Hum Help

[Daver]

Mike,
I've got a 3 prong cord and grounded outlets. Why does the AC only affect the second stage? I can try swapping the PT primary's. It'll be a pain, but if that works, so be it. Maybe I'll try unbolting the PT first to see if it helps. Thanks!
Dave

[loudthud]

How is the primary of the amp wired? Are the fuse and AC switch on the same side of the power line? One side of the AC line should be 120VAC to the safety ground and the other just a couple of volts (the neutral side). Please verify this with your DVM. The switch and fuse should both be on the 120V side. I can see where if the switch is on the neutral side of the line that the 120VAC is connected to the transformer when the power is switched off.

[mfratus2001]

Mike,
I've got a 3 prong cord and grounded outlets. Why does the AC only affect the second stage? I can try swapping the PT primary's. It'll be a pain, but if that works, so be it. Maybe I'll try unbolting the PT first to see if it helps. Thanks!
Dave

The second stage is where you have 3 pots grounded to the chassis that feed into a high-impedance input. The second stage input is also a shielded wire that is grounded.
If your ground is polluted, everything connected to it is polluted.

But the proof is waiting for when you break the connection between the power transformer and ground, or swap the wires.

EFK: You need to follow this thread from the start. There's a lot of questions you need to answer to see if it is anything else first. Hum is sometimes (as you know) hard to solve. This is not a common problem.

Mike.

[Daver]

loudthud,
The hot side of the AC is fused and also has the power switch. The neutral is connected directly to the PT.

Mike,
Well, I do have sheilded cable to and from the volume pot previous to the second stage and at the input. These don't seem to be causing hum. At any rate I will try unbolting the PT and see what happens. I'll report back.

Thanks, guys!

Dave

[Daver]

****UPDATE****

I unbolted the PT and put some thin cardboard in between it and the chassis where it might still touch. The hum is still there. So I'm assuming that swapping the primary leads isn't going to change anything. Crap. I was hoping we were on to something. Thanks.

Dave

[drewl]

"Swapping the primary wires on the power transformer is only cheating. The real problem (if that takes away the hum) is leakage from primary to core."

Can you elaborate on this in lay language? I'm having what I believe is a very similar problem, hum that cannot be eliminated by any means. Old Selmer amp. Is the leakage just symptomatic of an old/poor quality transformer? how does swapping primary wires affect the hum? I'm very interested. Thanks!

This sounds like the problem I'm having with this Ampeg M-15.
The hum/buzz sounds like it's being radiated from the AC line, and goes away as soon as the amp is turned off.
i tried all of the above ideas, swapping AC line, removing filament ground and creating a virtual ground, I tried everything but isolationg the PT.
Now I just fixed an old 50's masco that used a hum pot across the filament windings and tied the wiper to the output tube Cathodes (cathode biased amp).
Think that would help this amp?

[Alex R]

Just to elaborate on what Mike's been saying - a hum that goes away when you touch the chassis, and only fades away when you switch off, has a good chance of being induced from your body picking up atmospheric hum and radiating it to the amp as you bend over it trying to fix it. Just turning the amp over or putting it back in the box can stop this, or standing six feet away from it and using it normally rather than looming over it looking for hum...

[Daver]

Alex R.,
The amp came back to me because of the hum. It hums in the head case, just a little less. Me being near it or standing back from it doesn't change the hum in pitch or volume. I wish it were that easy! It's like a ghost has gotten inside. He hums because he doesn't know the words. Thanks.
Dave

[mfratus2001]

If hum is in the second stage, you need to find out if it is on the input or the output... the grid or the plate.
the grid should be grounded in most guitar amp gain stages. Measure the grid with a DVM and make sure it is at ground. Make sure the cathode is somewhere between 0.8V and 2V. Really, for Fenderish amps, it will be at 1.5V or so.
If the grid is grounded through a 470K resistor, short it to ground, and see if the hum goes away. Monitor the cathode voltage to make sure you aren't changing the bias somehow when you ground the grid.

