Odd 60Hz hum out of left channel only

[DerekVa]

Title pretty much sums it up. The hum is present on the left channel, while on the right channel it's pretty much dead silent - the amp in question is my P2P ST35 (Dynaco schematic as opposed to Shannon's schematic). If it were problems in the power supply caps, I'd expect the problem to present itself in both channels. Just to rule out coupling caps, I replaced them (both channels) with NOS K40-Y 0.1uF/1000V caps. No change. The only major physical difference I can see between the two channels is that on the left channel the (highly twisted) heater wires pass over the tube sockets whereas on the right channel, due to the orientation of the sockets, the heater wires are off to the side.

Any ideas (before I rewire the heaters)?

-D

[dcgillespie]

This could only be one of a few possible things: The driver tube (if it's got a heater/cathode short in it), so try swapping those to see if the hum moves with it. The heater winding ground -- if you're using a Dyna transformer with two separate heater windings, is the winding for the noisy channel grounded? If not, that would cause a hum. Finally, lead dress. If the heater leads are routed close to high impedance circuits (grids), then hum can be induced. Does moving the heater wiring of the offending channel change the nature of the hum? If it does, that's a clue that it's the problem.

If the hum is truly 60 Hz (as opposed to 120 Hz), then the issue must be AC power generated -- either from the power cord wiring, general power transformer wiring, or heater wiring. Hum from the B+ would be 120 Hz in nature.

Dave

[DerekVa]

This could only be one of a few possible things: The driver tube (if it's got a heater/cathode short in it), so try swapping those to see if the hum moves with it. The heater winding ground -- if you're using a Dyna transformer with two separate heater windings, is the winding for the noisy channel grounded? If not, that would cause a hum. Finally, lead dress. If the heater leads are routed close to high impedance circuits (grids), then hum can be induced. Does moving the heater wiring of the offending channel change the nature of the hum? If it does, that's a clue that it's the problem.

If the hum is truly 60 Hz (as opposed to 120 Hz), then the issue must be AC power generated -- either from the power cord wiring, general power transformer wiring, or heater wiring. Hum from the B+ would be 120 Hz in nature.

Dave

Hey, Dave! I see we frequent the same web sites. If it's not the Fisher, it's the Dynaco.

Does not follow the driver tube. I verified that it is 60Hz with my Rat Shack SPL Meter. The heater windings tied to ground via 100k resistors (one per lead) and also have 0.1uF snubber caps between the windings. This is identical on both channels. I tried moving the heater wiring but there wasn't a big change, however, I wasn't able to move the wiring where it passes under the tube sockets. I'll reroute that and see if it helps.

Thanks!

-D

[dcgillespie]

Hi Derek --

Yeah, I kind of keep tabs on the Yahoo Forums, AK, and here. Lots of new folks getting into the "old" technology today and are still working up the learning curve. I was there as a young boy, and could only wish I had something like the internet when I had all the same questions so many years ago, so I try to help folks when I can now. I have all the 'bibles" of the industry that I learned from, but the internet is invaluable for those learning today.


So is this a new build that has had this problem from the start, or something that just developed in a unit that has otherwise worked great? That will determine a lot here. If it's new, could it be transformer orientation so that the power transformer is coupling into the left output transformer? I take it the heaters are balanced to ground with 100 ohm resistors (as opposed to 100K resistors, which will not allow a proper reference to ground), although I'm not clear on how the .1 caps are connected into the heater circuit. If that is the general scheme though, then that should not be the problem. A couple of other tests might help narrow things down as well:

1. Does the hum go away with the driver tube is removed? If not, then the problem is centered in the output stage, although that is unlikely.

2. Does the nature of the hum change when the input is shorted to ground? If so, and if this is a new build where the problem has always existed, the problem could possibly be a ground loop, although they are almost invariably a 120 Hz event.

Solving a problem like this starts with getting as many facts as possible, and then using a logical process of elimination to help pin-point the area to target. It will be interesting to know if moving the heater wiring makes a difference.

Dave

[DerekVa]

You know, now that I think about it, I'm not sure if the hum existed all the time or not. I did have this amp hooked up to less sensitive speakers (Axiom M22TI) than the current Heresy IIs. It might have been less noticable simply due to this.

Either way, I rewired the left channel heaters so that the wiring was not crossing directly under the sockets and that has reduced the hum dramatically (although it isn't entirely gone). I also noticed while I was in there that the ground on the left channel RCA was loose (stupid Cardas RCA jacks are too damn hard to solder) so I re-soldered it. At some point I'll redo the right channel heaters as well as even though they are quiet, the heater wiring offends my aesthetics (I wired this amp up originaly a long time ago when I wasn't nearly as anal-retentive).

With regard to tying the heaters to ground - they are currently going through 100 kOhm / 0.5W resistors - is this the correct value? It appears from the Rev. D that Shannon is using 100 Ohm / 2W resistors. I'm wondering if changing the resistance will lower the noise floor?

-D

[dcgillespie]

Hi Derek --

The resistors balancing the heater supply to ground should definitely be 100 OHM resistors - not 100 Kohm. In addition to balancing the supply to ground, these resistors also act to reference the supply to ground as well. 100K resistors are much too high a value to do this, and will definitely increase your hum levels. While the spec is for 3 watt resistors in this location, even 1 watt resistors would only be dissipating only 10% of capacity. Most often, 1/2 watt resistors are speced for this service.

Dave

[Shannon Parks]

Post a couple close up shots if you can. This does seem like one of the heater resistors isn't doing its job. Did you say you had a cap in there too? Maybe remove that.

Shannon

[DerekVa]

Post a couple close up shots if you can. This does seem like one of the heater resistors isn't doing its job. Did you say you had a cap in there too? Maybe remove that.

Shannon

I'll take some shots later this evening. The cap is a 0.1uF snubber across pins 4&5 of each EL84 a la VoltSecond. Sounds like my first problem is I'm using about 1000x too large value-wise heater resistors to tie the heaters to ground. Once I get that addressed, I'll work from there.

-D

[Shannon Parks]

Post a couple close up shots if you can. This does seem like one of the heater resistors isn't doing its job. Did you say you had a cap in there too? Maybe remove that.

Shannon

I'll take some shots later this evening. The cap is a 0.1uF snubber across pins 4&5 of each EL84 a la VoltSecond. Sounds like my first problem is I'm using about 1000x too large value-wise heater resistors to tie the heaters to ground. Once I get that addressed, I'll work from there.

-D

Not so sure about that. It is merely a ground reference resistor, so almost any value can work. This will be an easy and fun one to figure out. :)

Shannon

[dcgillespie]

You and I will politefully have to disagree on this one Shannon. While a 100K resistor might reference ground acceptably for dc purposes, it is much too high an impedance from an AC standpoint for audio work. Consider the MK III and Stereo 70's heater circuit grounding, where they wanted to float the heater supply to the value of the heater cathode leakage within the driver tube to minimize stress on the triode section's heater/cathode insulation, and noise in the pentode section: Using a high value resistor might have accomplished the same thing, but a capacitor was used instead to keep the AC impedance of the ground connection low, while still allowing for the bias to be developed. With the heater circuit closely referenced to ground AC wise, it minimizes any radiated 60 Hz energy in the heater wiring from being picked up as hum.

Dave