Bob Cordell's Power amplifier book - Page 148 - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

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 6th February 2011, 12:42 PM   #1471
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by Olivier73 View Post
Hi Bob, Hi Macleod,

I understood that CL phase and gain measurement are pretty useless. So even if the CL gain goes throug zero at 10MHz and the phase at that point is already way beyond 180° lagging, that is not a direct concern... right?
Indeed measuring openloop (while breaking the loop with a large inductor) the result seem promising. If this figure defines the PM and GM I am happy.
FYI, for now, the compensation capacitor over the VAS is a simple miller cap, however stretching over a darlington and cascoded vas...

Bob, according to me the 36dB is normal. Where do you deduct from that the gain should be 26dB?

I will now try to incorporate TMC where I follow Edmonds view. But anyway according to your book and the many forum topics here the TMC camp seems to be more convincing than the TPC camp. My main reason is that I don't like the peaking in the bode plot...

Olivier
Hi Olivier,

The closed loop gain for the amplifier in my book should be 20, since the feedback network consists of a 19k and a 1k resistor, which results in a 20:1 attenuation in the feedback path. A voltage ratio of 20:1 corresponds to 26 dB. That's why I'm not sure how you got 36 dB.

I agree that TMC is preferable to TPC, at least in my opinion. However, the TMC vs TPC discussion that we had here brought more understanding to the problem, particularly in regard to analysis for stability.

Cheers,
Bob
  Reply With Quote
Old 6th February 2011, 02:01 PM   #1472
diyAudio Member
 
Join Date: Dec 2009
Location: Brussels
Default PM & GM

Hi Bob,

OK I see what you mean. In my circuit the feedback resistors are set at 15K and 270R and therefore the midband gain is around 36dB.

About compensation... I didn't go through all of the book yet but I read and saw several times on the net that placing a resistor in series with the compensation capacitor includes a zero in the bodeplot.
In TPC and TMC it is possible to insert 2 resistors in one would like... but what could be the gain of this? Which side to put it? Both sides?

Cheers,
Olivier
  Reply With Quote
Old 6th February 2011, 09:40 PM   #1473
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by Olivier73 View Post
Hi Bob,

OK I see what you mean. In my circuit the feedback resistors are set at 15K and 270R and therefore the midband gain is around 36dB.

About compensation... I didn't go through all of the book yet but I read and saw several times on the net that placing a resistor in series with the compensation capacitor includes a zero in the bodeplot.
In TPC and TMC it is possible to insert 2 resistors in one would like... but what could be the gain of this? Which side to put it? Both sides?

Cheers,
Olivier
Hi Olivier,

Great question about compensation. In most of my designs, I have made provision for a resistor in series with the Miller compensation capacitor. This includes the Miller input compensation scheme that I used in my MOSFET power amplifier with error correction. You are correct - that resistor adds a zero in the forward path by "stopping" the continuation of the 6 dB/octave rolloff of the Miller compensation. This zero must be used judiciously or it will cause a loss of gain margin. The zero is usually placed well above the gain crossover frequency.

Although I have not used it with TMC, I know of no reason why it cannot be carefully applied to a TMC compensation scheme. Without having given it much thought, I would add the resistor between C1 of the TMC network and the input base of the VAS.

Cheers,
Bob
  Reply With Quote
Old 8th February 2011, 08:24 PM   #1474
diyAudio Member
 
Join Date: Dec 2009
Location: Brussels
Default VAS Current limiter

Hi Bob,

I incorporated the classic VAS Iq limiter. The base-emitter of the protection Q stands across the VAS Q (not the cascode Q but the amplifying Q). The collector of the protection Q goes back to the base of the beta enhancer Q. Nothing special here.

However I would like to have a led lighting up when the circuit goes into protection. How can this be done? It should not interfere with the signal nor with the protection function itself and it should be quite binary, I mean led on when protection is on and vice versa ... Reducing in betweens as much as possible ...

cheers,
Olivier
  Reply With Quote
Old 9th February 2011, 12:31 AM   #1475
godfrey is offline godfrey  South Africa
diyAudio Member
 
Join Date: Nov 2009
Location: Cape Town
Quote:
Originally Posted by Bob Cordell View Post
The stability of power amplifiers can be rather different in different operating regions...

I can't tell you how many times I've seen amplifiers break out into a burst of oscillation on the falling slope at high amplitude of a sinewave just under clipping...
Hi Bob
You gave a perfect example of this back here:
Quote:
Originally Posted by Bob Cordell View Post
I've done some more simulations on CMC, TPC and TMC and have attached a zipped folder of three simulations ...

It is easy to see that both TPC and TMC do not recover from clipping as well as CMC. However, TPC is much worse than TMC...
I was surprised by how bad the ringing was with TPC and decided to do a little investigation. My starting assumption was that reduced phase margin will increase the tendency to ring, but that you still have to kick a thing to make it ring. E.g. Rail sticking during clipping will give the circuit a nice kick when it "unsticks", and could initiate nasty ringing.

So my first step was to modify the VAS to get inherently clean, fast, symmetrical clipping. This resulted in a big improvement, but there was still visible ringing, and it was still much worse on the positive half.

That made me curious about whether the phase margin was varying with output level, so I tried a quick test with a staircase waveform. The first run showed the ringing getting slightly worse at higher levels, but about the same on both sides. Increasing the output level a bit revealed the problem, though.

When the NPN output device is conducting more than about 5.5A, the amp is totally unstable. Testing with various loads and output levels seemed to confirm that. Adding an extra pair of output devices fixed the problem (although it would probably still have problems above 10A, I didn't bother testing further.)