Come to think of it, what precisely happended to this amp? Why did the user change the transformer? Was there noise before that, or did it appear after the change? What happened just before the noise appeared?

My rule is, "If the findings don't fit the facts, check the facts." In this case, NOTHING seems to be wrong. But yet there IS something wrong. We don't have a complete picture, or there is a wrong fact here.
Have you verified that EVERY resistor is what it should be, and I mean by actually measuring what value it is NOW?
Have you verified that all the solder joints are actually soldered thoroughly?
I am badly nearsighted. To see up close clear enough to really see solder cracks and incomplete wetting of joints, I HAVE to take off my glasses and get real close. Solder breaks and cold solder connections that have failed are darn near impossible to see. They are literally microscopic cracks, and hard to find!

I'm not so worried about caps, but if leaky they will throw voltages off, and that can be measured. Open caps, unless filter caps, won't cause this problem.

The symptom you have seems to be some sort of floating input. But why doesn't shielding help?

I use a metal probe with a clip lead on the other end that is just there to pick up more static.
The probe is covered with heat-shrink tubing, and I use it (instead of my finger) to find sensitive areas and points.
Move it around the area and build up a model of where the hum is picked up.

Another technique is to use a cap on a clip-lead to ground, a small value like 0.1uf will do, but at least a 600V cap... alternately discharge it to ground and then touch areas like the grid, (discharge), the plate (discharge) and so on through the amp. There will be places where the hum is reduced or eliminated.

The rest is just good tech work. You've gotten lots of tips and directions, but something is not just right with this picture. I'm looking forward to finding out what it was.

Mike.

[Daver]

Mike,
Thanks for taking time out of your holiday to help out. The owner changed the OT because he's a big tone hound. He chases every little thing. He's the kind of guy that has special power cords, gold plated fuses, rechargable DC power for his pedal board, etc. He can "hear" everything. I'll reserve my own public comments. I have walked him through, over the phone, swapping out signal caps, changing the type and some values and some other minor tweaks. The best that I can get out of him is that the hum probably started when he disconnected the negative feedback. My conclusion is that the hum was always there to some extent and that removing the negative feedback and maybe some other tweaks revealed it.

I know that grounding the grid makes the hum go away. Gounding the output of the tone stack the stage drives removes most of it. It has to be getting into the tube or being generated in the tube. I have checked every resistor for correct values. I've reflowed all of the solder joints connected to anything involving that stage, including replacing the tubes socket and all the tone pots. I have normal voltages on the plate, grid and cathode. I'll try using the probe idea. I have one made up here with a .047uf 630V cap. I have also removed one end of all coupling and tone caps to check for voltage leaks. I get the same amount of hum if I'm using sheilded wire going to the grid or plain wire. I've taken the day off from this thing and will be back at it on Friday. I'll report what I find. Thanks!

Dave

[mfratus2001]

Some things came to mind as I read that...
If the problem is in the second stage, then removing all the tubes except that one, and running the output of the treble control (which goes to the PI) to a monitor amp should give you the hum, as well.
You've done all this work, and it is not anything you have looked at or replaced so far.

If you can re-create the problem in just the one stage, then that eliminates a whole lot of amp. I still think it can be systemic, where it is ground loops, or bad ground, or something else than what is obvious.

I'm thinking that if it started when he messed with the negative feedback, maybe he did something there. After all, negative feedback is an input, too. Maybe it, or the cap to ground, is not connected any more.

If it is from the tone controls then grounding the grid would not help.

Remember to check resistor values from terminal-to-terminal, not on its leads.

And you said that grounding the output of the tone stack removes most of it... it should remove ALL of it, if that is the lead that goes to the PI and is grounded with a 1Meg resistor. If what you said is true, maybe you don't have just one source of hum.

Isolate... recreate... eliminate possibilities.

