I'm looking for ideas for building my own solid state class A/B amplifier, which should be meet the following specifications:
- Class A/B design
- Capable of delivering 50W rms into 8 ohms.
- I don't care if it is based on MOSFETs or BJTs, although I do have a slight preference for MOSFETs. The main thing is the performance.
- Exceptional THD performance up to maximum power levels over the audio frequency range.
- I don't care how complicated it is since I'm a professional electronics engineer. But I believe things should be no more complicated than they need to be.
Last edited:
Care to put some figures on that? <0.001% ? <0.0001% ??
BTW Maximum power levels always involve increased distortion, you mean the rated power level?
Do you mean either class A or B, or class AB?
To be more precise, I'm after a class AB design with THD of less than 0.01% at rated power. For me the sonic qualities and ability to reproduce complex waveforms are more important than a fantastically low THD value when fed with a sine wave signal.
For that pure class A may be a better option but efficiency then becomes a problem. Best limit power to levels that are actually needed.
The other aspect is how to know that you have the actual spec numbers you want? Measure yourself maybe?
H
HAYK
Welcome to the world of audiophiles. Yes, THD numbers are meaningless for sound quality. What is important in feedback amplifiers, is the NFB/frequency. At 20khz a high end sound needs 70db or wide band open loop response. @LKA is the great master on this website in this domaine, have a look to his creations. https://www.diyaudio.com/community/search/931298/
I passed from class A to class D, I only have a composite chip amp in AB using 1.5$ chinese Shanxin TDA7494 with 70db @20khz NFB.
I passed from class A to class D, I only have a composite chip amp in AB using 1.5$ chinese Shanxin TDA7494 with 70db @20khz NFB.
Attachments
The wolverine and AMB beta 24 both come to mind as well thought out designs with boards readily available, excellent documentation and active forums to support them.
There are many other excellent designs on this forum, but sourcing boards and build notes may not be as good.
There are many other excellent designs on this forum, but sourcing boards and build notes may not be as good.
The less THD, the more true to the input waveform. There is no other way to see it.
Many audiophiles cant have both sonic qualities (subjective) and ability to reproduce complex waveforms (objective)... 🙂
//
High end (sound) and A/B, i.e. push pull, are mutually exclusive. But an example of a super PP that is very close:
https://www.diyaudio.com/community/threads/zero-feedback-impedance-amplifiers.42259/
Otherwise: as few amplifying stages as possible, half-wave symmetrical, one psu (voltage) for all.
My recommendations.
https://www.diyaudio.com/community/threads/zero-feedback-impedance-amplifiers.42259/
Otherwise: as few amplifying stages as possible, half-wave symmetrical, one psu (voltage) for all.
My recommendations.
Discussing what topology or design philosophy is best is very polarizing. It’s like asking what music type is best or which sport is best. Everyone has their opinions and what one person prefers may not appeal to someone else.
I’d suggest it’s more appropriate to answer the original question with designs that are popular and/or well received even if the approach doesn’t suit your tastes.
I’ll use Rod Elliotts designs as an example. The P03A and P101 are popular designs with boards available, build notes and excellent support. But they are a polarizing design. Some love the simplicity, but it’s not for everyone. Many builders will find it too simplistic. They also don’t measure well compared to something like the Wolverine, but there is a fan base for them that falls into the “they sound good even if they don’t measure well” camp.
Douglas Selfs blameless design (available for signal transfer company) measures well, but FirstWatt class A fans will likely say it doesn’t sound good.
At the end of the day, it’s your time and money. Build whatever appeals to you.
I’d suggest it’s more appropriate to answer the original question with designs that are popular and/or well received even if the approach doesn’t suit your tastes.
I’ll use Rod Elliotts designs as an example. The P03A and P101 are popular designs with boards available, build notes and excellent support. But they are a polarizing design. Some love the simplicity, but it’s not for everyone. Many builders will find it too simplistic. They also don’t measure well compared to something like the Wolverine, but there is a fan base for them that falls into the “they sound good even if they don’t measure well” camp.
Douglas Selfs blameless design (available for signal transfer company) measures well, but FirstWatt class A fans will likely say it doesn’t sound good.
At the end of the day, it’s your time and money. Build whatever appeals to you.
If you have access to PCB manufacture, maybe you are interested in this.
i consider stability to be important above all.
you may also wish to consider slew and clipping characteristics.
a list from another thread
https://www.diyaudio.com/community/threads/diy-class-ab-ss-amplifiers-top-list.305856/
https://www.diyaudio.com/community/threads/diy-class-ab-ss-amplifiers-top-list.305856/
Bob Cordell's BC-1. I just built it and it works great. PCBs here: ebay item number 225339175513
Read more in diy thread "Bob Cordell’s BC-1 audio power amp".
Read more in diy thread "Bob Cordell’s BC-1 audio power amp".
.01 be at 20 kHz high frequency and .001 would normal for 1 kHz
plenty of common amplifiers or design topology can do that.
Wolverine is a very good example and Bob Cordells BC-1
and most of Bobs published books or Magazine articles
will show you how to do so.
Far as getting 20 kHz below .01% has also been done by Bob
with feedback error compensation using mosfets.
which if I remember was around .0001% could be wrong
cause .001 at high frequency is rather amazing once you become
familiar with stability at high frequency with high current.
There is plenty of advance design in small signal opamps.
Topologies with incredible slew rate and distortion as low
as .00001% So ideas can be applied to high power amplifiers.
just maintaining stability at high frequency/ high current is the
challenge.
as said before there's plenty common
home receivers and common brands that can do .01%
Any differential circuit will do it. Usually involves common
current mirror for the differential first stage. then further
reduction of 2nd stage using beta enhancement. And those
design are incredible common. And the more known writers
explaining the topology is Bob Cordell and Douglas Self.
Of course the credit actually goes to the inventors.
And numerous engineers. Those are just common known
published writers for Diy
So again BC-1 just shows more advanced differential
circuits. Which spawns from 1960's advancement of the
opamp
plenty of common amplifiers or design topology can do that.
Wolverine is a very good example and Bob Cordells BC-1
and most of Bobs published books or Magazine articles
will show you how to do so.
Far as getting 20 kHz below .01% has also been done by Bob
with feedback error compensation using mosfets.
which if I remember was around .0001% could be wrong
cause .001 at high frequency is rather amazing once you become
familiar with stability at high frequency with high current.
There is plenty of advance design in small signal opamps.
Topologies with incredible slew rate and distortion as low
as .00001% So ideas can be applied to high power amplifiers.
just maintaining stability at high frequency/ high current is the
challenge.
as said before there's plenty common
home receivers and common brands that can do .01%
Any differential circuit will do it. Usually involves common
current mirror for the differential first stage. then further
reduction of 2nd stage using beta enhancement. And those
design are incredible common. And the more known writers
explaining the topology is Bob Cordell and Douglas Self.
Of course the credit actually goes to the inventors.
And numerous engineers. Those are just common known
published writers for Diy
So again BC-1 just shows more advanced differential
circuits. Which spawns from 1960's advancement of the
opamp
Marantz MA9-S2 has a thread here: https://www.diyaudio.com/community/threads/marantz-ma9s2-choke-input-power-supply.396823/ More than 50w though.
Why don't you design your own amplifier then? Then you can tailor it to whatever you want.
- Status
- Not open for further replies.