space2000 said:
You can almost take any class AB design topology and increase bias current to the class A power level you want. If you want maximum power in class A you will still need bigh heatsinks even though push-pull class A requires half the current as single ended class A.
If you want power amp examples of maximum class A in push pull,. check for the "F4" and "F5" power amps on the Pass forum.
On my side, I use a compromise, a 70W Amp with about 5 to 20 W in class A (8ohms) since I usually listen to no louder than that. The Amp operates in class AB for the remaining watts until 70W when you want your neighbords to join your party...
Good luck
Sorry to butt in
I've read various info on this. Apparently increasing bias raises distortion and listening fatigue, and how do you know if you're in Class A anyway? An Led indicator type circuit would be nice.
http://www.tnt-audio.com/clinica/bias_e.html
TNT Audio discusses doing it. Or am I barking up the wrong tree???
Mike
I've read various info on this. Apparently increasing bias raises distortion and listening fatigue, and how do you know if you're in Class A anyway? An Led indicator type circuit would be nice.
http://www.tnt-audio.com/clinica/bias_e.html
TNT Audio discusses doing it. Or am I barking up the wrong tree???
Mike
Hi,
Doubling the bias current in a class aB amplifier is pointless
and will cause more distortion than the optimum setting.
The tntaudio article is very poor and ignorantly generic.
There will be an optimum setting for bias which depends on
the topology of the output stage, any other values are wrong.
Cranking it up "to be more class A" is very mostly wrong.
/sreten.
Doubling the bias current in a class aB amplifier is pointless
and will cause more distortion than the optimum setting.
The tntaudio article is very poor and ignorantly generic.
There will be an optimum setting for bias which depends on
the topology of the output stage, any other values are wrong.
Cranking it up "to be more class A" is very mostly wrong.
/sreten.
Hi,
but, if you increase the ClassAB bias to give an effective half power while still in ClassA that transfers to ClassB for the final 50% of max power, then an amp that has 20dB of overhead will have 17dB of ClassA overhead and an additional 3dB of ClassB overhead.
When you listen to this amp you will as likely as not remain in ClassA for 99.9999% of the time and only use the last 3dB of max power during the very few occasions that have extreme transients.
It is quite likely that the extreme transient that needs this last bit of power is so loud and short lived that the extra distortion due to transferring to ClassB from the High Bias ClassA will go unnoticed.
Effectively the high bias has hidden the non optimum ClassAB bias by ensuring that most signals never require the amp to go to ClassB.
but, if you increase the ClassAB bias to give an effective half power while still in ClassA that transfers to ClassB for the final 50% of max power, then an amp that has 20dB of overhead will have 17dB of ClassA overhead and an additional 3dB of ClassB overhead.
When you listen to this amp you will as likely as not remain in ClassA for 99.9999% of the time and only use the last 3dB of max power during the very few occasions that have extreme transients.
It is quite likely that the extreme transient that needs this last bit of power is so loud and short lived that the extra distortion due to transferring to ClassB from the High Bias ClassA will go unnoticed.
Effectively the high bias has hidden the non optimum ClassAB bias by ensuring that most signals never require the amp to go to ClassB.
It's swings and roundabouts. Upping the quiescent current ( if the heatsinks and thermal management of the amp can cope long term ) will increase the ripple on the rails. Many amps have poor supply rejection for various reasons and compromised wiring layouts. You may find the increase actually worsens the noise and distortion rather than improve it at lower levels. And it's at low levels where it matters most IMO.
high bias in push pull
I wonder what Nelson Pass would say about that....
Maybe he should reconsider his F4, F5 and X amps series based on some of you...
I made the experiment on low or no overall feedback design and increasing bias lowered the THD . Unless someone prefers to rely on high NFB as a cure all...
And I am not talking about the overall sound improvement due to class A...
Maybe you should read this article from Nelson "Leaving class A" http://www.diyaudio.com/forums/showthread.php?postid=1553062#post1553062 and reports your comments...
Or am I missing something here?
I wonder what Nelson Pass would say about that....
Maybe he should reconsider his F4, F5 and X amps series based on some of you...
I made the experiment on low or no overall feedback design and increasing bias lowered the THD . Unless someone prefers to rely on high NFB as a cure all...
And I am not talking about the overall sound improvement due to class A...
Maybe you should read this article from Nelson "Leaving class A" http://www.diyaudio.com/forums/showthread.php?postid=1553062#post1553062 and reports your comments...
Or am I missing something here?
I am sure Nelsons designs do "what they say on the tin" and don't suffer from compromised layout, wiring and design flaws. Unfortunately many commercial products do to say nothing of D.I.Y. projects, and upping the the current won't neccessarily improve matters, it can make things worse in some cases.
to say nothing of D.I.Y. projects, and upping the the current won't neccessarily improve matters, it can make things worse in some cases.
Just got a Marantz PM7200. It has a class A switch for quick comparison with class B mode. The difference between classA and classB is too little to be noticeble IMO, probably because the high NFB already reduces much of the distortion even in class B mode. So, increases the biasing doesn't always gurantee better sound. The overall design of the amp itself is more important.
Hi,
KSA50 is a push pull ClassA amplifier, 50W into 8r0.
It will automatically go into ClassB if you demand more than ClassA current from it, eg. 100W into 4r0.
The normal ClassA bias is 1.9A. The maximum ClassA output current is~3.8Apk.
The 2pair MJ15003/4 output stage can survive 2ohm loading, putting out peak currents of the order of 20Apk, way in excess of the ClassA maximum current. The transient peak current capability is probably of the order of 40Apk.
KSA50 is a push pull ClassA amplifier, 50W into 8r0.
It will automatically go into ClassB if you demand more than ClassA current from it, eg. 100W into 4r0.
The normal ClassA bias is 1.9A. The maximum ClassA output current is~3.8Apk.
The 2pair MJ15003/4 output stage can survive 2ohm loading, putting out peak currents of the order of 20Apk, way in excess of the ClassA maximum current. The transient peak current capability is probably of the order of 40Apk.
Ipanema said:Just got a Marantz PM7200. It has a class A switch for quick comparison with class B mode. The difference between classA and classB is too little to be noticeble IMO, probably because the high NFB already reduces much of the distortion even in class B mode. So, increases the biasing doesn't always gurantee better sound. The overall design of the amp itself is more important.
I can hardly understand how a bias switch can do a good job here. The thermal balance of the same whole amp will be different between 2 bias settings. This can be one reason (as many others like the conditions of the sound test done) why it does not sound different or better. It may even sound worse at high bias if thermal balance is modified. On my side, choosing the bias is done at the early stage of the design...
One can not simply take any completed amp and increase the bias with eyes closed. The conditions when the amp goes from class A to B must be carefully analysed. But, as long as the sound stays in class A power level then there is no output transistors switching artefacts thus there should be one problem to be put aside...
Re: Sorry to butt in
I have one installed in one of my amp. Its my own design and it may not be perfect (depending on dynamic variation of DC offset at output and filtration of the very low frequency) but it gives a good idea.
See http://www.diyaudio.com/forums/showthread.php?postid=703051#post703051
It should work in follower power amp output stage but not verified if other topology would...
mikee55 said:...how do you know if you're in Class A anyway? An Led indicator type circuit would be nice.
Mike
I have one installed in one of my amp. Its my own design and it may not be perfect (depending on dynamic variation of DC offset at output and filtration of the very low frequency) but it gives a good idea.
See http://www.diyaudio.com/forums/showthread.php?postid=703051#post703051
It should work in follower power amp output stage but not verified if other topology would...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.