Graham's Class A/JLH output - Page 11 - 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 11th March 2005, 11:53 AM   #101
diyAudio Member
 
BrianDonegan's Avatar
 
Join Date: Dec 2004
Location: TPA HQ
Many thanks Graham. I'm going to build one to drive my Klipsch's. I'll report back.
__________________
Twisted Pear Audio
  Reply With Quote
Old 19th March 2005, 01:59 AM   #102
diyAudio Moderator Emeritus
 
lumanauw's Avatar
 
Join Date: Oct 2002
Location: Bandung
Send a message via Yahoo to lumanauw
I've got a question. Why is in GM's type or JLH type output stage or non compliment type (like I attached here), they never use RE as emitor degeneration? Is RE forbidden in this type of output stage?

How if we wanted to parrarel the output stages, wouldn't RE is a must?
Attached Images
File Type: jpg noncompliment.jpg (25.2 KB, 1563 views)
  Reply With Quote
Old 19th March 2005, 05:12 AM   #103
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
if you want to add emitter resistors to balance the current in parallel output devices I don't see any difficulty - you would only be doing at the packaged transistor level what is implemented internally in many power transistor constructions

the real issue is whether you really want more power in a Class A amp; to accommodate impedance dips you might want 2x the nominal current bias giving 100 W continuous dissipation for 25 W rated output - 1 KW for 250W rms? - here it could be a "winter amp" - maybe you could run a Laundromat on the waste heat if you go 5.1 (looks like that should be a night-time amp with Bandung’s climate, maybe you could drive an absorption cooler during the day?)
  Reply With Quote
Old 19th March 2005, 05:44 AM   #104
diyAudio Moderator Emeritus
 
lumanauw's Avatar
 
Join Date: Oct 2002
Location: Bandung
Send a message via Yahoo to lumanauw
Why is that I get impression that many members here have been to Bandung?

JCX,

The thinking is simple. It's not about more power, but about safety. I got worries with hot transistors. They can act weirdly. 50W dissipation will be safer to be handled by 2 or 3 pair than 1 pair, right?

If you see the non-compliment article, the biasing method is quite different than GM's. GM is using bootstrap to upper rail, and non-compliment using fixed voltage referenced to output node.

Is these 2 resulting the same thing?
  Reply With Quote
Old 19th March 2005, 08:42 AM   #105
diyAudio Member
 
Graham Maynard's Avatar
 
Join Date: Dec 2003
Location: UK
Hi Lumenauw.

You have attached the 'non-complementary' circuit due to Bengt Ollson, I think circa 1994, though there seems to be a drawing error for the pcb/heatsink split. It was class-AB.

Bengt went on to manufacture 250W.rms 8 ohms amplifiers using single push-pull pairs of output Mosfets with no emitter resistors. His solution regarding temperature compensation was in the little bootstrapped sub-circuit shown boxed beside TR1.

Two reasons for me to not like this circuit at the time were that it effectively had Miller compensation at the splitter driver, also it used a series output choke.

Bengt actually removed the choke in his next version, but increased Miller C with an additional component such that the phenomenal open loop gain started rolling off around 100Hz. Again I could not like that.

Bipolars need current bias; through base wrt emitter - in a JLH via the collector resistors and bootstrapped to the upper emitter.
Mosfets need voltage bias; to gate wrt source - Bengt also uses Vcc current flow, but voltage regulates and similarly bootstraps it to upper source.

Both lower output halves are already driven wrt to their emitter/source terminations, and the interposing of an emitter/source resistor here also increases the imbalance of the splitter drive wrt input stage output, though there is less effect upon a differential/mirror stage output.
__________________________________________________ ______________

With correctly biased single output devices, emitter/source resistors actually degrade output stage performance, though they are essential for the correct hot running of parallel output devices.
Low value resistors do not degrade parallel performance however, due to the combined device decrease in effective emitter/source resistance.
__________________________________________________ _______________

JCX is of course correct relating to heat versus class-A power output, but modern devices are rugged, and your concerns can even more allayed by implementing a low drive resistance capable of quickly removing base/grid stored charges.


Cheers ............. Graham.
  Reply With Quote
Old 23rd March 2005, 05:24 PM   #106
mikelm is offline mikelm  England
diyAudio Member
 
Join Date: May 2003
Location: England
Send a message via Yahoo to mikelm
Default A question for Graham

In an amplifier with global feedback but with little or no compensation - the kind that you prefer

What would you think of as being an acceptably low propigation time

or to ask the same thing in a different way at what kind of frequency would you hope to see 1 degree of phase shift ?

