Loop gain measurement apparati - diyAudio
Go Back   Home > Forums > Design & Build > Equipment & Tools

Equipment & Tools From test equipment to hand tools

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 4th May 2014, 08:11 PM   #1
tomchr is offline tomchr  United States
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Greater Seattle Area
Default Loop gain measurement apparati

Measurements of loop gain and the ever popular gain and phase margins are causing confusion for many. I a recent thread, a couple of good posts on this topic emerged. I had a few thoughts myself, so I figured I'd get a new thread started collecting this information.

One way to measure the loop gain is to inject a disturbance signal using a current probe. That is covered well in this article from Maxim: Injection Transformers for Closed-Loop DC-DC Converter Network Analysis - Application Note - Maxim

Current probes generally have rather limited bandwidth, hence, for measurements closer to DC, an op-amp circuit may be more useful. Such a circuit is described in an old app note for the HP 3577A network analyzer: http://cp.literature.agilent.com/lit.../3577-5131.pdf

Enjoy.

~Tom
__________________
21st Century Maida Regulator, Universal Filament Regulator, Damn Good 300B, Novar Spud, and more...
Neurochrome : : Audio - http://www.neurochrome.com/audio - Engineering : : Done : : Right

Last edited by tomchr; 4th May 2014 at 08:14 PM.
  Reply With Quote
Old 4th May 2014, 11:52 PM   #2
diyAudio Member
 
sumaudioguy's Avatar
 
Join Date: Jan 2010
Location: Midwest U.S.A.
Hope all this is more or less on subject.

Maybe something of use about dynamic range-

What has been the most detailed driver you have worked with?

And maybe this one-

Another view of damping factor?

This is just to help in general.

For open loop gain I simply put a low frequency DC servo on the amp so it will operate without any feedback above 0.01Hz (-3dB on the servo) . It is straight forward from there to measure gain, time delay, frequency response, phase margin, or what ever else is desired.

To test damping factor and impedance of the output, driving the output of the amplifier under test with a different amplifier through various resistors or networks while monitoring signals of the amplifier under test at its' output and the input can be very informative. This is much more informative than loading the amplifier and comparing the no load output voltage to the loaded output voltage to calculate damping factor. A complex model can be created using this method.

More in this thread-
Solid State Testing- Tug of war
__________________
What the other guy said----Standing on the shoulders of giants.
New avatar- no more little array
  Reply With Quote
Old 5th May 2014, 12:19 AM   #3
tomchr is offline tomchr  United States
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Greater Seattle Area
So... I decided to build a Rogowski transformer. I started with a small ferrite toroid that I picked up at a surplus store a while back. I wound 10 turns on it and measured the inductance: 660 uH for 10 turns.

A 15:5 transformer on this core had a usable bandwidth of about 13 MHz. That's not quite enough for the loop gain measurements I'd like to do. So I turned it into a Rogowski transformer by splitting the 50 Ω termination into two 22 Ω resistors (the closest to 25 Ω I had available). With two 15-turn primaries connected in series with 22 Ω*across each primary and the same 5-turn secondary, the bandwidth is now 70+ MHz. I was pretty blown back that I could push 70 MHz through this Daddy Longleg...

~Tom
Attached Images
File Type: jpg Rogowski.jpg (186.0 KB, 186 views)
File Type: gif RogowskiTransformer.gif (16.2 KB, 182 views)
__________________
21st Century Maida Regulator, Universal Filament Regulator, Damn Good 300B, Novar Spud, and more...
Neurochrome : : Audio - http://www.neurochrome.com/audio - Engineering : : Done : : Right
  Reply With Quote
Old 5th May 2014, 12:25 AM   #4
tomchr is offline tomchr  United States
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Greater Seattle Area
Quote:
Originally Posted by sumaudioguy View Post
For open loop gain I simply put a low frequency DC servo on the amp so it will operate without any feedback above 0.01Hz (-3dB on the servo) . It is straight forward from there to measure gain, time delay, frequency response, phase margin, or what ever else is desired.
Now, that's cool. I hadn't thought of using a DC servo. One has to be careful with the input signal amplitude, though.

Ironically, that's the exact same method I use in simulation, though, rather than a DC servo, I use a monster LC filter to reject everything except DC. I typically use 1 MH, 1 MF in simulation. That's capital M for mega (1E6) not lower case m for milli (1E-3).

~Tom
__________________
21st Century Maida Regulator, Universal Filament Regulator, Damn Good 300B, Novar Spud, and more...
Neurochrome : : Audio - http://www.neurochrome.com/audio - Engineering : : Done : : Right
  Reply With Quote
Old 5th May 2014, 01:25 AM   #5
diyAudio Member
 
panson_hk's Avatar
 
Join Date: Aug 2005
Blog Entries: 10
Quote:
Originally Posted by tomchr View Post
So... I decided to build a Rogowski transformer. I started with a small ferrite toroid that I picked up at a surplus store a while back. I wound 10 turns on it and measured the inductance: 660 uH for 10 turns.

