Glad someone is paying attention 😀
Theoretical maximum efficiency of a class-A amplifier is 50%, so 46% is certainly possible, however 53% should be taken with a grain of salt. Higher efficiency is only possible with deliberate use of non-linearities (such as in square-law output stages).
thanks guys. good stuff. i like pma's opa627 front end idea. i've always seen the opa627/op637 as a VERY nice (albeit very expensive) part that generally has never really gotten the respect it deserved.
mlloyd1
mlloyd1
No, I mean class AB that drives center point of the power supply that supplies small class A amplifier with full load current capability. This saves a lot of heat, though it is quite complex. It was Technics in seventies/eighties who came with the idea.
You will find a thread even here
http://www.diyaudio.com/forums/solid-state/104259-my-1024w-rms-class-amplifier.html
You will find a thread even here
http://www.diyaudio.com/forums/solid-state/104259-my-1024w-rms-class-amplifier.html
I think this is what you're referring to?
Looks overly complex to me, just to save a few Watts?
An externally hosted image should be here but it was not working when we last tested it.
Looks overly complex to me, just to save a few Watts?
No. I think there is a link to a patent in the thread I gave a link to.
In case you wish to have 200 - 300W true class A you have no other chance, re heat dissipation.
In case you wish to have 200 - 300W true class A you have no other chance, re heat dissipation.
20V amplitude of 20kHz sine makes 2.513V/us at zero crossing. You need 5 - 10x more to get reasonably low 19+20kHz CCIF.
Bridge needs 2x higher slew rate.
If you would like to learn the formula, it is SR = max.(dv/dt) = 2 x pi x F x Vp, Vp is amplitude of the sine wave.
You also need to handle possible EMI spikes from D/A converters etc.
Bridge needs 2x higher slew rate.
If you would like to learn the formula, it is SR = max.(dv/dt) = 2 x pi x F x Vp, Vp is amplitude of the sine wave.
You also need to handle possible EMI spikes from D/A converters etc.
Last edited:
And if you design the amplifier down to that minimum slew rate you find that distortion increases enormously well before the limiting slew rate is reached.
Many have concluded that the required slew rate should be around 5times h#what the signal slew rate is and some have gone straight to 10times as adequate slew capability to cater for maximum slew in the signal.
Now back to the nonsense statement
edit,
I see PMA is stating the same 5 to 10times requirement.
Many have concluded that the required slew rate should be around 5times h#what the signal slew rate is and some have gone straight to 10times as adequate slew capability to cater for maximum slew in the signal.
Now back to the nonsense statement
edit,
I see PMA is stating the same 5 to 10times requirement.
For my amusement I calculated the output errors for ExtremeAB, that is, class AB output at 1/2 max output current bias. At this level there is no crossover at all and only a small amount of compressive 3rd's. With .28 Ohm emitter resistors the open-loop output resistance modulates only .14 Ohm to .165 Ohm, of course with the feedback it is constant .14 Ohm. Open-loop distortion at full 100W is .024% thirds. At only 48dB gain at 20K all distortion above PPM levels will be from the rest of the circuit and the build. The AB does not hit the wall at 5A.
I agree with PMA that there are probably many front ends that work including an op-amp.
I agree with PMA that there are probably many front ends that work including an op-amp.
I suppose you mean that when Iout =< 2 x Iq, then there is almost no distortion - I agree. But when Iout > 2Iq, distortion rises rapidly.
- Status
- Not open for further replies.