This was quite an eye-opener for me. I have thought it a good idea to operate output devices "below the knee" for other reasons (e.g. distortion), but it never occurred to me before that it might be important for stability too. It also served as a gentle reminder that checking bode plots under quiescent conditions doesn't even begin to tell the whole story.

Cheers - Godfrey
Attached Images
File Type: gif 8r1.GIF (10.3 KB, 330 views)
File Type: gif 8r2.GIF (11.8 KB, 329 views)
File Type: gif clip.GIF (11.5 KB, 702 views)
  Reply With Quote
Old 9th February 2011, 08:18 AM   #1476
Macleod is offline Macleod  France
diyAudio Member
 
Join Date: Dec 2008
Location: Toulouse, south of France
By the way, is there a possibility to make ion LTSpice AC analysis at various level? It seems changing the AC voltage doesn't change anything while I may observe that stability is level dependant.
  Reply With Quote
Old 9th February 2011, 08:44 AM   #1477
diyAudio Member
 
Join Date: Jun 2005
Location: Zürich
AC analysis is by definition carried out at infinitesimal AC level. To vary operating point add DC voltage sources in the feedback path (shifts output voltage) or DC current sources from the output to ground (shifts output current).

Samuel
  Reply With Quote
Old 9th February 2011, 03:31 PM   #1478
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by Olivier73 View Post
Hi Bob,

I incorporated the classic VAS Iq limiter. The base-emitter of the protection Q stands across the VAS Q (not the cascode Q but the amplifying Q). The collector of the protection Q goes back to the base of the beta enhancer Q. Nothing special here.

However I would like to have a led lighting up when the circuit goes into protection. How can this be done? It should not interfere with the signal nor with the protection function itself and it should be quite binary, I mean led on when protection is on and vice versa ... Reducing in betweens as much as possible ...

cheers,
Olivier
Hi Olivier,

If I understand you correctly, you are talking in the context of a single-ended VAS that is loaded with a current source (correct me if I am wrong, and what I say might apply to certain complementary VAS stages anyway). If you are in te context of the single-ended VAS, the LED will only light when the amplifier is clipping in one direction (trying to turn the VAS Q on too hard).

The challenge in illuminating an LED when this classic VAS current limiting scheme is used is that it appears you have only one Vbe plus the VAS emitter resistor drop to work with when the circuit is protecting. One approach might be to use a matched differential pair of transistors for the protection transistor and make it part of a current duplicator, where the added transistor is arranged to conduct approximately the same current as the protection transistor. Maybe put a little bit of degeneration in each emitter of the pair, but no so much that it interferes with the protection action. Also, you might want each of the pair of transistors to have their base go the the VAS emitter through a 100 ohm resistor to make life safe for each of the pair of transistors and to prevent the added transistor from dragging down the base of the protection transistor if the added transistor saturates. Use the current flowing in the collector of the added transistor to light the LED. Just thinking out loud here.

Cheers,
Bob
  Reply With Quote
Old 9th February 2011, 05:10 PM   #1479
diyAudio Member
 
Join Date: Dec 2009
Location: Brussels
Default Led VAS protection

Hi Bob,

I am using a complementary IPS & VAS. On one side I think it's too much components and too much potential interference to have just an led indication but on the other hand, the used amplifier topology is very complex so it doesn't come to one or two more components... I am unsure what to do for now.
But I choose not to go that road for now, the led can come in later when the circuit proto works in reality.

As you know I have to tame the fighting VAS Iq. I use a CMCL for it which seems to work well in simulation and also in reality for some part (that is in my circuit). My major problem for now is that my phase lag grows faster than it should with rising frequency.

I will redesign my pcb because I hope it suffers PCB layout problems such as : parasitic capacitors. One that bothers me much is the node between the IPS output and VAS input (on both sides). All summed this track is very long because it connects many components : Q mirror, Q diff, IPS clamp diode, VAS input Q, RC filter to rail, CMCL Q, protection Q, VAS clamp diode, compensation circuit.
Not only the length of the track is an issue but its impedance value must probably play a role as well as the amount of connections it has...

Do you think my trouble could have its origin on the pcb? Or is its impact usualy soo small I should search elsewere....

Is there a particular point to keep in mind?

Cheers,

Olivier
  Reply With Quote
Old 10th February 2011, 08:07 AM   #1480
Macleod is offline Macleod  France
diyAudio Member
 
Join Date: Dec 2008
Location: Toulouse, south of France
Quote:
Originally Posted by Samuel Groner View Post
AC analysis is by definition carried out at infinitesimal AC level. To vary operating point add DC voltage sources in the feedback path (shifts output voltage) or DC current sources from the output to ground (shifts output current).

Samuel
Wow! It seems so obvious now you point it out!

thank you for the tips, I will now step a param for output voltage to check that!
  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
Another realization of Bob Cordell's THD Analyzer giulianodes Equipment & Tools 37 9th October 2013 02:48 AM
best audio amplifier book?? Bouvett Everything Else 30 13th August 2012 02:43 AM
Amplifier Design Book pixie Everything Else 27 11th June 2010 08:36 PM
Project 11.1 from Slone "High-Power Amplifier" Book Karl71 Solid State 46 6th October 2008 03:47 AM
book-audio power amplifier by Douglas Self mikee12345 Solid State 8 16th November 2003 01:16 AM


New To Site? Need Help?

All times are GMT. The time now is 10:15 AM.


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