DC coupled 71A amp- its alive but noisy

[vetmed]

It works,  but loud 120 Hz hum I looked at Doc's missive on noise, shorted the inputs, tried a cheater plug, and concluded it must be a cold solder joint in the power supply so reflowed every soldered connection I could and no luck I'm using a pair of regulator boards to power the 71A and I am fairly convinced this has something to do with it, but what I am unsure. I've consulted with the creator of said boards but he is as much at a loss to explain it as I am. The regulator board consists of full wave bridge with Schottky diodes, lots of filter capacitance, a LT1085 regulator, with adjustable voltage output. There are optional pads on the board for a common mode choke which I am using. It sounds very good if I turn it up loud enough to drown out the hum Your help is greatly appreciated and needed.

Robert Lees

[Grainger49]

A 'scope will let you see where the hum is creeping in.  Use it AC coupled, to block the DC.

[vetmed]

No scope I'm afraid. And if I had one I wouldn't know what to do with it Seriously I have 32 mv of noise measured as ac volts at the OT outputs. That translates to some serious ripple getting in somewhere. I'd be better off to use AC and a humpot. And what is most annoying is that this power supply less the regulator boards was powering a "Darling" amp until recently without any problems

Robert Lees

[vetmed]

FWIW I removed the common mode choke from the regulator board and the AC measured at the output went from 32mv to 18mv Don't know what to make of that.

Robert Lees

[Doc B.]

Try measuring AC on the filaments. That might give you an idea of how smooth the filament DC is.

[2wo]

Are the reg boards for the filaments? Remember they need to be completely separate. Try running one channel, first removing one 71, then disconnecting one of the reg boards completely...John     

[Paul Birkeland]

I'd also recommend drawing a schematic, then posting some photos of your build.

-PB

[vetmed]

Following Doc's advice and I measure 1-2 mv ac on the filaments, more or less the limit of resolution of the meter I am using. One board per 71a,both measure identically. Certainly this would seem to rule out the regulator boards but for the fact that removing the CMC lead to a reduction in noise Some sort of interaction between the board and the PS? And the PS seems to be otherwise working correctly since I get correct DC measurements. And it worked before, and the only thing I changed there is a new socket for the 5AR4 to replace the 80, both to increase the B+ and provide a soft start. I'm at a loss as to how to proceed

Robert Lees

[Grainger49]

If the hum is from the heater supply then when you turn off the amp the hum will immediately go away before the amp shuts down.  If it is elsewhere in the circuit it will stay till all sound dies away.

[vetmed]

Thanks Grainger, when turned off sounds stop immediately. Yet there is a lot of capacitance in the filament regulator board and no AC current. There is a separate 12.6 volt AC transformer with CT that was used to heat the 5965 and both 1626 tubes when it was a "Darling" amp, I have retained this to power the 5965 but use 6.3 volts. Could this be the problem?

Robert Lees

[vetmed]

Changed the 5965 heater to the same winding the filament regulator uses, no change I have tried to attach a drawing of the schematic I made. First was told it was too big, re-scanned it at a lower resolution and was told it was an unacceptable file format; I can't attach jpeg files so if some kind soul would help me change to a format that can be uploaded I could then share my misery with you

Regards
   Robert Lees

[Grainger49]

Robert,

Look in my FAQ thread, Grainger's Corner, and there are two ways to post pictures here.  You will need to host them.  I use Photobucket, it is free and kind of easy to use.  But then that is just me.

[Paul Birkeland]

There is a separate 12.6 volt AC transformer with CT that was used to heat the 5965 and both 1626 tubes when it was a "Darling" amp, I have retained this to power the 5965 but use 6.3 volts. Could this be the problem?

That's potentially quite a problem.  You can't just use one 6.3V leg to heat the 5965 with no load on the other half, but if you use the 12V configuration, that shouldn't present a problem.

Also, how much was changed when going from Darling to this amp?  Having started with a working Darling would be helpful in the diagnosis. 

[braubeat]

Hello

I have had hum on several amps I made which turned out to be improperly grounded filaments. All filaments must be grounded to the signal somewhere.

Michael 

[Paul Birkeland]

I have had hum on several amps I made which turned out to be improperly grounded filaments. All filaments must be grounded to the signal somewhere.

Filaments shouldn't be grounded unless the amp is running fixed bias.

Driver stage heaters do indeed need a reference.

-PB