Hum Help

[Daver]

OK, now I made a ground buss connected to the chassis only near the input. I connected my star grounds from each stage to the buss. Absolutely no change. It still hums. It may even be a little worse than before.

Dave

[R.G.]

I attached the PT center tap directly to the filter cap (-). No help. Pulling the power cord stops the hum right away.
...
Again, this hum seems to be completely surrounding the second stage.

OK, so the hum stops when the power cord is pulled. It's conducted in on the power cord, not picked up from outside.

First - have you replaced that second preamp tube? Tube swapping is always the first step.

Next - what are the DC voltages on the suspect stage, all pins?
Next - does shorting the grid to ground with a BFC (Big Freaking Capacitor) change anything?

[Daver]

R.G.,

Yes, I have tried several different tubes. That's one of the first things I tried.

The second stage is a paralleled 12AX7. 100K plate R and 820 ohm cathode R. I've tried using just one half and then the other. The voltages are as follows.

B+ 327V
Plate 163V
Cathode 1.33V
Grid 0v

Clipping a 220uf cap to the grid seems to eliminate the hum, but I do get some extra noise from the clip leads and I pick up the local rock station fairly well.

Thanks for the help.

Dave

[Mark Black]

Wow - tough dog indeed!

In re-reading this epic journey I find myself returning to previous statements about the hum fading when the power switch is flipped off but stopping abruptly if the AC cord is unplugged from the outlet. Something bothers me about that scenario.

Have you verified the AC cord is properly wired within itself? That is to say the AC plug ground pin really is connected to the green wire, neutral blade to the white, and hot to the black (assuming USA color codes). Each of those should be very low resistance, and there should of course be infinite resistance between any of the wires (when not connected in the amp).

I think it was Rick Erickson who found & reported here about some mis-wired molded AC cords he ran across some time ago.

Just a thought...

Good luck!

[Daver]

Mark,
Thanks for putting yourself through this. The power cord is a detachable type. I just checked it and it's wired correctly. I even tried another, different, cord just for luck, but no dice. Thanks!
Dave

[mfratus2001]

Daver,

You really need to read my last reply. My amp now has almost no hum, even with the gain all the way up.

Mike.

[Daver]

Mike,

Sorry if it seems like I was dismissing your post. Some of those things I've done and a couple weren't clear to me. Let's go throught them.

1. Is the twisted pair comprised of both "to" and "from" wires or are you using the gound wire in there?

2. I've done star grounding and now a ground buss with no change. The tone stack and volume pots always had separate grounds.

3. The tone controls have all the leads twisted together.

4. Head case shield helps some.

5. Tried the ungrounding and regrounding of the PT.

6. It the twisted pair here the "from" wire twisted with a ground wire?

7. I have tried both the center tap and 100 ohm virtual center tap. What exactly do you mean by "One of the pins to ground SILENCED what hum was left!"? What pin?

Thanks for any clarification you can provide.

Dave

[EFK]

I'm extremely interested in the 'twisted pair' classification as well; I'm also chasing an incessant hum around (see the equally-long Selmer amp thread nearby) and have tried a LOT of shielded wire which has not made much difference. I'm a little unclear here as to what mfratus is describing: it *sounds* like a twisted pair of wires wherein only one is carrying signal, the other grounded at one end? This would be a neat approach as supposedly shielded wire is often higher capacitance and too much of it may not be desirable? Looking for some clarification here. Thanks!

[mfratus2001]

Daver,

Yes, by twisted pair I mean a ground and signal twisted together. I make a long length before I build by putting the two wires in a variable-speed hand drill, then twist away.

1. In theory, and pretty much in practice, only the ground lead at the driven end of the wire gets grounded, the other end it is left floating. However, in the case of the gain control, the wire from the 1st stage output is paired and grounded at the control end as well, but not the pot's metal body... You're trying to avoid situations where you have multiple grounds on the same wire, as a chassis that could have even a tenth of a volt of AC from one end to the other, and that impressed onto a signal of a few hundreths of millivolts (like your guitar pickup) would be a sizable hum.

The output from the gain pot (the wiper) is twisted with a ground wire as well.

2. I don't think that for preamp wiring it makes much as a difference whether it is star ground or buss ground, as long as there is only one single ground and not chassis ground and some other ground as well. That is hard to put into words, isn't it?
What R.G. was talking about was right, and a few other posters, too. The high-current users (output stage, phase inverter) should ground back to the filter caps to avoid causing sizable current to flow through the chassis or any ground buss.

3. Wires have fields around them, and the ground wire conducts currents in that field to ground. Each of the three wires, treble control, bass control, and mids, should all be run separate and be a twisted pair, one of which is ground. Since the ground is not used at the control end, that is left floating, but the circuit board end (being the source of the signal) is grounded.
So you should have three twisted pairs running to the controls, and one twisted pair from the treble wiper as the output to the next stage.

4. I found grounding the case helped only in keeping the hum down when you touched the control. Cutting it away from the mid control's ground made less hum.

5. False alarm on the transformer ground. But there IS hum on the chassis, I bet.

6. Yes. (had to look at what I said before!)

7. I had the two resistors from the two filament pins that joined and went to ground. I disconnected them from ground, but still joined together. MAJOR hum.
Then I took a clip lead from power supply ground and touched one of the filament pins. (Pin 2, but I'm using 6SN7's, not 12AX7's for this amp.) There was a DRAMATIC drop in hum, to almost nothing. I touched the other pin (Pin 1) and that did not help as much. In fact, I then put the wire back on Pin 2, and took off the resistors, and then I had less hum than at any time.
Note: There is no center tap on his filament transformer. Don't try this if you DO have one.

