Stray hum pickup *preamp* cathode R-bias vs. battery bias & 12AU7vs.5687
This one's got me. (and yes, I've searched on hum til my eyes bled) -- can you say "star ground"? I can!
So, here it is: my preamp (12AU7 common-cathode followed by cathode follower) was all quiet (just tube rush) when I had a bypassed resistor bias on the gain stage cathode. Switching to a battery bias brought on considerable hum. (I tried both traditional battery-on-cathode as well as a fixed-bias scheme for fun, C on input, grounded cathode - same noise).
Here's more on the noise: basic 120Hz noise that escalates with increasing level on a BentAudio stepped attenuator. This is a PCB-relay based unit with SMD resistors. It is a "hybrid-ladder", the shunt resistor is switched out every 5 steps only. Indeed, you can hear the noise go from 0-level on up like:
That is until it gets to MAX, i.e no voltage division/full signal = NO NOISE!!! So it's really loud at "60" and quiet at "61" (max). Wha? Clearly the noise is entering the attenuator. So:
I tried shielding the entire attenuator module of one channel in a Hammond Aluminum cast box, grounded that, and ran a shield to the tube socket... very little improvement here unfortunately. Yes, I am using grid stoppers (510R). I CAN make the noise go away by running the filaments etc. from a battery; this is not B+ noise. Trying to make a quieter filament supply only helps a little (adding capacitance, trying a capacitance multiplier). The filament supply is lifted 75V. Removing the lift has no effect.
The most recent thing I tried REALLY got me twisted: subbing in 5687 tubes for the 12AU7s makes the noise disappear again! What gives??? I thought maybe the 12AU7s were dumping excessive grid current through the attenuator, but measuring voltage offset shows that this is not the case. Different brands of 12AU7 all do the same. I know, I know, the 5687s sound better and fix the noise, but I just need to know what's going on here!
Anybody?:spin: :spin: :spin:
best way to learn is without any help...
... that way you learn more.
I have basically fixed the problem - by finally upping the value of the capacitor which 'floats' the filament supply at 90V in reference to ground. Increasing the value from 0.10uF to 100uF did the trick! The rest is all to do with sub-optimal layout on my part.
In the process I found that one of the resistors in the voltage-divider for the filament supply float had gone bad and I was only getting 22V float. 90V is more like it. I'm using a 1M-680K divider. What do the rest of you use for this network?
And seriously, why would a battery bias attarct more noise to begin with, and why would a different but not too dissimilar input triode be that much quieter??? seriously, I want to learn something here.
OK... im going to take a cue from programming here and ask - are you sure it was teh switch to the battery bias? You could have done something inadvertently in the switch... and are misplacing the blame? You could have even just moved a wire inside the unit an inch or two near a hum producing component for instance... I guess Im saying, it's not always the obvious.
I haven't ever messed with a battery bias supply so I can't comment much on that, but by technical reasoning... I don't see how switching from a resistor to a battery should matter... did you remove the bypassing capacitor as well? You might want to leave it in place, even with the battery...??
|All times are GMT. The time now is 05:02 PM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio