I`m inclined to say just play it as it is.
You see the bias problems are also due to the fact that Metaxas uses two seriescoupled 1N4001 diodes as control element. These are glued into epoxy, which is not normally a problem, but it is in this instance. Because you have so small heatsink area you can`t really get a stable temperature and as a result the bias will fluctuate.
The spread in bias values are due to differing data on the transistors.
A better way would be to utilize a Vbe transistor, as this can be properly bonded to the heatsink.
R
You see the bias problems are also due to the fact that Metaxas uses two seriescoupled 1N4001 diodes as control element. These are glued into epoxy, which is not normally a problem, but it is in this instance. Because you have so small heatsink area you can`t really get a stable temperature and as a result the bias will fluctuate.
The spread in bias values are due to differing data on the transistors.
A better way would be to utilize a Vbe transistor, as this can be properly bonded to the heatsink.
R
The bias is actually quite stable, I'm just not that happy with the wildly spread bias values (the original transistors actually seem to be more poorly matched than the on semi I replaced them with).
Both amps have some oscillation on bottom and top half (more so on the bottom part) of the sine wave when I get to about 3/4 of the max rated power. It's a little worse on the one I repaired but only very very slightly so I'm guessing this is normal for this design (I must add this is with the top cover removed).
Both amps have some oscillation on bottom and top half (more so on the bottom part) of the sine wave when I get to about 3/4 of the max rated power. It's a little worse on the one I repaired but only very very slightly so I'm guessing this is normal for this design (I must add this is with the top cover removed).
I believe this is Bode Nyquist instability, Depaj. The only way to fix this is to increase the lag compensation cap - usually ceramic NPO or even silver mica - around 25%.
The amp has insufficient phase margin; that is, at the high pole we loop gain is at unity the phase shift if OVER the 180 degree from audio frequencies. This means negative fb presented to the negative feedback node the fb is now POSITIVE and the amp is in an oscillator.
If you can identify that cap, it is easy to fix, BUT, the sound will be different, 'slower', and possibly a 'dead' midrange.
Setting this cap precisely is one of the difficult aspects of good design. I know Metaxas; he is a clever guy, but he cannot be told. He has been told about this issue but refuses to acknowledge it. Such is life, people are people......
Nice looking amp, too.
Hugh
The amp has insufficient phase margin; that is, at the high pole we loop gain is at unity the phase shift if OVER the 180 degree from audio frequencies. This means negative fb presented to the negative feedback node the fb is now POSITIVE and the amp is in an oscillator.
If you can identify that cap, it is easy to fix, BUT, the sound will be different, 'slower', and possibly a 'dead' midrange.
Setting this cap precisely is one of the difficult aspects of good design. I know Metaxas; he is a clever guy, but he cannot be told. He has been told about this issue but refuses to acknowledge it. Such is life, people are people......
Nice looking amp, too.
Hugh
Hello Aksa,
Yes Rm suggested this in a previous post. The cap (styroflex) was originally 10pF, I already ramped it up to 56pF. As it is known to be somewhat unstable I did this little mod before actually testing the amp so I don't know if the oscillation was as bad before and if it sounded any different than it did before (which is pretty nice on my pair of ditton 66).
Did I go too high, not high enough ??
Yes Rm suggested this in a previous post. The cap (styroflex) was originally 10pF, I already ramped it up to 56pF. As it is known to be somewhat unstable I did this little mod before actually testing the amp so I don't know if the oscillation was as bad before and if it sounded any different than it did before (which is pretty nice on my pair of ditton 66).
Did I go too high, not high enough ??
I replaced the caps with 22pF silver Mica. It didn't change a thing to the oscillation but the amps definitely sound better. They were slightly "muddy" with the 56pF and lacked presence in the high frequencies.
They sound really good on my ditton 66, I think they are probably the first that really get the low frequencies exactly right on these speakers.
They are very susceptible to mains noise and changes though. when the fridge turns on or off or some of the light switches it gets right through to the speakers.
Thanks for your precious help
They sound really good on my ditton 66, I think they are probably the first that really get the low frequencies exactly right on these speakers.
They are very susceptible to mains noise and changes though. when the fridge turns on or off or some of the light switches it gets right through to the speakers.
Thanks for your precious help
No change at all by adding this cap.
The noise level is also not that great (with or without input). Unless I short the input there is quite some hiss and a buzz but above mains as well as rectified frequencies. Again it's exactly the same on both amps (repaired and original), is this, together with the oscillation, just the way they were designed in order to get this "extreme" speed they are advertised for ?
The noise level is also not that great (with or without input). Unless I short the input there is quite some hiss and a buzz but above mains as well as rectified frequencies. Again it's exactly the same on both amps (repaired and original), is this, together with the oscillation, just the way they were designed in order to get this "extreme" speed they are advertised for ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.