• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Why do valves go microphonic?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
My phono preamp (self-built) has as its input valve a triode-connected D3a. It has been playing quite happily for months but yesterday the right-hand channel went microphonic contributing its own monotonic drone to the music. Nothing I did to decouple the preamp chassis from its environment had any effect, but substituting a fresh D3a did the trick. Now the D3a is a high-gain (mu = 70), high gm (40ma/V) triode, but I am currently playing Berlioz's Damnation de Faust at a satisfying volume-level without any microphonics at all. Is it something to do with the frame-grid distorting, or what?
 
Administrator
Joined 2004
Paid Member
I use these in the front end of my phono stage as well, triode connected like you. I have floated the filament supply about 35V above ground which it seems from prior experience prevents this monotonic drone - "singing" if you will in this and related types. So far over the last couple of years neither D3A has begun to sing along atonally or otherwise with my music..

While I have not investigated the mechanism thoroughly, I have encountered this problem with 5751, 12AX7A and some other types in circuit topologies where the difference between cathode and filament potentials was either small or the cathode was significantly more positive than the filament supply reference point. (Ground or a bypassed voltage divider.) Making the filament supply somewhat positive relative to the cathode seems to help for in most instances.
 
Microphonics is an old topic... basically it's caused by any mechanical excitation in which one or more elements or structures move thereby altering the electron flow. A reasonably detailed explanation and some coping concepts are contained in an old book I found recently on Pete Millet's site by Tomer in 1960.

http://www.pmillett.com/Books/Tomer_1960_Getting_the_Most_Out_of_Vacuum_Tubes.pdf

As tubes are used, many stresses are placed on the elements and their mounting. Also, on/off cycles (heating/cooling) contribute to this over time as grid wires may sag or get limp, the cathode structure may become loose in the mica insulators, etc. You seem to have found that a tube has gone bad in your case.

Tubes with higher mu factor generally exhibit greater sensitivity to microphonics as the same mechanical movement of the elements will have a greater effect, hence 12AX7s and 5751s will generally show as being more sensitive than a 12AU7 for example.

Lifting the filaments from the cathode potential can solve some hum coupling but generally not a "true" microphonic situation caused by mechanical excitation. Also interesting to note that you can also make the filaments negative with respect to the cathode... same effect, but is rarely, if ever used, as most circuits that employ this are done to compensate for certain tubes having their cathodes running at fairly high DC voltages such as phase inverters.

DHTs can be a real problem here due to having multiple runs of filament wire suspended between points. I have found single-plate 2A3 triodes to be highly microphonic... literally a handclap within 2 meters will show quite an effect on a scope as will whistling at a resonant frequency... tapping on the chassis with a fingernail will make them go bonkers. The same applies to the EML 45 new manufacture triodes. Each have many cathode wires and are quite long. Traditional NOS 45 ST glass seems very immune to microphonics... moreso than older balloon glass versions.

Microphonics is also considered a lifespan parameter... i.e., when a tube becomes microphonic, you replace it.

Regards, KM
 
I guess the valve going "bad" (evil would be closer to the mark) is the more likely explanation as the D3a's heaters are floated on +45V in my preamp I suppose the frame grid is very close to the cathode to get the gm and mu figures I quoted previously, but I assumed the frame grid technique produced a structure that was less prone to microphonics. Perhaps not. I'd better lay some more stocks in. Thanks for your replies, chaps.
 
Administrator
Joined 2004
Paid Member
In this case the mechanism does not appear to be microphonic in the conventional sense which is why I did not use that term in my response.

I have actually measured pre-amp circuitry where I could see a continuous tone being generated at anywhere from 400Hz to a couple of kHz, and did not appear to be an external feedback or layout induced oscillation either. The level measured was often on the order of just a couple of mV on the plate and very hard to see.

I have not been able to come up with a satisfactory explanation for the observed phenomena, and my current building/design techniques seem to preclude its occurrence in my gear, so it is quite possible that when I encountered it I was not equipped to investigate to root cause (very limited budget and antique test gear) - in my case it was definitely related in some way to filament vs cathode potentials and changing them eliminated the problem at the time.

Now that leaves me wondering whether or not the noise I was hearing was actually the 6th or higher line frequency harmonic in the rectifier ripple, not fully removed by the finite ripple rejection of the voltage regulators in my filament supplies, and coupled by excessive capacitance between the filament and cathode sleeve?

I would assume that as the filament heats up it expands slightly and if the cathode does not expand at the same rate that filament to cathode capacitance could increase. My experience with this problem is that it started a few minutes after the unit was turned on.

Is it possible that the increased electrostatic field generated by biasing up the filament supply actually reduces the capacitance between the filament/cathode and/or more likely reduces leakage currents between them?

This is all unverified suspicion on my part...
 
kevinkr said:
In this case the mechanism does not appear to be microphonic in the conventional sense which is why I did not use that term in my response.

I have actually measured pre-amp circuitry where I could see a continuous tone being generated at anywhere from 400Hz to a couple of kHz, and did not appear to be an external feedback or layout induced oscillation either. The level measured was often on the order of just a couple of mV on the plate and very hard to see.

I have not been able to come up with a satisfactory explanation for the observed phenomena, and my current building/design techniques seem to preclude its occurrence in my gear, so it is quite possible that when I encountered it I was not equipped to investigate to root cause (very limited budget and antique test gear) - in my case it was definitely related in some way to filament vs cathode potentials and changing them eliminated the problem at the time.

Now that leaves me wondering whether or not the noise I was hearing was actually the 6th or higher line frequency harmonic in the rectifier ripple, not fully removed by the finite ripple rejection of the voltage regulators in my filament supplies, and coupled by excessive capacitance between the filament and cathode sleeve?

I would assume that as the filament heats up it expands slightly and if the cathode does not expand at the same rate that filament to cathode capacitance could increase. My experience with this problem is that it started a few minutes after the unit was turned on.

Is it possible that the increased electrostatic field generated by biasing up the filament supply actually reduces the capacitance between the filament/cathode and/or more likely reduces leakage currents between them?

This is all unverified suspicion on my part...

As you pointed out... not mechanically induced. It's possible to have a diode effect occur between the filament wire and the cathode sleeve due to breakdown of the insulation at one or multiple points. Imposing a DC bias on the filament can help alleviate this. A DC filament also helps as it can result in a static diode effect (if that's the cause). This would still be more perceptible on high mu tubes however.

Regards, KM
 
It wasn't a "cot death" as that valve had been working fine for six months or more : probably just a catastrophic failure. Having involuntarily done a 12Mb download of the Tomer book (shouldn't it still be in copyright, by the way, being less than 75 years and having been recently republished by Audio Amateur?) I should probably read at least the first chapter.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.