Safe to put V1a coupling capacitor on a switch for variety?

[Mr. Bill]

Thanks Chuck. I have a couple questions for clarity.

1. Can I take figure 1 straight off the screens cap positive end?
2. Is the capacitance of the bypass cap critical to the noise in this instance? Any harm in it being larger than it needs to be?
3. The center tap doesn't need to be an artificial one to be effective, correct? I can use the actual center tap on the 6.3V winding? I don't see why not, but just clarifying since I think I did it wrong the first time.

Figure 2 is more eloquent indeed. I don't have the values for it in house though.

[Enzo]

Don't guess, find out. If you think your switches are picking up hum via the wires, take the switches off the panel, and stick them straight up in the air on their wires, getting them as far from the heater wires as possible. Does this reduce the hum or not. If not, then the switches were not at fault. If it does reduce the hum, then you are correct.

[Chuck H]

1. Can I take figure 1 straight off the screens cap positive end?

Yes.

2. Is the capacitance of the bypass cap critical to the noise in this instance? Any harm in it being larger than it needs to be?

Yes, the cap needs to be large enough to offer a low impedance to 60Hz. No harm in a larger cap to some extent. In fact 22uf is larger than needed. Knowing the actual tech is beyond me, but going too large may cause issues due to charge time or inrush voltage spike on the filaments? Maybe that's not an issue with DC filaments because the caps in those circuits are comparably enormous.

3. The center tap doesn't need to be an artificial one to be effective, correct? I can use the actual center tap on the 6.3V winding?

Correct.

Figure 2 is more eloquent indeed. I don't have the values for it in house though.

NBD really. I did it as per figure 1 in my last build. It only occurred to me later to incorporate the already existing totem. Next time.

[Mr. Bill]

Sorry to double post, but..... Something went amiss the first time I tried elevating the ground reference. I'm not sure what I did wrong, but this time I used the 220k/27k with a 330uf/50volt bypass on the 27k, and it helped A LOT. Not quite as good as the battery heating, but the noise is down from 186mV to 75mV. I'm dancing around the table, ha. That's with the washer and dryer running too, which always adds a little noise in the garage. I'll check it again tonight and see if it's actually quieter than that.


Also, thanks for confirming those question Chuck.

Also, Enzo, I will try doing that later tonight as well. Those 3 switches are fairly heavy, so I'll have to find a way to hoist them up. Maybe one at a time and I can physically hold the knobs.

[Chuck H]

Don't guess, find out. If you think your switches are picking up hum via the wires, take the switches off the panel, and stick them straight up in the air on their wires, getting them as far from the heater wires as possible. Does this reduce the hum or not. If not, then the switches were not at fault. If it does reduce the hum, then you are correct.

THAT^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

[Enzo]

I use a towel. If I want to lift a board from a chassis and power it, I fold a towel under it in the chassis to prevent it touching anything. Your switches? Pull them up while stuffing a towel around them. Or get some string and tie to them and anchor it off something overhead. It isn't permanent, we just need a moment to see.

I also sometimes use a magazine or a newspaper folded. Be creative.

[Mr. Bill]

Well I broke something. Amp has no sound now and I'm getting 400+ volts on all preamp plates.

[Chuck H]

If all your preamp cathodes are tied to the same ground point, it's possible that ground is faulted open so no current is flowing through the tubes. That's about all I can think of that would raise preamp plate voltages that much and shut the signal down.


EDIT: Assuming the tubes are glowing then

[Mr. Bill]

I fixed it, sorry for the false alarm. I had temporarily alligator clipped the beginning of the preamp filament string together since I was going back and forth between the battery and the ac testing noise levels. The clip either came loose or shorted to another pin. Re-soldered in the AC heaters and all is well. Heaters being off would also cause no current to flow through the tubes, correct?

[g1]

Correct.

[Mr. Bill]

Would elevating the heaters cause mains fuse to blow when playing loudly? Here's how I have it tacked in for now.

[Chuck H]

That looks correct. And no, the elevation and the elevation circuit should not cause the fuse to blow. Did you get to play the amp loud before installing the heater elevation? And the fuse didn't blow? My guess would be that excess drive to the power tubes, due to all your extra gain, may be over dissipating the screens. Modern EL34's can be tender about this. On the only two EL34 amps I've built in the last five years BOTH were prone to excess screen current with too much grid drive voltage. I solved the problem in both builds by reducing grid drive and increasing screen grid circuit resistance. It's also possible to increase screen grid impedance (same thing for this purpose) with a higher Henry choke, but that shouldn't be necessary. Since your amp derives most of it's gain from the preamp design it's unnecessary to drive the power tubes deep into clipping. Simply reducing grid drive, but not such that the amp won't produce full power, should be alright. I'm not able to positively identify this as your problem based on the data, but it's a good possibility based on my experience. And most modern high gain preamp design amps DON'T overdrive the power tubes much at all.

[Mr. Bill]

Thanks Chuck. Yes, the circuit hasn't changed in a while besides messing with heaters. It's got hours of playing on the high gain channel. The fuse actually blew while playing loudly on the straight JCM800 circuit settings bypassing the first stage. Hasn't done it again. Maybe vibrations for the heater elevation just being clipped in with gaters. I'll solder it in tonight and try and make it look neater and see if it happens again.

For the sake of learning. Is there a reason you chose 220k/22k in your example? Wouldn't a larger 1mg/100k or some such combination steal less current? Also, where is the line of diminishing returns on elevation? I'm currently at 50V round about, would 60 be quieter? 40?

[Chuck H]

Since the main filament hum problem is with your earliest gain stage you actually only NEED about 15V of elevation to be sure that the filament is positive WRT the cathode. This should keep electrons from going the wrong direction and allowing hum from the filaments to head that way. But since your amp also uses a cathode follower the higher elevation can only help, since cathode followers commonly stress the heater to cathode voltage differential allowance.

I chose 220k/22k arbitrarily because at that resistance it's not taxing the HV supply significantly, they are common values that you are likely to have on hand, they've been used for this circuit by others with good success and, for some reason that MAY have technical bearing I'm not aware of, very high resistances may be of some consequence to a circuit that must both elevate AND decouple AC relative to amplifier signals. I mean, the CT is commonly grounded! How high then should we be comfortable elevating it? Unfortunately I'm not electronically educated enough to answer this one better.

I will say that, sans my knowledge of the circuit beyond rote implementation, 330uf is much larger than typical. If that's causing any problems I can't say. 10uf would reduce hum volume by some 45+dB (that's a lot) so 330uf would be a lot more. It takes time to charge that cap. Some modern tubes "talk" to their filaments more than the golden age tubes did, so it may make a difference. I had one build where the pilot lamp (run off the filament supply) would modulate in brightness with the trem function. There's no explanation for this I can find, but there it is. Maybe the 330uf cap is taxing the circuit and this interaction is causing a problem? Not sure. But 22uf is PLENTY for the elevation circuit.

[g1]

Is there a schematic of this build yet? Is it the HT fuse blowing or mains?
I may be grasping at straws but I don't think I've seen cap bleeders as low as 56K before. Maybe at high volumes you were right at the edge of the fuse value and this extra elevation string made a difference? Not likely but thought I'd throw that out there about the 56K bleeder value anyway.
Edit: digging around a bit, I guess the 56K bleeders were stock in some Marshalls for the screen supply?