By now you could have built a new board to put in there.

Mike

[Daver]

Mike,

I removed all the tubes except for the second stage and the rectifier. I used clip leads to attach the treble control to a guitar cord, grounding the sleeve. I plugged the cord into another amp. The hum is still there. It's weaker, of course, but still there. The bass control still changes the hum volume as before. So, it's in the second stage as I thought from the beginning.

Next I moved the voltage divider off the board and onto a terminal strip right next to the tube socket. The resistors are soldered onto the tube socket directly, shielded cable still feeding the divider from the volume pot. The hum is still there, but somewhat less in volume. I then tried moving the voltage divider ground around to various places and that didn't change a thing. The hum is now NOT affected by me touching the chassis. Maybe we're getting closer.

So, I've replaced the tube, socket, most components around this stage and tone pots. I've moved or rewired grounds, reoriented the socket 90 degrees, checked and reflowed all solder joints and moved components. I've verified voltages and isolated the noise to this second stage. I'm running out of ideas even as it seems the problem is getting closer to a resolution.

Thanks so much for the help, Mike!

Dave

[mfratus2001]

Well, its a good thing, then, that I decided to pick up an amp I had built.
Ya see, I changed the power transformer since it was too small, and it worked just fine.
So I decided to tidy it up, and go ahead and mount the controls on the wooden front panel. The amp is made from terminal strips and sockets, and everything can be done on it with just a screwdriver (see Vintage Amp Repair and Guitar Effects in Houston, Texas) and I was "finished" except for that. When I was done, and slid it into the wooden box, the next chance I had to work on it I found it unusable because of the terrible hum!

Thinking it was a misplaced wire or something, I went to work, tracking it down and moving things around. Here is a list of what I found that eventually eliminated the hum altogether!

Oh, it is laid out as IN>1st Pre>gain control>2nd Pre>tone stack>Phase inverter>push-pull output. Outputs are self-biased. filaments have a virtual ground from 2x 100 Ohm resistors. Supply filtering is minimal, with full-wave center-tapped transformer. No choke.

1. The shielded wire for the gain control, though good quality, was crappy shielding. 4 wires, bare wire ground, and foil shield. Holding your hand on the middle of the wire span got lots of hum, less at the ends. Replaced with two twisted wires from the 1st Pre to the pot, two more from pot to the 2nd Pre. Hum reduced.

2. The grounding was daisy-chained, and I added on terminal strip just for a star-ground. It was not until I broke the common ground for the 1st and 2nd Pre's that a major hum reduction occurred. I had to wire each stage with a separate ground to the star ground. I also cut the wire running across the pots so the gain pot was by itself and the tone pots were together.

3. I rewired the tone controls with twisted pairs, finding out I probably had something mis-wired with all the confusion of grounded shield only on one end... more than one wire had both ends grounded! It should have been only one. They worked much better after that, too. Hum much reduced.

4. I put a shield on the bottom of the wood that all was built on. Ended up not needed when built right.

5. Grounded power transformer, but again not needed.

6. Twisted pair went from treble pot to PI. more Hum reduced. I could then turn the gain up halfway with no input.

Finally,

7. Tried changing the virtual ground on the filaments. Disconnected the junction of the 100 Ohm resistors to ground. Hum increased, of course. I tried a clip lead from there to different grounding points, and they were all pretty much the same. I then ran a clip lead from ground to the filament terminals. One of the pins to ground SILENCED what hum was left!
I could then turn the gain up all the way and still the hum was not annoying. At that level, signal would have brought the neighbors complaining.

I removed, one at a time, other grounds and found what was not needed, and besides the filament, the single major improvement was re-doing the 1st and 2nd Pre's to make them ground apart, instead of daisy-chaining the grounds.

