high frequency oscillation and decoupling question

[gary h]

Hi all,

having enjoyed my Aikido line stage for several months now I have noticed one issue that needs to be fixed; very faint high frequency singing. Having searched the forum I assume that this is microphonic oscillation of the 6sn7s. I am also under the impression that it is caused by the transformer since it occurs when no program material is being played, sources are turned off and unplugged. It also can be heard on a speaker pair in another room. The transformer is large and supplies much more power than needed but does get quite warm and audibly hums up close. It is mounted to the top of the chassis as is the Aikido PCB.

I would like to avoid moving the transformer and PS to another box if possible, has anyone found success with gaskets, caulk, or the tranny or other treatments for the tubes themselves?

Thanks,

gary

[nigelwright7557]

What does the oscilation look like ?
Is it short spikes at mains frequency ?

I had a similar problem with caps that slugged the diode causing brief oscilations. It was just a matter of getting the right cap value.

[gary h]

Hi Nigel,

I don't have a scope to put it on but it doesn't have the hum or buzz of 60 or 120 Hz. It is intermittent coming and going like a breeze, sounds like microphone feedback at a constant high pitch but very faint. And when I put my hand on the chassis or tap lightly it stops.

gary

[nigelwright7557]

Originally posted by gary h
Hi Nigel,

I don't have a scope to put it on but it doesn't have the hum or buzz of 60 or 120 Hz. It is intermittent coming and going like a breeze, sounds like microphone feedback at a constant high pitch but very faint. And when I put my hand on the chassis or tap lightly it stops.

gary

It might be an earthing problem ?
Have you tried earthing the chassis ?
Is the screen on the input grounded too ?

[gary h]

chassis, input, output, and PS grounded via a star-ground.

gary

[kevinkr]

How have you handled the filament supplies? Are you sure that filament/cathode insulation ratings are not exceeded in any stage? Have you floated the filaments on some level of dc in order to mitigate this issue, and if so what is the voltage superimposed on the filament circuit?

I have had similar problems in the past when I did not take this issue into account. I am not sure what the mechanism is but in several cases I traced it back to what was apparently electrostatic stress (voltage gradient) between the filaments and their cathodes which in all of these cases were much more positive than the filaments heating them - fixing that issue took care of the annoying, sporadic low level singing/whining noises. (As per your observations taping the chassis would trigger or terminate the phenomena.)

[Geek]

Radio frequency pickup?

[drj759]

Loose tube socket, socket or pin corrosion/dirt?

I just fixed an oscillation on a 12AT7 driver stage that produced a randomly occurring faint singing/hissing sound with no signal input. Scope showed a 68 kHz oscillation. Didn’t directly hear the 68 kHz oscillation but the side effects did cause an audible faint singing in the speakers. The fix was to clean the socket.

Another possibility: cold solder joint.

[gary h]

Hi guys,

thanks for the replies.

Kevin, the 4 6sn7 heaters are getting exactly 6.3VDC from a pair of voltage regulator boards. The plates are getting 295VDC +/- a volt or two from a 5AR4 based PS. I put a circuit between the two to reference the heater voltage at 1/4 B+ above the cathode. I'll enclose a schematic but, if I have interpretted his instructions correctly, I think this is what JB recommends for the Aikido. Perhaps I have done this wrong, it sounds like you and I experience the same symptom.

Gregg, what is a possible solution to RFs? btw, there is no am/fm receiver connected to the preamp, nor one in the house, though I doubt this matters?

drj759, the tubes are so well seated that I initially had problems getting them into their sockets. The sockets are new and clean and JBs boards allow for beefy solder joints. I'll recheck. Come to think of it, the 5ar4 can be wiggled slightly but that is the entire socket, not the tube, and it's leads are soldered well.

thanks,

gary

[Geek]

Hi Gary,

Originally posted by gary h
Gregg, what is a possible solution to RFs? btw, there is no am/fm receiver connected to the preamp, nor one in the house, though I doubt this matters?

Not receiver, but if you have a transmitter nearby, it can cause trouble. A commercial AM, FM or TV transmitter locally can wreak havoc on the high input impedance amps.

I'd start with a grid snubber - 330 to 1K in series with the grids as close to the tube as possible.

Next is bringing out the chokes - 10 to 100uH in series with the grids with a small (10-100pf) cap from grid to ground.

Cheers!