Audio Power Amplifier Design book- Douglas Self wants your opinions - Page 75 - 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 22nd April 2013, 11:10 AM   #741
dadod is offline dadod  Croatia
diyAudio Member
 
dadod's Avatar
 
Join Date: Apr 2006
Location: Zagreb
This blasphemy, probably, but I was not satisfied with the sound of the JLH 80W MOSFET amp, something was wrong. That was long time ago, and when finally I've got an oscilloscope I noticed small oscilations sometimes, and I did not like to high gain and to high value resistors used in the NFB, so I modified it.( JLH 80w mosfet power amplifier - modifying it ).
It is not anythig close to MIC anymore, but I like it better now. I used different small MOSFET as I don't have VN1210 model, used BSS123, so it could be some differnces there.
LG picture, green original JLH, blue modified.
Damir
Attached Images
File Type: jpg JLH80W-mod.jpg (117.7 KB, 146 views)
File Type: jpg LG-comparison-JLH_JLH-mod.jpg (235.6 KB, 140 views)
  Reply With Quote
Old 22nd April 2013, 11:32 AM   #742
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by steveh49 View Post
Ed Cherry also in 1982, although briefly (what I've quoted in post 723 is all of it AFAIR).
Hi Steve,

Thanks for this additional reference to MIC. Ed Cherry was prolific in his writing at the time and I am not surprized if he touched on what we are now calling MIC.

Cheers,
Bob
  Reply With Quote
Old 22nd April 2013, 11:43 AM   #743
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by kgrlee View Post
This has always been JLH's favourite method of compensation.

I've just checked that he doesn't compensate his 1969 Class A amp at all. But he certainly discusses it in his 1975 HiFi News 75W amplifier series.
_______________

Bob I think it may be worth discussing the need for zillion V/us slew. A 100W@8R amp gives V = 40 exp(jwt) Volts at max power.

The slew rate for this is dV/dt = j w 40 exp(jwt)

If the amp has a first order roll-off at f3dB, then the maximum slew rate the amplifier needs is twice that for a rail to rail sinewave at f3dB. You get this for an unclipped square wave whose rise time will reflect the bandwidth of the amplifier.

Any other roll-off will require less slew rate until when you have a brickwall filter (eg Red Book CD) the requirement is the same as for a a rail to rail sinewave at f3dB.

This means that you'd only see zillion V/us slew when the amp is severely overloaded. Beyond a certainly level, the slew only reflects the linear bandwidth of the amp.

Overload recovery and the lack of nasty artifacts are usually more important than high slew when coming out of overload.

In fact, one could argue that slew rate should be deliberately limited to what's required for low THD on a rail to rail sinewave at 2 x f3dB. The harmonics on overload would be more 'benign' cos the result would be a truncated triangle wave instead of a square wave.
_____________________

On much surer psychoacoustic ground is deliberate bandwidth limiting. I don't think there is any need for a power amp to have a 3dB bandwidth exceeding 100kHz. Some parties, eg the Broadcast Organisations, in fact insist on much lower bandwidths.

ALL the reliable blind listening tests on band limiting of electronic signals, both this and the previous millenium, show a clear preference for the band limited signal from those who can reliably tell the difference.

Our 100W 8R, 100kHz amplifier only needs a slew rate of

2 x 40 x 2 x pi x 100,000 = 50.26 V/us

This is halved if its 3dB bandwidth is 50kHz and reduced by nearly 10x if it is only playing Red Book CD.
Hi kgrle,

You are right about the matter of how much slew rate is really needed in an amplifier. A long long time ago I wrote an article in Audio called "Another View of TIM" (I think I have it on my web page at CordellAudio.com - Home). In that article I touched on the slew rates of various sources and followed much of the logic that you presented here. At the time TIM was the rage, and I got hammered for putting the need for slew rate into numerical context.

It is certainly true that most sources don't produce great amounts of slew rate. For example, a 20kHz sinewave produces about 0.125 V/us per volt peak of the signal. However, Amplifiers with higher slew rate usually have a lot more margin against HF nonlinearity, so in comes the valid matter of what was then called soft TIM. Bottom line is that I usually like to see an amplifier have at least 10X the slew rate needed for it to do full power on a 20kHz sinusoid.

What really counts is HF linearity in the open and closed loop.

I agree, the 300V/us of my 1983 50-watt amplifier was way more than needed, but achieving that high slew rate was not a primary goal of the design. It just turned out that way because of my use of Miller Input Compensation.

Cheers,
Bob
  Reply With Quote
Old 22nd April 2013, 01:44 PM   #744
diyAudio Member
 
Join Date: Feb 2003
Location: Australia
Quote:
Originally Posted by michaelkiwanuka View Post
If I remember corrrectly, J. L. Hood used 220p MIC on his MOSFET designs. I am not certain he degenerated his input stage though.
I've just checked the few JLH articles I have. The MIC components were:
- 1982: 5pF + 47k; cf 33k feedback resistor
- 1989: 10pF; cf 56k feedback resistor
- 1993: 5pF + 120k; cf 39k feedback resistor

Perhaps the 220pF was only very early on in his writings, or you have misremembered?
  Reply With Quote
Old 22nd April 2013, 03:39 PM   #745
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally Posted by kgrlee View Post
This has always been JLH's favourite method of compensation.................
I have regularly noticed that this is a favorite of most of the 70's into 80's British designers.
Would I be correct to surmise they were talking to each other?
__________________
regards Andrew T.
  Reply With Quote
Old 22nd April 2013, 03:40 PM   #746
Waly is offline Waly  United Kingdom
diyAudio Member
 
