Hello,
I would like to start a discussion about microphonic affections and effects in low-level high-gain solid-state circuits such as phonopres or mic-amps having cl-gain of 10.000 or even more. This might include discreet as well as opamp circuits.
Maybe I choose the wrong clue-words, but I only found one closely related thread here , and no solution or even explanation has been offered.
Every now and then I used to knock an certain components in previous designs, then thinking ‘strange to hear that knock in the speakers, what might that mean to accurate music reproduction?’ but then forgetting about it immediately.
My actual ‘project’ shows some (in my view) excessive microphonic behaviour. At maximum volume, cautios tapping on the top of decoupling caps makes quiet a boomy bassdrum, one cap in the voltage regulator serves just too well as a microphone at that same volume. It’s a mc-phonostage, but instead of asking for a fix I’d like to take a broader perspective:
Rüdiger
I would like to start a discussion about microphonic affections and effects in low-level high-gain solid-state circuits such as phonopres or mic-amps having cl-gain of 10.000 or even more. This might include discreet as well as opamp circuits.
Maybe I choose the wrong clue-words, but I only found one closely related thread here , and no solution or even explanation has been offered.
Every now and then I used to knock an certain components in previous designs, then thinking ‘strange to hear that knock in the speakers, what might that mean to accurate music reproduction?’ but then forgetting about it immediately.
My actual ‘project’ shows some (in my view) excessive microphonic behaviour. At maximum volume, cautios tapping on the top of decoupling caps makes quiet a boomy bassdrum, one cap in the voltage regulator serves just too well as a microphone at that same volume. It’s a mc-phonostage, but instead of asking for a fix I’d like to take a broader perspective:
- How much microphonic behaviour is acceptable (read ‘unavoidable’)?
- Which flaws and flakes provoke an overly microphonic behaviour? (maybe aside from loose ground nodes)
- How do you cope with it?
- When all is done what could be done circuitwise, what mechanical additions can you think of? In old radios you sometimes see some sort of wax poored over certain areas. Some people seem to embody the whole circuit in epoxy, a technique which I would not feel too well with, and it’s no option in the making. Speaker wool comes to my mind, but it certainly affects to many parameters of the circuit, e.g. temperature.
Rüdiger
OK,
in this case, it was a degraded op-amp (or two) that were accidentially fed with one rail side
(on of them serving as servo),
and the issue is gone for now with new ones. Do I presume right then, that microphony is always a failure in solid state?
Rüdiger
in this case, it was a degraded op-amp (or two) that were accidentially fed with one rail side
(on of them serving as servo),and the issue is gone for now with new ones. Do I presume right then, that microphony is always a failure in solid state?
Rüdiger
That's a new one for me! I've encountered two sources of microphony in new SS gear I've built- capacitors (especially the premium audiophile-grade ones) and Teflon shielded cable. In any of these cases, yes, it's a defect. A properly operating piece of solid state amplification should emit no noise when tapped.
In older gear, the most common cause is connections- cables, plugs, switch contacts, dodgy solder joints.
In older gear, the most common cause is connections- cables, plugs, switch contacts, dodgy solder joints.
Hi SY,
probably, I was a bit unclear. Surely it were mostly the caps. But it was so severe, that all you tapped made a noise, and all those noises sounded a bit different. All this was due to those faulty opa's.
But one might get into thinking: what is with this special cap that served as a microphone that you could hear your voice through the speakers? Now that is gone, but maybe it would pay if you apply some damping on that device? Many audiochains tend to sound worse when played very loud, it would not surprise me if that is partially through acoustic coupling.,
Rüdiger
probably, I was a bit unclear. Surely it were mostly the caps. But it was so severe, that all you tapped made a noise, and all those noises sounded a bit different. All this was due to those faulty opa's.
But one might get into thinking: what is with this special cap that served as a microphone that you could hear your voice through the speakers? Now that is gone, but maybe it would pay if you apply some damping on that device? Many audiochains tend to sound worse when played very loud, it would not surprise me if that is partially through acoustic coupling.,
Rüdiger
If a cap is microphonic, out it goes! Why band-aid it when there are plenty of good, non-microphonic caps widely available?
Onvinyl said:OK,
in this case, it was a degraded op-amp (or two) that were accidentially fed with one rail side
and the issue is gone for now with new ones. Do I presume right then, that microphony is always a failure in solid state?
Rüdiger
Hi Rudiger,
Microphony is often a problem with high impedance, high gain circuits. Particularly bad are ceramic capacitors which exhibit a piezo electric effect. These should not be used in the signal path!Coaxial cables can also be microphonic, since they are essentially a long capacitor when feeding a high impedance. Some high end manufacturers go to great lengths to provide acoustic isolation for their PCB assemblies, with rubber mountings, and acoustic dampening panels inside the equipment box, and external isolation feet etc.
Paul J
Just about anything with voltage across it or current through it is potentially microphonic, I have used pcb track as a strain guage in the past.
It is important to mount a pcb rigidly, just supporting the corners of a large board is asking for it to be flexed. Once that happens you are straing the leads of components and will have some effect on the value.
Capacitors with a dc bias on them are very microphonic and the effect is proportional to value, so this is one reason why large ecaps in mc and mm preamps have a bad reputation.
It is important to mount a pcb rigidly, just supporting the corners of a large board is asking for it to be flexed. Once that happens you are straing the leads of components and will have some effect on the value.
Capacitors with a dc bias on them are very microphonic and the effect is proportional to value, so this is one reason why large ecaps in mc and mm preamps have a bad reputation.
I have not read this anywhere but by common sense, the wax must be preventing the component from changing their original position. If they are moved then the RF circuitry tuning will be lost. At list that is the reason IMO for the wax in FM frontend of ss radios.
Gajanan Phadte
- Status
- Not open for further replies.