A 15:5 transformer on this core had a usable bandwidth of about 13 MHz. That's not quite enough for the loop gain measurements I'd like to do. So I turned it into a Rogowski transformer by splitting the 50 Ω termination into two 22 Ω resistors (the closest to 25 Ω I had available). With two 15-turn primaries connected in series with 22 Ω*across each primary and the same 5-turn secondary, the bandwidth is now 70+ MHz. I was pretty blown back that I could push 70 MHz through this Daddy Longleg...

~Tom
Cool! You are quick.

My one is entirely based on the Maxim AN as shown in the picture.
Attached Images
File Type: jpg injection probe.jpg (453.2 KB, 175 views)

Last edited by panson_hk; 5th May 2014 at 01:43 AM.
  Reply With Quote
Old 5th May 2014, 12:51 PM   #6
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
Often, you can't break the feedback loop to neatly measure phase and gain. Omicron labs describes a method for non-invasive loop measurement by examining Zout and phase angle -- most of the PDF is used to describe how their current injectors and Bode-100 analyzer are to be used, but the method is easily adapted to the equipment we have at hand:

http://www.omicron-lab.com/fileadmin...ility_V1_0.pdf

Condensed version here, but annoyingly you have to watch the pop-up:
Evaluate Feedback Stability | Electronic Design | Test & Measurement content from Electronic Design

J
  Reply With Quote
Old 5th May 2014, 04:37 PM   #7
diyAudio Member
 
sumaudioguy's Avatar
 
Join Date: Jan 2010
Location: Midwest U.S.A.
Quote:
Originally Posted by jackinnj View Post
Often, you can't break the feedback loop to neatly measure phase and gain. CLIP

J
That being the case, the amplifier is unconditionally unstable because with a bad load and assuming voltage feedback the delay of the feedback signal caused by the bad load will "break" the loop. I do recognize certain designs work well with known loads which require feedback to function. This is not a good choice for audio because the load is indeterminate.
__________________
What the other guy said----Standing on the shoulders of giants.
New avatar- no more little array
  Reply With Quote
Old 6th May 2014, 01:10 AM   #8
diyAudio Member
 
panson_hk's Avatar
 
Join Date: Aug 2005
Blog Entries: 10
Here is another article describing signal injection using transformer. Ridley Engineering | - AP300 Loop Injection

I will post some data later.
  Reply With Quote
Old 6th May 2014, 04:33 AM   #9
diyAudio Member
 
panson_hk's Avatar
 
Join Date: Aug 2005
Blog Entries: 10
Default LM3886 loop-gain measurement

The LM3886 shown below is configured as non-inverting amplifier with 1k and 33k as gain setting (feedback) resistors. Loop gain is measured by HP3577A with signal injection method outlined in the Maxim app note.

The signal injection probe has a useable bandwidth from approx. 200 Hz to 10 MHz. Hence, HP3577A frequency range is set to 100 Hz to 10 MHz. From the figure of measured response, we can see roll off around 100 Hz and "noisy" curve due to very high gain. This is not a problem for amplifier loop analysis where high-frequency roll off characteristic is our major concern.

The 0 dB point is where open-loop and closed-loop gain intersect. It can be seen the frequency and phase of this point are 347 kHz and 87.5 (~90) deg, respectively. Phase margin of this particular circuit is thus 90 deg.
Attached Images
File Type: jpg LM3886.JPG (572.8 KB, 125 views)
File Type: jpg LM3886 loop gain.JPG (638.4 KB, 122 views)

Last edited by panson_hk; 6th May 2014 at 04:54 AM.
  Reply With Quote
Old 7th May 2014, 01:11 AM   #10
diyAudio Member
 
sumaudioguy's Avatar
 
Join Date: Jan 2010
Location: Midwest U.S.A.
Nice!

Can I use my 50MHz TEK current probe for this?

Anyone ever use lissajous plot method for this? Have many times in the past.

Would really like to put my 3886 amp in this test just to see. Its gain is really high, 240dB at DC, so there is huge loop gain.
__________________
What the other guy said----Standing on the shoulders of giants.
New avatar- no more little array
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Control-to-Output or Open-Loop Measurement experimentador Power Supplies 7 21st February 2014 11:30 PM
Loop gain measurement in pspice enzo88 Solid State 9 5th September 2011 03:17 PM
Throwing away open loop gain with an opamp near or at unity gain okapi Pass Labs 3 25th December 2009 01:23 PM
Gain Blockwith Variable Loop Gain Bonsai Solid State 17 11th July 2008 02:40 PM
Open Loop Gain Measurement? Fotios Need Help! fotios Solid State 21 22nd April 2008 12:02 AM


New To Site? Need Help?

All times are GMT. The time now is 07:27 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2