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

[Mr. Bill]

That picture is from the "stock 2203 layout" and schematic. They had 56k bleeders. I based it loosely off of that amp with an added gain stage.

[Mr. Bill]

Some updates.

I've managed to get 77mV (down from 185 post elevation) of hum down to 44mV by messing with the routing of the heater wires around V2, the cathode follower. I theorized that moving the negative loop around to the low impedance side of the cathode follower would reduce the noise and it did. I'm going to try rewiring this section as tight as the others and see if there are any further benefits. At this point I find the filament noise totally acceptable, but I love to learn and tinker, plus I'm still obsessing over the battery giving me noise levels of around 11mV. If you don't mind, I would like to ask about a few things that I "think" I understand just to check my logic as I hunt down further noise.

My circuit chain is currently as follows. Keep in mind V0a is unused with the pins tied together.

V0b>V1a> gain knob > V1b> V2a>V2b(cathode follower)> tone stack > master volume >V3a>V3b(PI)> NFB and Output.

So questions...

1. With the gain shut off and the MV at max, logic tells me I should not be hearing the hum induced into V0b or V1a since they are sent to ground by the gain knob being off. Is this correct? My experiments hunting the hum would seem to back his up. Pulling the preamp tubes with a meter hooked to the output shows that pulling V0 made no difference in hum levels; pulling V1 brought the level down from 77mV to 66mV ( a little change since half of that tube is not sent to ground ); pulling V2 drastically dropped the noise down to 10mV, and pulling V3 further reduced that small amount to 3mV. This is what prompted me to attack the cathode follower to try and reduce noise first.

2. As I understand it, running high current out of phase wires close together focuses the fields emitted by the current into each other, sort of trapping it as they approach null. My question is, does twisting in phase high current wires have the opposite effect? Does it strengthen the emission field or make no difference at all?

3. I asked once already, but I'm not sure if anyone noticed because it didn't seem relevant. Going back to question 1 about V0 and V1a being shunted to ground with the gain knob off..... I have noticed that probing the voltage on V1a's plate and coupling cap on either side causes a very loud hum to come from he amp. How is this possible if that part of the circuit is already shunted to ground by the gain knob? I've noticed this on a few builds, not just this one, and the logic of why has always escaped me.

4. Can heating an unused triode, like the V0a in my amp, induce noise in V0b or is that filament contained to that half of the tube with no effect at all on noise levels?

5. Last one. Noticing that the cathode follower seems to be my prime inducer of hum in the circuit has me slightly baffled. I'm sure I'm missing the obvious, but as I understand it a cathode follower has theoretical gain that approaches 1, but does not exceed it. Since there is no "amplification" taking place there, how can it induce so much noise?

I have learned more in the past few months here than I have in forever. I sincerely thank you guys for all your help in my journey to better understand these things.


Edit: Thinking harder about #5, I believe it is because both triodes in a cathode follower are in phase, so you get no help from cancellations.

[Chuck H]

1. With the gain shut off and the MV at max, logic tells me I should not be hearing the hum induced into V0b or V1a since they are sent to ground by the gain knob being off. Is this correct? My experiments hunting the hum would seem to back his up. Pulling the preamp tubes with a meter hooked to the output shows that pulling V0 made no difference in hum levels; pulling V1 brought the level down from 77mV to 66mV ( a little change since half of that tube is not sent to ground ); pulling V2 drastically dropped the noise down to 10mV, and pulling V3 further reduced that small amount to 3mV. This is what prompted me to attack the cathode follower to try and reduce noise first.

You are correct-ish. If the hum is due to grounding and the first or second stage have any circuit grounded where there is heavy current fluctuation (near, shared or in between power amp or power supply grounds) this can be read as AC on the signal chain. Even though it's coming straight from ground! The first couple of stages in a cascade amp are EXTREMELY sensitive. Since you're chasing filament hum right now, this may not be your problem. As your testing and reasoning has proved out. Since much of the hum left seems now to be with the cathode follower I would suggest trying yet another tube in that position. And NOT a Rusky made tube. Some of them are especially sensitive to the filament/cathode voltage differential and have been known to fail as cathode followers. Some reports say the manufacturers have corrected for this, other reports say they haven't. Why take a chance? When I build an amp I order five preamp tubes for every three I need. I'm buying from vendors, not bulk in any way. I do this because I have, indeed, had two noisy tubes, new out of the box in a purchase of five. This is a relatively recent phenomenon that is the new "normal".

2. As I understand it, running high current out of phase wires close together focuses the fields emitted by the current into each other, sort of trapping it as they approach null. My question is, does twisting in phase high current wires have the opposite effect? Does it strengthen the emission field or make no difference at all?

Interaction is interaction. Proximity and current determine the amount. The result is (at least partly) determined by phase. Two identical, out of phase currents will cancel. And two like phase currents will feed back. That is, other limiting factors not withstanding they will amplify each other indefinitely until the universe is consumed in a seething plasma. Fortunately there are always limiting factors Determining phase is sometimes tricky though because it changes with frequency and shifts are complicated and exacerbated by clipping. It's not entirely a crap shoot though. You can usually count on two concurrent gain stages being sufficiently out of phase and this can help with determining safe lead dress and grounding for high gain circuits.

3. I asked once already, but I'm not sure if anyone noticed because it didn't seem relevant. Going back to question 1 about V0 and V1a being shunted to ground with the gain knob off..... I have noticed that probing the voltage on V1a's plate and coupling cap on either side causes a very loud hum to come from he amp. How is this possible if that part of the circuit is already shunted to ground by the gain knob? I've noticed this on a few builds, not just this one, and the logic of why has always escaped me.

