I've played with a value of R16 to adjust Aleph current source and found that with its value 1k I can get the distortion 0.02% with 8Ohm load at 1kHz and 1W output. My bias current is set at 1.8A. This value of R16 seems too low for me for 1.8A bias or it's OK?. I'm using audiophile sound card Asus Xonar Essence to measure distortions which gives 0.0003% THD with self closed loop.
Any comments?
Any comments?
Hmm, such low R16 value has a side-effect on higher power level. The amplifier is getting unstable for example at 4Ohm 10W - the distortion gets much higher (~3%) with third harmonic dominating, looks like the amplifier gets converted into push-pull one.
You have figured it out correctly. As you raise the current
gain of the bias source, the distortion of the circuit goes
down at lower power, but it will go into Class B at a lower
power level. You might want to take a look at Wen-San's
SRPP thread for more information and solutions.
😎
gain of the bias source, the distortion of the circuit goes
down at lower power, but it will go into Class B at a lower
power level. You might want to take a look at Wen-San's
SRPP thread for more information and solutions.
😎
Thank you, Mr. Pass
Did you mean "it will go into Class B at a higher
power level." instead of "it will go into Class B at a lower
power level."?
Did you mean "it will go into Class B at a higher
power level." instead of "it will go into Class B at a lower
power level."?
It's like this:
Greatest Class A efficiency occurs when the top current
source is supplying about 50% of the AC output current,
this being true of all aleph, mu follower, sepp and srpp
circuits (all of which are simply variations of each other).
If you set the gain of that current source to deliver less,
then you must increase the bias to avoid clipping. If you
set it for more, you must increase the bias to avoid
dropping out of Class A.
In the extreme case of a constant current source, the
maximum undistorted power is at output = bias, which
is 1/4 the potential power of a 50% arrangement.
In the case of very high gain on the current source, it
goes into class B (or AB if you like) at output = bias,
which again is 1/4 power.
And so it would be lower.
😎
Greatest Class A efficiency occurs when the top current
source is supplying about 50% of the AC output current,
this being true of all aleph, mu follower, sepp and srpp
circuits (all of which are simply variations of each other).
If you set the gain of that current source to deliver less,
then you must increase the bias to avoid clipping. If you
set it for more, you must increase the bias to avoid
dropping out of Class A.
In the extreme case of a constant current source, the
maximum undistorted power is at output = bias, which
is 1/4 the potential power of a 50% arrangement.
In the case of very high gain on the current source, it
goes into class B (or AB if you like) at output = bias,
which again is 1/4 power.
And so it would be lower.
😎
- Status
- Not open for further replies.