Kegger/blueglow KT88 PS hum

[jank9]

Here's a power supply that has been designed twice, once for 60hz mains and again for 50hz.

Thanks dubadub. I also considered adding additional RC filter but that would require a new PT....

 

[jank9]

What does an FFT of the output show? I wouldn't use anything else to hunt down a hum or hiss issue.

Not sure if FFT is available on my measuring equipment (I am using Analog Discover 2)...

 

[jank9]

Chasing hum in p2p tube builds--especially SE--can be frustrating.

I know we're talking 100hz and not 50hz, but I'd still explore the low hanging fruit. Before messing with the circuit too much, I'd do as others advise above and post pics of the top and bottom of the chassis.

Additionally, you haven't mentioned the rest of your system. If you have something else where signal ground is connected directly to earth (including video or internet), that could be your loop. Sometimes that makes noise; sometimes it doesn't. You can try isolating signal from earth with a 10R or so resistor for testing purposes. If that works, you'd want to add some diodes going each way in parallel for safety and likely a little cap for rf.

Finally, be sure that the electrical hum is all that is irritating you. Edcor PTs (though not only Edcor PTs) often buzz a bit. If you are hearing combined noise, mechanical isolation may be improve your overall experience.

There are many threads on these things with embarrassing situations by some of us fighting bigger problems. 2mv isn't awful in the grand scheme and counts as "quiet" for many with less sensitive speakers, who sit farther from them, or both.

Paul

Thanks Paul, I've had my fair bit of frustration looking for source of this hum, but since I see a measurable improvement after increasing C1 in CLC filter in PS I consider the source of hum to be identified. The hum was with no signal attached so I guess there is no issue with the rest of the equipment.

 

[jank9]

The ripple is smaller in triode mode 2,7mV versus 3,7mV in UL mode (both with C1=20uF in PS).

BTW, shouldn't it be the other way around? Shouldn't UL tap act as a negative feedback here?

 

[20to20]

I'd guess it's higher on the screen tap because it's closer to the raw B+. The ''UL'' connection has been called a NFB loop but David Hafler had adamantly pushed back against that definition with an obvious point that it doesn't reduce the gain of the tube until you get over 50% toward triode loading of the tap point. ....To paraphrase. It's more of a ''tuning'' parallel that gives lower distortion with very little gain loss, until the ''UL'' point is crossed out of its narrow envelope. The screen tap and triode loading have two different affects on the tube/primary Z. So that was the reason for my question, if you had seen a difference in output ripple. Good data.