Waly's Avatar
 
Join Date: Jan 2011
Location: London
Quote:
Originally Posted by Bob Cordell View Post
It is certainly true that most sources don't produce great amounts of slew rate. For example, a 20kHz sinewave produces about 0.125 V/us per volt peak of the signal. However, Amplifiers with higher slew rate usually have a lot more margin against HF nonlinearity, so in comes the valid matter of what was then called soft TIM. Bottom line is that I usually like to see an amplifier have at least 10X the slew rate needed for it to do full power on a 20kHz sinusoid.
That would be the small signal slew rate, which is tied to the frequency corner where the closed-loop gain meets the amplifier open-loop gain curve. The small signal slew rate is usually discussed in terms of "rise time" and the well known expression Tr=0.35*Acl/Fu where Acl is the closed loop gain and Fu is the amplifier ULGF.

This being said, your requirement of a 10x the slew rate to do the full power on a 20KHz sinusoid is in fact a requirement for the amp ULGF: SR=Vout/Tr=Vout*Fu/(0.35*Acl)=Vin*Fu/0.35 => Fu=0.35*SR/Vin. Which, in all truth, I don't think it makes much sense, and here is why:

The full power on a Fo=20KHz sinusoid is achieved at a slew rate of 2*PI*Vout-peak*F=2*PI*SQRT(2)*Vout*F. If you want the amp to have a k times more slew rate than at Fo, then this maps to (see above, 0.35 is 2.19/(2*PI) where 2.19 comes from the 10-90% SR limits) k*2*PI*SQRT(2)*Vout*Fo=Vout*Fu/(0.35*Acl) => k*SQRT(2)*Fo=Fu*Acl/2.19. This is in fact a condition for the amp ULGF, Fu=k*SQRT(2)*2.19*Acl*Fo. Putting in some numbers, k=10 as you said, Acl=26dB (x20), Fo=20KHz we get Fu=12.35MHz. You would of course agree that this is way larger than practically achievable in an audio amp.

Now, maybe you are talking about a constraint for the large signal slew rate, as an effect of the input stage nonlinearities? That would be a different story, the large signal slew rate (what is actually called "slew rate" in most data sheets) has little to nothing to do with the small signal slew rate. I would fully agree that a large signal slew rate of 10x the slew rate required to do the full power bandwidth at 20KHz makes sense, and that would be (again see above) k*Fu/(0.35*Acl) for each volt at the output, which leads immediately to an amp SR requirements (for the same Acl=26dB, k=10 and taking Fu=1MHz) of 1.42V/uS for each volt at the output. This is a good and perfectly achievable value, significantly larger than the old rule of thumb of 0.5V/uS for each output volt.
  Reply With Quote
Old 22nd April 2013, 03:49 PM   #747
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally Posted by steveh49 View Post
I've just checked the few JLH articles I have. The MIC components were:
- 1982: 5pF + 47k; cf 33k feedback resistor
- 1989: 10pF; cf 56k feedback resistor
- 1993: 5pF + 120k; cf 39k feedback resistor

Perhaps the 220pF was only very early on in his writings, or you have misremembered?
In at least one article he suggested the resistor be "tuned" for best squarewave shape, using a pot wired as rheostat.
This implied that different builds would require a different R even though all were supposed to be using the same PCB and the same components.
It seems clear that he recognised that the stability margin does vary with device parameters.
__________________
regards Andrew T.
  Reply With Quote
Old 22nd April 2013, 05:29 PM   #748
wahab is offline wahab  Algeria
diyAudio Member
 
Join Date: Nov 2009
Location: algeria/france
Other than increasing slew rate MIC has no practical value
linearity wise , a moderate TMC yield lower distorsion even
at the upper side of the audio spectrum.
  Reply With Quote
Old 22nd April 2013, 05:49 PM   #749
dadod is offline dadod  Croatia
diyAudio Member
 
dadod's Avatar
 
Join Date: Apr 2006
Location: Zagreb
Quote:
Originally Posted by wahab View Post
Other than increasing slew rate MIC has no practical value
linearity wise , a moderate TMC yield lower distorsion even
at the upper side of the audio spectrum.
TMC could be used as MIC(TMIC ?), but again if to get same phase margin and gain margin it should be internaly compensated and we are back to similar distortion as ordinary TMC.
  Reply With Quote
Old 22nd April 2013, 07:28 PM   #750
diyAudio Member
 
Join Date: May 2011
Location: Silicon Valley
I'm sure y'all are aware of the A40 (40 watts class A) amplifier that Nelson Pass published in Audio Amateur (1978). Here is the full article and the circuit schematic is below. Please note frequency compensation capacitor C4.
Attached Images
File Type: jpg pass_a40.jpg (147.0 KB, 108 views)
  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
The design of active crossovers- Douglas Self wants your opinions DouglasSelf Analog Line Level 283 12th August 2014 12:01 PM
Comments on Audio Power Amplifier Design Handbook by Douglas Self Samuel Groner Solid State 68 16th March 2013 11:26 PM
Douglas Self's Active Crossovers book forr Analog Line Level 8 31st December 2010 06:20 PM
New Douglas Self Book blmn Solid State 7 22nd July 2009 07:00 PM
book-audio power amplifier by Douglas Self mikee12345 Solid State 8 16th November 2003 02:16 AM


New To Site? Need Help?

All times are GMT. The time now is 09:42 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