diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   Class A biasing for the AD797? (http://www.diyaudio.com/forums/solid-state/123613-class-biasing-ad797.html)

Onvinyl 23rd May 2008 09:38 PM

Class A biasing for the AD797?
 
Hi!
It has probably been discussed before, but if so, I couldn't find it.
Is it possible to bias the output stage in class A by referencing a current to the output node? If so, V+ or V-? Is it worth it sonically? Is the output stage class AB, and if so, which is the class A area currentwise, so a (very) light load resistor might work instead?
thank you,
Rüdiger

analog_sa 23rd May 2008 09:58 PM

Re: Class A biasing for the AD797?
 
Quote:

Originally posted by Onvinyl
Hi!
Is the output stage class AB,


Quiescent current 8.2mA and +/-13v in 600ohm. Does it sound like class A? :)

What load do you have in mind?

Onvinyl 23rd May 2008 10:16 PM

Hi,
with light load I meant the use of an external buffer stage. Say 50k.
Rüdiger

aparatusonitus 23rd May 2008 10:16 PM

Re: Class A biasing for the AD797?
 
Quote:

Originally posted by Onvinyl
Hi!
It has probably been discussed before, but if so, I couldn't find it.
Is it possible to bias the output stage in class A by referencing a current to the output node? If so, V+ or V-? Is it worth it sonically? Is the output stage class AB, and if so, which is the class A area currentwise, so a (very) light load resistor might work instead?
thank you,
Rüdiger

I don't know is it possible to bias AD797 in Class A I have never tried that mod, but if the answer is yes, you should always reference to -V, because you want to NPN output transistor to do the hard work, not PNP;)

Is it worth sonically?...Yes and no, you should try that for yourself, but be aware that the THD numbers will be up, but at the same time H2 will be up too, which can be pleasing to some ears.

Regarding standing current in AD797 output transistors, according to my simulation, it's about 0,5mA.

jcx 23rd May 2008 10:39 PM

the datasheet claims 500uA output stage bias

push-pull operation operation means ~+/-1 mA load current would be delivered in Class A without additional bias, which turns out to be not much when using feedback impedance levels low enough to take advantage of the input noise performance

the AD797 has a unique local feedback option that essentially cancels the dominant pole compensation around the output stage local Cn loop, giving high loop gain for correction of output stage error:
http://www.diyaudio.com/forums/showt...28#post1513928

with the optimum Cn local feedback cap I doubt external Class A bias would give a measureable improvement

a possible difference between Class A and AB is that poor power supply routing could allow Class B rail currents to magnetically couple to low level signal loops - Self's "distortion #6:
http://www.dself.dsl.pipex.com/ampins/dipa/dipa.htm#2

I would suggest adding a buffer/gain stage with a CFA op amp in the feedback loop as shown in the datasheet fig 49 if ultimate low distortion performance was your goal

scott wurcer 23rd May 2008 10:47 PM

Quote:

Originally posted by jcx
the datasheet claims 500uA output stage bias

push-pull operation operation means ~+/-1 mA load current would be delivered in Class A without additional bias, which turns out to be not much when using feedback impedance levels low enough to take advantage of the input noise performance

the AD797 has a unique local feedback option that essentially cancels the dominant pole compensation around the output stage local Cn loop, giving high loop gain for correction of output stage error:
http://www.diyaudio.com/forums/showt...28#post1513928

with the optimum Cn local feedback cap I doubt external Class A bias would give a measureable improvement

a possible difference between Class A and AB is that poor power supply routing could allow Class B rail currents to magnetically couple to low level signal loops - Self's "distortion #6:
http://www.dself.dsl.pipex.com/ampins/dipa/dipa.htm#2

I would suggest adding a buffer/gain stage with a CFA op amp in the feedback loop as shown in the datasheet fig 49 if ultimate low distortion performance was your goal


This is correct, you rapidly get into touchy bypassing issues. You really have to think about where the harmonic energy in the supply current is going.

nelsonvandal 24th May 2008 07:18 AM

Is it possible to tap the signal from the decompensation pin 8 (point B in the simplified schematic) and connect it to a discrete output stage and wrap it in the feedback loop instead of the built in diamond buffer, just like you can from the compensation pin in AD829 and AD744?

lumanauw 25th May 2008 06:04 PM

Hi, JCX,

Quote:

with the optimum Cn local feedback cap I doubt external Class A bias would give a measureable improvement
Is this applicable to power amp too? To make a better classB/AB output stage?

lumanauw 25th May 2008 06:07 PM

Quote:

You really have to think about where the harmonic energy in the supply current is going.
To the output node via Cbc?

P.Lacombe 25th May 2008 07:28 PM

One good solution is to connect a constant current sink at the output. Very good results are possible with bipolar sink, in the range 8...12 mA. Fets exhibits poor linearity at low voltage, and thus are not usable.

PL.


All times are GMT. The time now is 02:13 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