OOOOooo! I just read where you DO have a center tap! If you try this, lift the center tap AND virtual center tap.

EFK, hope this explanation helps. You will see twisted pairs in a lot of great sounding amps. They work!

These two links show tone control wiring. Though they don't use twisted pairs, they do show a good way to run the wires.

JCM800 2204

http://www.charlestonarea.com/octane...e_Controls.jpg

Mike.

[EFK]

Mike - great explanation. Thank you. I'm going to implement a number of your recommendations in my complete overhaul of an old Selmer that has caused me to lose most of my hair. Your #2, above, is one of my particular problems - too many varied paths to ground and I'd guess the entire chassis is 'lit up' with signals. The twisted pair approach to many of these signal wires sounds like a very effective approach to eliminating excess noise.

[Daver]

Mike,
I'm still unclear on #7. Pin 2 on a 6SN7 is the plate. So are you saying you connected your virtual center tap to the plate of that tube? How exactly is this wired? Thanks.
Dave

I just tried biasing up the filaments up on 50VDC. No chnage. I replaced some of the shielded wire with twisted pairs. That didn't help at all.
I've built many amps. This thing is by far the worst trouble maker I've had to try and sort out. None of the amps I've built made this much noise. Perhaps it's time for an exorcism.

[mfratus2001]

Sorry, Daver. I wasn't looking at the amp when I said that. It's actually pins 7 and 8, and I had 8 grounded for the least noise.
I don't have a virtual ground, and I don't have center-tapped transformer for the filaments.

I don't think you told us what happened when you tried to find the source of the hum by sticking something (just not your finger, but something that carries a hum) around the wires to find out where the sensitive points were.

This cannot be that difficult. Someone else said to take a capacitor and ground different parts of the circuit (going by the schematic) and see where you were able to kill the hum.

Let's look at it this way:

If the hum is in this stage, it is either generated by the tube itself or it is on one of the signal lines. Have you ever pulled the tube out with the power on? If the hum just disappears then it is generated inside the tube (from the filaments) or is being amplified by the tube.
Another quick test. Cut the filament wires to just that tube. If the hum disappears immediately, it is from the filaments. If it fades away, it is elsewhere. This test is only valid if you can take the filament away when the tube is warm.
This separates the filament and the tube stage.

If it is in the tube stage, is it from the power supply? Put your DVM on AC volts and measure your filter caps. There may be some AC on the first filter caps, but there should be only millivolts of AC on the rest of the caps. Otherwise you will hear hum.

Other than that, you are looking at the problem, but not seeing it. We can't see your amp. Sounds to me, though, like you need to unmount the tone controls and gain control from the chassis, and start again. You did something different with this one, or there is some defective component that you did not find or did not test.
Architecture or parts, one or both are bad.

Mike.

[Daver]

Mike,

I now disconnected the filaments from the AC supply and ran the preamp tubes and PI off a 6VDC lantern battery. No change. That should eliminate the filaments. I then disconnected the tone controls by lifting the treble cap off the board at the plate and lifting the slope resistor from the bass and mid caps and removing the wire from the treble pot to the PI. I tacked in a .022uf cap from the second stages plate and connected its very short lead to the shielded wire feeding the PI. With no tone controls and more available gain the noise is worse. With the tone controls hooked up with no second stage tube in the socket and the PI just "floating" with the tone stack as its input, there is very little noise. That should eliminate the tone controls and wiring. Every pot has been replaced.

With the tone stack and the first gain stage removed, if I ground the grid of the second stage, I still get noise. I replaced the filter cap for the second stage with a different brand and value (bigger). No change.

Yes, this shouldn't be that hard to figure out. Yes, there could be something I'm overlooking. I've done some good trouble shooting. I've eliminated many variables and made good substitutions. I've also cursed a lot. I'm starting to think I need to move the tube socket to somewhere, anywhere else in the chassis. It's like its in the path of every possible stray signal. Thanks for hanging in there with me, Mike.

Dave

[mfratus2001]

With the possibility of your chassis being a bad ground, I would have to ask, "where did you ground the grid to?" The safest (as far as least-contaminating) would be to the ground end of the cathode resistor of that stage.

I am sure, but gotta ask, did you check your clip leads?

I found a lot of difference between where I grounded too. I finally added a terminal strip just for grounds. The cathode resistor and grid resistor go to one terminal near the tube, then one wire from that runs to the terminal strip ground.

I don't know what is connected to your chassis, but I don't even have a metal chassis. I can only think of one point you should have the chassis connected to, and that is the ground wire for the cord and one wire from the first filter cap.

Mike.

[Daver]

Some Good Progress/Solution

Well, I've been able to reduce the hum greatly, but I'm not sure why. How's that? If I connect a .1uf cap from ground to one side of the heater string the hum drops a lot. Pin 9, that is. But, only one side and it has to be connected to the second stage tube socket! If connected to some other tube socket it doesn't help. This amp is haunted.
So I've found a solution, but I'm not experiencing that sense of satisfaction. No amount of troubleshooting, substitution or rewiring grounds made any difference. Thanks to all who helped out here. My brain hurts.

Dave