I was thinking that 15 - 25 nano seconds was pretty good

What say you ?

mike
  Reply With Quote
Old 24th March 2005, 06:32 AM   #107
diyAudio Member
 
Graham Maynard's Avatar
 
Join Date: Dec 2003
Location: UK
Hi Mike,

What a question !

The propagation time is more a function of the circuit, than what I might be looking for.

I look out for the way an amplifier reacts when simulated with a virtual loudspeaker, ie. I null the output at the virtual loudspeaker and study the error waveform and node currents with phase shifted crossovers. Thus the voltage output can be delayed wrt input, but the zero current crossover can arise several volts before the end of a first sine cycle, and non-class-A amplifiers always develop voltage crossover wiggles about that zero current crossover.

So it is not just propagation delay, but whether fully stable back emf induced error correction is completed. Any amplifier that can do all of this in say 50nS wrt first input base/gate is going to be excellent..


Cheers ............. Graham.
  Reply With Quote
Old 25th October 2005, 11:27 AM   #108
diyAudio Member
 
Join Date: Jan 2002
Location: Moscow
Hi Graham!
Why you did not use current source instead resistor in emitters of input transistors ?
  Reply With Quote
Old 26th July 2006, 07:41 AM   #109
diyAudio Member
 
Join Date: Dec 2000
Location: Önnarp, Söderslätt
Hi there.

Quote:
Originally posted by Graham Maynard
Hi Ejam,

My amplifier has a differential input stage which immediately makes it a different beastie to JLH designs.
The original JLH circuit is basically stable and as such did not need feedback compensation if sensibly constructed and used, but when modern high frequency transistors it might become necessary, with an output Zobel too. ( I use 47n and 2.2 ohms for mine where 100n and 10 ohms has been common.)
Cheers ....... Graham. [/B]
I'm in the middle of building the 25-8, and have a few question to the Zobel network:

1: Why choose 47nF + 2,2ohm instead of 100nF + 10ohm. Is it more "musical" due to the higher Fo??

2: Will it make any difference if placed at the speaker terminals instead of on the PCB??
__________________
Regards
Hoffmeyer ;0)
  Reply With Quote
Old 26th July 2006, 03:04 PM   #110
diyAudio Member
 
destroyer X's Avatar
 
Join Date: Feb 2004
Location: Recife - Brasil Northeast
Default I am afraid that Graham are not tracking our forum anymore.

He had some problems...some troubles inside our forum, someone that already disappeared from our forum had disturbed Graham`s mind and he gave up from our forum.

But i can remember that he said that this zobel filter values are something interesting to create a lower impedance patch into the earth, to avoid start of oscilations....as amplifier will be continuouly loaded related hi frequencies...... when starting oscilations the zobel will represent, in his design, something alike an enormous troat to drag those oscilations.

Depending of the source signal frequency, some oscilations could start, as he was using hi speed transistors and all them in the threshold of unstability, also Graham do not like Miller capacitors, and normally he used to avoid them, as those miller compensation capacitors normally turns sinusoidal input wave into triangle waves when frequency goes up into the spectrum...so, it can be a distorter too.

But as you know, Miller capacitor produces stability against oscilations too..... having not the Miller capacitor he decided that could be more safe make some modifications in the zobel filter values.

Graham continues to produce designs, but they are not published in our forum anymore, as he understood that bad educated guys can disturb, that no one will have protection, that we have only the forum rules that are not so strong to control those bad guys that sometimes appear in our forum...also that he is too much old to face those foolishes.... to see young boys, some of them without experience, some others with clearly bad intentions, to enter to discuss things already tested and that belongs to our Universal knowledge...things already accepted as true by almost half of our scientific world.

He is in some other forum...i heard that that forum collects people that do not accept this forum rules...well, i do not have the adress, as i accept almost all our forum rules...with some exceptions.

regards,

Carlos
__________________
Dx Super A, Hajame test, 35V, 4 Ohms, T-line speaker, flat https://www.youtube.com/watch?v=bYIZejeSoeQ
  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
Graham Maynard Stocker Everything Else 0 28th March 2007 03:47 PM
deleted Graham Maynard thread SY Solid State 0 16th October 2005 11:52 AM
First cycle distortion - Graham, what is that? peranders Solid State 115 29th April 2004 03:24 PM


New To Site? Need Help?

All times are GMT. The time now is 11:08 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