Touching tube pins with your meter does two things. It adds a significant length of unshielded lead hanging off the pin and it impresses whatever circumstances your meter uses for testing into the circuit. It's common to hear SOMETHING. Sometimes a just pop, sometimes hum or buzz. It's possible that there is a fault in your decoupling or grounding if probing the plate prior to a volume control is causing significant noise. Hard to know though. Like I said, it's common enough that probably everyone here (including me) read that and figured "Yeah, well..." And went on, distracted, to answer about another question that might have seemed more significant.

4. Can heating an unused triode, like the V0a in my amp, induce noise in V0b or is that filament contained to that half of the tube with no effect at all on noise levels?

Many amps incorporate an unused triode in their design. Typically the pins are left floating. I'm not even sure what tying them together would do. Nothing I would think since there is no voltage on any of the elements.

5. Last one. Noticing that the cathode follower seems to be my prime inducer of hum in the circuit has me slightly baffled. I'm sure I'm missing the obvious, but as I understand it a cathode follower has theoretical gain that approaches 1, but does not exceed it. Since there is no "amplification" taking place there, how can it induce so much noise?

Again... Grounding could be an issue. If any part of the CF or it's buddy triode is grounded with, near, or between high current grounds this can inject fluctuation in current as AC into the signal path. And read above where I mention noisy tubes too. And since the CF is direct coupled it's possible that differing ground points for any circuit on either triode could form a loop. So maybe check for that. Use individual leads for each ground associated with circuits from either triode (rather than daisy chain) and run them to the same ground point nearer the preamp end of the buss.

[Mr. Bill]

Thanks again Chuck!

27mV and counting. Tried 3 different tubes in position. All were within a mV or 3 of each other. Got reduction further reduction by other means of trial and error.

[Mr. Bill]

I think I'm getting excess noise with the gain up because of how I have my switches wired. From cathode, all RC loops are wired in and getting voltage and the ground is switched. It was easier to wire this way, but I'm pretty sure I'm creating little antennas on every tube. If I get time tonight I'm going to change that and see if it helps with the noise.

[eschertron]

If I was a betting man, I'd "bet a packzi to a donut" that the low-ish impedance at the cathode means less hum introduced there than you imagine.

[Mr. Bill]

I broke 1 of the 9 loop and it brought the noise down to 19mV from 27mV..... That was the only one that was an easy break, but yeah, looks like I need to rewire these switches like I should have in the first place.

Edit..Inverting all switches made it worse. But only by a tiny amount. However, it brought the hiss down so it's still a net gain in reduction.

What's a good target level in mV at the tap with both gain and volume maxed on a high gain 4 stage like this?

[Chuck H]

However, it brought the hiss down so it's still a net gain in reduction.

This seems odd. The impedance at that circuit is very low and hiss is typically a byproduct of gain relative to Johnson noise (thermal noise) which is a byproduct of circuit resistance. I can't think of anything about the cathode circuits that should change the hiss level noticeably. I fear something you've done has reduced gain or your testing was done in two different gain modes. That is, I don't think you've actually lowered hiss unless you've lowered gain because that's impossible (though there are exceptions that don't apply here).

[Chuck H]

What's a good target level in mV at the tap with both gain and volume maxed on a high gain 4 stage like this?

They're all different due to grounding scheme limitations. And how can differentiate between hum and hiss just reading AC noise on the output? I can only tell you that with that many stages maxed there should be so much hiss that hum is a moot point That's not necessarily a bad thing, it's just the brakes. I have my own amp open on the bench right now, but it's not a high gain preamp type amp. It's an overdrive the piss out of the power tubes amp and it's only 17 watts. Still... Two stages into a third that boosts to OD the PI that is already ODing the power tubes, so, five stages with the last three clipping. I'll go take a measurement and we can adjust for wattage Be back in a bit.

[Chuck H]

I have 46mV of hash on my output with the tone stack pots at 5, presence 0 and the volume dimed. Note the presence setting!!! And that 46mV is for a 17W amp. I consider this entirely acceptible. Of course, since I'm dimed and I have no no master volume this is measured with the power amp wide open. Not sure if you're testing like this. At these settings my amps gain is in pseudo metal territory. Probably not dissimilar from yours.

[Mr. Bill]

With presence and all tone controls at noon with both gain and volume dimed I'm getting 134mV at 100 watts.

Will draw schematic to help investigate how hiss was reduced. Total noise came down by about 10mV.

[Chuck H]

We're actually ham handing this. That mV indication is meaningless without knowing the load, so... I'm using an 8 ohm load. So that's .000265-ish watts of noise And by your numbers, unless you're using a very low load, I would think you've done better considering it's a 100W amp.

[Mr. Bill]

8 ohms That rogue wire trick of yours shaved off another 10 or so!

Here's how the switches were wired before.



Side note. Since elevating the filaments, the amp makes a couple quiet pops while warming up, then never again. Is this maybe something to do with the elevation? Perhaps my larger than needed cap value?

[Chuck H]

Perhaps my larger than needed cap value?

Perhaps. If noise goes up when you lower it to 22uf, just bump it back up to 100uf (it'll be as quiet at 330uf fo reals) and since there's no significant current through the circuit regularly it shouldn't matter at all when using the amp other than those little noises at startup.

[Mr. Bill]

Popping was cathode follower slowly frying a tube that was good a few days ago. It was getting worse and I tracked it down. Hum with gain off / volume max is now down to 10mV with a gain max / vol max noise level of ~129mV. I kinda think I'm reaching an apex here of as good as it's going to get.