DC Heater Question

[Cameron]

Maybe this will work better... I agree, the lead dress could be much better on the filament wires.

Agree also ....maybe clean up the lead dress all around now that you know where you want things to go. You may find it doing this or at least you will be able to see the amp more clearly to find the problem.

[bal704]

Here's a pic of the heater wiring before I added all the other circuitry:

http://s755.photobucket.com/user/bal704 ... sort=3&o=3

I only had a few different colors, so I had to use the yellow and green for other purposes as well.

[Tillydog]

Here's a pic of the heater wiring before I added all the other circuitry

My comments stand...

[martin manning]

The heater leads should be twisted uniformly, including the sections where one of the pair goes to and from the main run. I circled two places where it looks like you have grid wires running close and parallel to heater wires. There may be more, and if you can isolate a stage that has the problem, that will help you find the place(s) where the hum is getting into the audio.

In your layout the sockets are close to the preamp tag board, so with the way you have run the heater wires they loop underneath. I think I would have oriented the 12AX_ sockets with pins 1 and 9 at the back and run the heater leads to pins 4 and 5 across the sockets. This would keep the heater wires on one side of the preamp sockets and away from the leads going to the board.The EF86 filament connections are at pins 4 and 5, so it is fine as-is.

[soma_hero]

Actually, I guess it has to be 60Hz since powering the filaments with a battery killed it. Schematic?

not true. 120hz from the HT diode transient can couple into the heater winding if they are on the same PT. That can then be amplified by the tubes.

[martin manning]

Yes since 120Hz current pulses are always present on the HT winding there will always be some 120Hz content on the filament winding. Being 2x the frequency, it will flatten both the peaks and troughs of the filament voltage equally, distorting the 60Hz wave. But it was probably far from clean anyway!

[bal704]

I recieved a 500 ohm pot today, and hooked up a 'humdinger'. I was able to dial out about 98% of the hum. I hooked up the center pot to ground and the Power tube cathode to see which was better. It's pretty close, but I think the ground may get rid of slightly more hum.

What exactly is this miracle device, er humdinger, doing?

[martin manning]

The pot provides a ground reference for the filament circuit which is adjustable from one leg to the other. I would not have expected a huge difference between an actual CT and the pot so that is surprising that you have almost eliminated it. Where is the pot set? Close to center or not?

[bal704]

It's about halfway between middle and one of the legs.

[soma_hero]

There are symmetrical and non-symmetrical sources of interference that the heaters can introduce/pick up. The humdinger allows you to adjust the reference of the heaters to ground to help eliminate the non-symmetrical sources that a symmetrical center tap can't take care off.

i just made an amp with an 12ax7 and 6sn7 in the preamp. You could humding so that one tube was happy but then the other tube would hum and vice-versa. Different heater filaments having varying sensitivity to noise sources.

[martin manning]

In this case I think what is happening is that the heater noise is coupled into the circuit asymmetrically, most likely through lead placement, and the off-center tap enables canceling it out.

[joesatch]

Find some space in your chassis to bolt one of these in. Power your 12ax7's at 12v, and never need to think about it again.

http://www.alliedelec.com/search/produc ... U=70069560

I'm very interested in this supply. I can i run all 12VDC heaters (pre and power tubes) with this unit? I'm looking at 2xEL34 and 3x12ax7. Would that work?

[martin manning]

No. The two EL34's (in series) require 1.5A, and then you need 0.45A for the three 12AX7's. 1.9A at 12V total. The linked supply is rated at 1.3A