if you model leakage as an external inductor, as you have done, but then you should only put it on one winding as it will be correctly reflected to the other windings by K=1.

Thirdly, inductance scales with the square of the turns ratio, so the secondary inductance should be 250uH, not 168uH.

Finally, yes, the practical way to measure leakage is to short out the primary (secondary) and measure the inductance across the secondary (primary), assuming there are just two windings; if there are more than two then technically you need to short out all of them except for one to measure the leakage, but unless the leakage is really high then just shorting any winding with a high VA rating (e.g. - the primary or the main secondary) will suffice.

 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #20 on: May 31, 2019, 04:43:05 pm »


Quote from: The Electrician on Yesterday at 11:35:44 pm
Can you post the transformer's spec sheet (or a link to it)?  Is it spec'd as an audio transformer?

Measure and plot the impedance at the primary with the load connected over the frequency range 10 kHz to several hundred kHz.



Thank you.  I see now this is the correct approach and is very helpful.  I should have done this from the start.

Published Specs: (@100kHz)

Sec Inductance: 156-173uH
Leakage (sec) 70uH max
IWC: 75pF max
Turns ratio: 1:10


Measurement and simulation:








These seem quite close, at least qualitatively (I did reduce IWC for the above plot for a better fit)


The big discrepancy must be the amplifier's output impedance, having a nontrivial reactive component.  This cleanly confirms it I think.
 

Offline The Electrician

  • Frequent Contributor
  • **
  • Posts: 747
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #21 on: May 31, 2019, 07:41:03 pm »
Does the spec indicate a recommended working impedance for the windings?  Something like 100:10k ohms?  Is this transformer wound on a ferrite core?
 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #22 on: May 31, 2019, 08:52:51 pm »
Does the spec indicate a recommended working impedance for the windings?  Something like 100:10k ohms?  Is this transformer wound on a ferrite core?

Nope, it only specified what's above (and some maximum resistances).  The values I used in the simulation were ones I measured myself though.  (I'm still a little uncertain if I did leakage correctly).

It is indeed wound on a ferrite core (over a bobbin).

 

Offline me_engineerTopic starter

  • Contributor
  • Posts: 15
  • Country: us
Re: Confusing behavior in LCR/stepup transformer circuit - Help?
« Reply #23 on: May 31, 2019, 08:55:40 pm »
Hmmm... what you have drawn is a series resonant LCR network, hence you should see a maximum in current at the resonant frequency.

Also, the preferred way to represent leakage inductance in LTSpice is by reducing the coupling coefficient, K, from 1 (leakage fraction is then equal to 1-K^2). The simulation will sometimes run faster if you model leakage as an external inductor, as you have done, but then you should only put it on one winding as it will be correctly reflected to the other windings by K=1.

Thirdly, inductance scales with the square of the turns ratio, so the secondary inductance should be 250uH, not 168uH.

Finally, yes, the practical way to measure leakage is to short out the primary (secondary) and measure the inductance across the secondary (primary), assuming there are just two windings; if there are more than two then technically you need to short out all of them except for one to measure the leakage, but unless the leakage is really high then just shorting any winding with a high VA rating (e.g. - the primary or the main secondary) will suffice.

Thanks.  I prefer having discrete L's to show leakage, as long as it's correct.  Why wouldn't there be leakage on both coils separately?

The values in my simulation are all measured using the methods I described; the 168uH is correct.  Perhaps the difference is from leakage?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf