AndrewT said:
Andrew,
I always use plastic standoff, that is why I did not insulate the mounting holes. I think this has been mentioned before.
\Jens
An externally hosted image should be here but it was not working when we last tested it.
Hi Pooge,
the problem is the shorting of the power ground plane to the heatsink.
Jens' solution in the previous post confirms that the ground plane must be insulated from the chassis.
When I tested the amps out of the chassis, the chassis did not connect the heatsinks together and so the extra earthing loops did not exist. And because the sink was isolated all the little loops from the 6fixing screws were of no consequence.
The amps did not hum.
It is the EXTRA connection from ground plane to heatsink to chassis for BOTH channels that is creating all the extra grounding loops.
Jens,
how is the stand-off connected to the sink, is it male/female at opposite ends? or is there a stud from sink to stand-off?
the problem is the shorting of the power ground plane to the heatsink.
Jens' solution in the previous post confirms that the ground plane must be insulated from the chassis.
When I tested the amps out of the chassis, the chassis did not connect the heatsinks together and so the extra earthing loops did not exist. And because the sink was isolated all the little loops from the 6fixing screws were of no consequence.
The amps did not hum.
It is the EXTRA connection from ground plane to heatsink to chassis for BOTH channels that is creating all the extra grounding loops.
Jens,
how is the stand-off connected to the sink, is it male/female at opposite ends? or is there a stud from sink to stand-off?
Power Output Transister
Hi Kiss, Pooge & Andrew T.
Thanks for all of your comments. I will check the Leach Clone thread again.
Can anyone suggest me for some of your favorite output transistor? It seems like threre are various opinions on this issue. My last question, is there any sonic different when we use different output transistor?
Thanks again, Kittikun
Hi Kiss, Pooge & Andrew T.
Thanks for all of your comments. I will check the Leach Clone thread again.
Can anyone suggest me for some of your favorite output transistor? It seems like threre are various opinions on this issue. My last question, is there any sonic different when we use different output transistor?
Thanks again, Kittikun
kittikun said:
Some say no, others say yes.
Using a higher Ft output device does not alter the design parameters in a circuit, but higher Ft devices have a faster response time. Without changing a single thing in an amp circuit, that alone will effect the sound(imho)
If you do a search/google you'll find others who "feel" alike, e.g. check the TNT page on output devices.
Hi folks!
Dropping in quite late here; hope not to give unnecessary advice. In general with semiconductor amplifiers, when going to different devices, particularly output, it is a good idea to check NFB stability (scope + square wave or such) in the final construction. As Jacco intimated, either a poorer or a better one can upset the apple cart, because each transistor has its particular capacitances. It is not the "sound" re a particular transistor in the first place but insufficient freedom of spurious oscillation or other instability that can cause a different sound, then wrongly attributed to a transistor. There has been a lot of heatsink discussion here including different isolating wafers/materials; in my experience that can also affect optimum NFB response, as can lead lengths - the lot.
Transistors are quite high frequency devices - in my own design I had to contend up to 10 MHz with a particular new construction. This is not common, but showed that it can be important; also one must test at several amplitudes up to maximum. It is a sort of "caveat emptor" thing sometimes neglected.
Regards.
Dropping in quite late here; hope not to give unnecessary advice. In general with semiconductor amplifiers, when going to different devices, particularly output, it is a good idea to check NFB stability (scope + square wave or such) in the final construction. As Jacco intimated, either a poorer or a better one can upset the apple cart, because each transistor has its particular capacitances. It is not the "sound" re a particular transistor in the first place but insufficient freedom of spurious oscillation or other instability that can cause a different sound, then wrongly attributed to a transistor. There has been a lot of heatsink discussion here including different isolating wafers/materials; in my experience that can also affect optimum NFB response, as can lead lengths - the lot.
Transistors are quite high frequency devices - in my own design I had to contend up to 10 MHz with a particular new construction. This is not common, but showed that it can be important; also one must test at several amplitudes up to maximum. It is a sort of "caveat emptor" thing sometimes neglected.
Regards.
KISS, no!
I was not aware of that. I have had a look at Prof. Leach's work and greatly respect him (have print-outs of his articles), but I have not gone into his amplifier that much. Penalty for designing one's own stuff! I will take a closer look then.
Yes, folks think I am funny, but I also use - well, not quite an overall ground plane, but copper here and there for screening/ground plane effects. I have to, that's also why I had to trim h.f. stability after a built-up board, grounding of heat sinks and the like. (The penalty for designing a wide open-loop bandwidth and other measures to kill high-order harmonic distortion in my case.) Designs with emitter/source follower outputs have the advantage there, but I use FCP.
Regards.
I was not aware of that. I have had a look at Prof. Leach's work and greatly respect him (have print-outs of his articles), but I have not gone into his amplifier that much. Penalty for designing one's own stuff! I will take a closer look then.
Yes, folks think I am funny, but I also use - well, not quite an overall ground plane, but copper here and there for screening/ground plane effects. I have to, that's also why I had to trim h.f. stability after a built-up board, grounding of heat sinks and the like. (The penalty for designing a wide open-loop bandwidth and other measures to kill high-order harmonic distortion in my case.) Designs with emitter/source follower outputs have the advantage there, but I use FCP.
Regards.
Hi,
after powering off at the mains and also shutting down and/or muting the pre-amp, the speakers go silent.
But after about two to three minutes by which time the voltage on the PSU caps will have dropped quite low, there is a loud hum that lasts about ten seconds. I mean louder than someone talking quietly.
Any thoughts on where I should look?
after powering off at the mains and also shutting down and/or muting the pre-amp, the speakers go silent.
But after about two to three minutes by which time the voltage on the PSU caps will have dropped quite low, there is a loud hum that lasts about ten seconds. I mean louder than someone talking quietly.
Any thoughts on where I should look?
AndrewT said:
That isn't a problem with my 4.5 version boards from Leach, nor with Brian Bell's board, but an earlier 2.x version appears to have a similar problem. Both channels oscillate and I never determined exactly what is going on; the problem moved with the boards to another chassis. I did note that the rails bleed down very asymmetrically, and I think therein may be a clue. There was substantial DC offset as well, which was not at all an issue when the amplifier was on, nor were there any significant turn-on transients.
I minimized the problem by adding some aggressive bleeder resistors to help force the rails down more evenly, but this did not completely end it. I think the problem is in the input stage, but I won't swear to that theory.
I was never completely happy with the earlier boards' sound; sometimes something about the sound was 'off'; the later boards even though nearly identical electrically had a more open and relaxed sound, and I wonder if there wasn't some marginal instability (many of Leach's earlier boards incorporated a ground plane).
Is this mains hum or at another frequency?
It seems like mains hum.
But I have discovered that it still hums for that brief period even when the amp is unplugged from the mains wall socket.
If it's mains hum then it can only be coming from the pre-amp. But it's off when I hear the hum.
It seems more likely that it is some kind of low frequency instability, but I am completely unsure, i.e. I'm guessing.
A solution would be an instant off relay on the speaker output. But that covers over the cracks.
But I have discovered that it still hums for that brief period even when the amp is unplugged from the mains wall socket.
If it's mains hum then it can only be coming from the pre-amp. But it's off when I hear the hum.
It seems more likely that it is some kind of low frequency instability, but I am completely unsure, i.e. I'm guessing.
A solution would be an instant off relay on the speaker output. But that covers over the cracks.
Hi,
opened it up and took some measurements.
With the amp plugged in, but switched off, the rail to rail voltage drops from 116V to 40V fairly quickly then slowly falls to 23V rail to rail. The noise builds up to 250mVac and then subsides and is effectively off when rail to rail is 18V.
With the plug removed from the wall the voltages decay similarly but the speaker output noise now only rises to a peak of about 30mVac
That 250mVac will produce about 65db in my speakers whereas the 30mV will be only 47db.
opened it up and took some measurements.
With the amp plugged in, but switched off, the rail to rail voltage drops from 116V to 40V fairly quickly then slowly falls to 23V rail to rail. The noise builds up to 250mVac and then subsides and is effectively off when rail to rail is 18V.
With the plug removed from the wall the voltages decay similarly but the speaker output noise now only rises to a peak of about 30mVac
That 250mVac will produce about 65db in my speakers whereas the 30mV will be only 47db.
Andrew,
How is your power supply set up?
Do you use two supplies e.g. one for input and one for the output stage of the amp?
You may need control the sequence of shutdown to ensure that the input + VAS has no power when the output stage is turned off.
When the input + VAS section is without power the amp should be silent.
\Jens
How is your power supply set up?
Do you use two supplies e.g. one for input and one for the output stage of the amp?
You may need control the sequence of shutdown to ensure that the input + VAS has no power when the output stage is turned off.
When the input + VAS section is without power the amp should be silent.
\Jens
Hi,
down from +-58 to +-25V they remain in sync.
The positive then falls much more slowly than the negative.
when it reaches about +19V-5V the hum starts and when it gets down to +17V-1.5V the hum has stopped.
So could it be very asymmetric voltages that are causing the problem?
I wonder if one of the smoothing caps (3pairs of 15mF=>+-45mF) has degraded in some way that is causing the asymmetry? That might account for only noticing this recently.
I did notice one cap bulged before I got as far as testing and discarded it.
Faster discharge using resistors may keep them more in sync, but it will increase the quiescent ripple. I'll try that.
down from +-58 to +-25V they remain in sync.
The positive then falls much more slowly than the negative.
when it reaches about +19V-5V the hum starts and when it gets down to +17V-1.5V the hum has stopped.
So could it be very asymmetric voltages that are causing the problem?
I wonder if one of the smoothing caps (3pairs of 15mF=>+-45mF) has degraded in some way that is causing the asymmetry? That might account for only noticing this recently.
I did notice one cap bulged before I got as far as testing and discarded it.
Faster discharge using resistors may keep them more in sync, but it will increase the quiescent ripple. I'll try that.
Andrew,
This will occur now and then, since there is no predicting how an amplifier wil perform at various points as the voltages deteriorate. I have found momentary h.f. oscillation even more often. One cannot even simulate this, as it will also depend on the parameter spread with practical transistors.
As I have a delayed loudspeaker connecting funtion in my amplifier (otherwise it gives quite a plop turning on and off), I forget about it. It becomes a little tall if one has to design for that as well - in fact, I am surprised if it will not happen (momentary, that is) in any but the most heavily rolled-off designs.
The most I would do is something as Jens suggested or a quick signal short somewhere suitable in the signal line until voltages are low enough. (One sometimes has to do that in instrumentation amplifiers.) Yes, I would prefer not to have a relay in the loudspeaker circuit, but all considered (plus using a good relay, not one of those warts/reeds) I am relieved that I am free of such a concern. There are enough others.
Regards
This will occur now and then, since there is no predicting how an amplifier wil perform at various points as the voltages deteriorate. I have found momentary h.f. oscillation even more often. One cannot even simulate this, as it will also depend on the parameter spread with practical transistors.
As I have a delayed loudspeaker connecting funtion in my amplifier (otherwise it gives quite a plop turning on and off), I forget about it. It becomes a little tall if one has to design for that as well - in fact, I am surprised if it will not happen (momentary, that is) in any but the most heavily rolled-off designs.
The most I would do is something as Jens suggested or a quick signal short somewhere suitable in the signal line until voltages are low enough. (One sometimes has to do that in instrumentation amplifiers.) Yes, I would prefer not to have a relay in the loudspeaker circuit, but all considered (plus using a good relay, not one of those warts/reeds) I am relieved that I am free of such a concern. There are enough others.
Regards
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.