Lessons I learned:

Signal leads need to be twisted-pairs and grounded in the same stage.
Tone carefully. It's confusing enough, so use short jumpers and twisted pairs (grounded only on amp-end, not control-end), grounded only to tone stage.
Short leads may not look nice, but they sound nice. Twisted pairs look nice, though. Twist red and white and it looks like candy canes in yer amp.

Hope my painful experience helps yours.

Mike.

[R.G.]

mfratus2001,
Removing the PI tube stops the hum. Dead silence. I've also tried different output tubes, too.

OK. On some fixed bias amps, hum can come from ripple on the bias supply. That would continue with the PI gone.

The filter for the power tube plates and screens is a 50/50 uf JJ can cap. The negative has a 2" long wire going to a ground lug on one of the PT bolts. The PT center tap is also grounded at that point. The center tap is fused, so the wire from the transformer runs to the back of the chassis to the fuse holder, then to ground. I've also bypassed the fuse holder and attached the CT directly to the negative of the can cap.

That's not optimal, but not automatically a disaster. What I was fishing for was ground ripple induced by capcitor charge/discharge pulses. If, however, your signal ground connects to the chassis somewhere else, it could still be a hum cause.

The ideal setup for this is for the PT center tap to connect (through a fuse, if needed) directly to the (-) of the first filter cap, and nowhere else in the entire amp. That forces the high current pulses to happen through the CT wires only. Any I*R drop in the wire is not connected to ground. The (-) of the filter cap then goes to the signal ground. Only DC currents and signal-frequency artifacts appear on that wire, because the charge pulses don't travel there. They share only the ESR of the cap to cause rectification buzz.
It's best if the (-) of the first filter cap is either the star point for the entire amp, or is not directly grounded at all, but a wire from its (-) goes to the (-) of a second cap, which is then the star point. That ensures that signals don't mix with rectification products in the ESR of the first filter. You want any signal grounding to be a single wire to chassis, either from the filter cap (-) or the input jack, or a separate star point. You should be able to cut one wire to the chassis and have signal ground completely disconnected. That forces the chassis to NOT be a signal or other current carrier and not a hum source therefore.

If I were you, I would take the transformer wire directly back to the cap, as this will eliminate a potential hum source in the future. The output stage cathodes should also have their own wire back to the first filter cap (-) as well, just to eliminate high current paths from signal ground. But this is merely good practice, probably not your main problem.

So - how *exactly* does your signal ground touch chassis? At the PT bolt? or there, plus jacks, plus pots plus terminal strips?

The OT and PT have their lams 90 degrees opposed to one enother. They are separated by about 4". The choke is in between them and oriented in the same plane as the OT. Are you thinking there is power supply hum coming from the output getting amplified by the second stage? Thanks for taking the time to help!

I was actually thinking the PT could be sending hum to the OT directly through M-field. Guitar pickups sense M-fields directly. I've seen one guitar amp that hummed in the speakers when the output tubes were pulled. The PT was sending directly to the OT.

You might want to try the old pull-the-plug trick. When it's operating and humming, pull the AC cord out of the wall. If it's hum pickup from the air, hum will continue. If it's from the AC power coming in the wall cord, it will stop instantly and you'll hear the hum quite in the few seconds while the tubes cool down and DC drops.

[Daver]

R.G,
I attached the PT center tap directly to the filter cap (-). No help. Pulling the power cord stops the hum right away.

The PT center tap, filament center tap, speaker and OT ground, bias supply and power tube cathodes are all grounded to the chassis at one PT bolt. The PI is attatched to its filter cap (-) and a wire also goes to that point.

The two preamp stages have all of their (-) tied to their respective filter caps including pot grounds. A wire from each filter cap (-) runs to chassis ground near the isolated input jack. So, essentially, I have two points where things are grounded to the chassis.

I've used this type of ground scheme before with no problems.
I have tried isolating either ground point from the chassis and running a wire to either point to have a single ground connection, but that didn't change anything. Maybe a buss ground?

Again, this hum seems to be completely surrounding the second stage.

Thanks for helping me out!

Dave