AC filament supply hum reduction methods

[Ritchie]

Built the crack as a full preamp and using the same ac filament arrangement. Although it is quiet for ac I was curious if you guys tried balanced center tap with resistors or elevating with dc. I was going to try these methods to see if I could get the filaments even quieter.
If you guys tried this and didn't notice a difference, I probably won't bother.

Thanks,
Ritchie

[Paul Birkeland]

When you elevate with DC, it's generally to keep the heater/cathode voltage within specifications.  Since the 6080 was designed as a voltage regulator, its tolerance to these voltages is incredibly high (+/-200V IIRC), so we optimized the Crack for the 12AU7, and the heather cathode voltage is a paltry 1.5V.

The center tap idea is a good one, give it a shot and let us know if you notice or measure any difference!

[Paul Joppa]

The PT-7, which we keep talking about using for a revised Crack but haven't done yet, does have a CT heater winding.

6080 heatar-cathode voltage rating is +/-300vDC. (Note the much-admired 5998 is only rated +/-100v.) Raining the heater voltage will of course reduce the heater-cathode voltage on the 6080, but in most cases the 12AU7 might be expected to dominate.

The issue with raising the heater voltage relative to the cathode is that is has two competing effects:

Raising the 12AU7 heater voltage will saturate the trace emission of the inside of the cathode sleeve to the (positive) heater, increasing its effective resistance and thus limiting the AC current that can flow between them. 20 to 50 volts is the usual amount. Note that you get the same effect by reducing the impedance from cathode to ground - Crack used an LED which has very low dynamic resistance.

But the heater winding is capacitively coupled to the adjacent high-voltage windings. In order to prevent this from causing even more hum, the raised voltage must be grounded at AC through a large capacitor. If this cap is too small, or has too high an equivalent series resistance, then you wind up introducing more noise than you are killing.

[Ritchie]

Thanks, I'm assuming the cap should be at least 10uf?

[Paul Joppa]

It is often 0.1uF.  I am not aware of any analysis, or any design which claims an analysis, justifying that. That was the value in the original Foreplay, for example. I happily speculate without evidence that 0.1uF is used by everyone because they saw an older circuit that used that value, and the primordial circuit is lost in the mists of time.

Any such analysis would require parameters of the particular transformer and of the particular tube which are never given in specs and rarely even hinted at in texts. The tube parameters (heater-cathode leakage) are undoubtedly nonlinear and a strong function of aging.

That leaves the only practical answer - try it and see! If it helps, then try two different capacitors and see if they give a different result. Then you'll know - and you will be the guru everyone asks so that they don't have to do the work.

[Ritchie]

Thanks Paul,
I have used 0.1 before but may try something larger.
Tried the artificial center tap and it was about the same. I am getting a little more hum out of one channel than the other so it may be something other than the filaments. I am not using any grid stoppers, could this cause a hum issue. I am being picky as I do not hear the hum until I'm about 3-4 feet from speaker.

[Doc B.]

Yes, have you tried different tubes? All bets are off on getting the noise down much lower if you don't have quiet tubes to begin with. Try some different ones if you have not, and when you have a set of tubes you find satisfactory you can go back to hunting down the residual noise. A scope is very helpful for this because you can see small changes that you might not be able to hear, and you can figure out what might be causing the noise by the shape of the waveform.

[Ritchie]

I have only tried two different tubes, are there certain brands known for being quieter than others? I have a scope, I'll give it a try, I'm assuming you're referring to the preamp output?

Thanks

[Doc B.]

If the noise didn't change when you changed tubes that more or less rules them out as a source. And yes, you would want to look at the noise floor by connecting the scope to the ouputs of the Crack. Remember that the peak to peak voltage you will see on a sine wave is actually 2.8 times the rms voltage value. For example if you see a one millivolt peak to peak, that is about .35mV rms.

[Ritchie]

Still have not been able to pin down where the hum is coming from. It is not the filaments and not noticing anything obvious with scope. I'm getting approximately 5mv rms on the output, but seems to be the same for both channels and the hum is only right channel and very slight, but not too the point I can ignore it. I'll try resoldering and maybe alternating things from one channel to the other like the speedball circuits. Just looking for possible tips.
Thanks

[Grainger49]

If you are willing a good way to eliminate AC hum from the heaters is to convert to DC.  It is not that hard.  Bottlehead uses 5 diodes a resistor and one cap.  Crack might not have enough voltage/current to go DC so you might need a transformer for DC heaters.

But first, do you know that the problem is the heaters?  The test is to listen with no music and turn off the Crack.  If the hum goes immediately, before the heaters cool and the high voltage bleeds off, it is the heaters.  If it stays, it is something else.

Just saying.

[Ritchie]

I'll try that. I have tried dc and no difference so I'm thinking it is not heater related.

[Doc B.]

I'm getting approximately 5mv rms on the output, but seems to be the same for both channels and the hum is only right channel and very slight,

I'm sorry, it's not clear exactly what you mean here. 5mV of something that is not hum?

[Ritchie]

Well, I finally found the source of the hum, it was poor wire routing on my behalf, I had the lead from the output cathode to the output capacitor to close to the power transformer. Makes you realize how important layout and wiring can be, my crack is in a completely different chassis as I needed more room to use it as a full preamp, so my layout and wiring is not as tight and direct as the bottlehead chassis.