Tube Amps / Music Electronics
For current discussions, please visit Music Electronics Forum. The sunn still shines online!

ampage archive

Vintage threads from the first ten years

Search for:  Mode:  

Grid leak bias hum problem

11/19/2005 3:03 AM
Greg Simon Grid leak bias hum problem
Hello. I recently overhauled my uncle's 1958 Gibson GA-20T, and it uses a 12ay7 in grid-leak bias for one of the channels. The other channel is a 5879 pentode, and it also has trem on the second channel. The 12ay7 channel hums quite a bit more than the 5879 channel, and it is before the volume control for the most part, since the volume makes the hum level go up, but it does hum a little with the volumes down too. I left the layout stock, and replaced all caps and resistors in the amp. I also added a 3 wire power cord, and added a two resistor 100 ohm artifical CT to the filaments and lifted the other end from ground.  
The question is, does grid leak bias just hum more and I have to be satisfied with it, or can it be made quiet? I think it is a 120 cycle hum, but I'm not positive.  
Thanks for any help or suggestions that anyone can provide.  
11/19/2005 11:06 AM
Rob Mercure
Hey Greg,  
On the road and just checking in but wanted to send you a quick reply. No, grid leak bias is not any more hum prone than other schemes - one of my primary amps uses it and is very quiet. I'd suspect layout or perhaps insufficient B+ filtering that may be being bucked out in the pentode stage.  
But I'm sure other's with more time (am on a library computer) will provide more clues/solutions.  
Good luck with it.  
11/19/2005 1:06 PM
Greg Simon
Thanks for the reply Rob. The layout, aside from having parts on both sides of the board is actually pretty good, with the signal coming into the jacks at the top of the chassis, going through the caps and resistors below it on the board, and then straight to the tube socket below that. Then it goes to the left and crosses the board on its way up to the volume control. Since I know its before the volume, maybe I'll try some shielded cable to see if that affects it. Hoping to have it solved before thanksgiving as I'm supposed to give him his amp back then! Anyone else with more suggestions?  
11/19/2005 2:33 PM
Ray Ivers
Apparently grid-leak biasing used to be employed as a low-noise biasing technique for phono-preamp and microphone input stages, way back when; a hard-grounded cathode is technically better than a positive-biased one for rejecting induced/coupled heater hum. The huge 10M grid resistance (your amp may have 1M) ain't so hot noise-wise, especially at very low frequencies, but for the lion's share of the AC passband at least it's shunted by the pickup's (or pedal's) input impedance.  
There's three different schematics of this amp in the latest Groove Tubes book's CD collection, but in each one the pentode stage's output has five cascaded high-pass filter sections, each one with an F3 @ 32Hz, for tremolo feedthrough suppression; it's reasonable to assume this will result in a very noticeable reduction in the 60Hz output of the pentode stage. If your hum is in fact 120Hz, the pentode's superior PSRR (Power Supply Rejection Ratio), along with the trem filter matrix, may well account for the lower hum level from that stage.  
This amp appears to have marginal B+ filtering; if you haven't already you might try at least upping the 10uF on the preamp's B+ node to 20 to 50uF, and if it were me I would at least double all of the B+ caps' values. If that doesn't help, perhaps it's a grounding issue, maybe related to the input jacks, volume control, etc. The shielded wire shouldn't hurt either; you might want to plug in a single-coil guitar and pass its pickup around the chassis a bit to see just how much hum is "in the air", so to speak, and if so, where it's concentrated.  
11/19/2005 2:51 PM
Greg Simon
Ray, thanks for the reply! I've always heard that grid leak was noisier, but maybe thats from people who don't know and are just passing around rumors. This amp does have 10M grid resistors, and .01uf gird caps, and the cathodes are grounded. I added a 100ohm resistor CT to the heaters and lifted one end from ground, but that didn't change the hum level much. I was wondering what they were doing with those caps after the I know! When you turn the volumes up, you can hear the hiss from the higher gain on the 5879, but its just hiss. When you turn the volume up on the 12ay7 in grid-leak bias, it hums the more the volume goes up, so I know the problem is before the volume control. I'll try a different cap and see if it makes a difference. I would think if that was the problem that the 5879 would be humming too since it shares the 10uf cap, but I'll try a larger one anyway. I got another suggestion from a local friend to check the inputs and make sure the jacks around grounding, so I'll check that too. I can induce hum in the chassis with a chopstick depending on where I put it, but I suppose I could try a pickup too.  
We tried to use the amp at a local jam last wednesday and the hum was so bad on the 12ay7 channel we used the other one. It still hummed more than a '64 Deluxe Reverb, but it was at least useable, whereas with the 12ay7 channel it was unuseable. I'll post updates soon later tonight hopefully. I need to have it fixed by thanksgiving which is when he gets his amp back. Thanks for the help!  
11/19/2005 4:18 PM
1. I assume you tried replacing the tube.  
2. Proper wiring at the input jack.  
3. Proper grounding of the 12AY7 stage.  
Reflow all solder joints.
11/19/2005 4:26 PM
Ray Ivers
I dug out the R.D.H. - almost wish I hadn't... :)  
p. 489: "Grid-leak bias is not very suitable for use with low-level (pre-amplifier) stages owing to hum".  
p. 786: "High resistance grid-leak bias has not been found satisfactory for low level operation".  
Fair enough - now p. 1281: "Pentode valves... are not so suitable for grid-leak biasing. However when used with [a high-value Rs] and [small input signal levels] they can be satisfactory, and remove some of the hum problems from [preamps] since there is no bias supply to be filtered, and the cathode is at ground potential so that heater-cathode conductance is not troublesome." (Items in [brackets] are mine.)  
So... I think it would be safe to say that back when grid-leak-biased "mu amps" were considered low-noise, there was still much to be learned about truly low-noise operation. :D I hate to throw this out at this late stage of the game, but do you know for sure the amp was quiet to begin with?  
The pentode will tolerate more B+ hum on the plate supply than a triode will, for the same output noise level. The screen supply is another story, but the .25uF screen cap provides quite a bit of additional filtering/hum suppression.  
You might try putting the heaters back as stock with clip leads - sometimes several conflicting 'wrongs' can make a half-assed sort of 'right' where hum is involved. I don't mean at all to endorse this kind of "engineering", but if it works you may save yourself a lot of grief. I worked on an AC30 a while back (hey, Joe!) and ungrounding/balancing the heater string uncovered a bunch of ground-loop hum weirdnesses that I (hopefully) got mostly squared away, after a lot of detective work.  

  Page 1 of 2 Next> Last Page>>