NPD: Peavey Valveking 100

[glpg80]

I’ll take a stab at cold cathode stripping and my perspective on standby switches.

Even though filter capacitors have increased their voltage rating to negate the sincere absolute necessity of standby switches, the problem they resolve by having increased voltage rating does not itself actually solve the problem at hand of AC transient surge currents at startup, it just mitigates the catastrophic damage that occurs from it. The surge current demand from the PT to fully charge the filter caps and transient conduction from the power tubes until the grid bias circuit and heater current stabilize, still exists.

At startup the filter and screen capacitors are trying to charge, and voltage across the filter capacitors is zero with current limited only by ESR and the VA rating of the PT. Recall voltage lags current in a capacitor. You then hit depending on the amp either an RC low pass filter filter with a series resistor or an LC low pass filter with a series choke. At startup the choke is not a dead short circuit until the circuit stabilizes, delaying the supply of screen caps needed current so that they can charge to H+.

So let’s summarize so far - we have filter caps acting as a dead short to the transient AC at initial turn on and a choke acting as a time dependent open circuit only slightly conducting as screen caps begin to charge as the B field stabilizes around the choke. Note screen caps are starved of current until the filter caps are charged and voltage across the choke equalizes where the choke is then out of circuit and considered a dead short.

Now going further down the line, heaters take considerably more time to actually heat the cathodes while all of this is happening. Without the cathode sufficiently heated to promote electron flow, you can begin stripping electrons from the coating of the cathode.

So we have filter caps charging, screen caps charging, a choke trying to limit the screen caps and potential tube conduction, all at the same time. This plays hell on the secondary of the PT. The stripping of electrons is a temporary minuscule effect until the heater sufficiently heats the cathode and bias controls the demand of electrons, but the real concern here is the stress a PT has to go through until everything is operating as intended.

Over time the cathode can in fact be damaged from repeat surge turn on effects from cathode stripping previously described, but I argue tube quality is so bad that it’s a third order concern and nothing compared to the transient stressors on a PT at initial turn on - especially with shot filter caps taking longer to charge which puts even further stress on the PT.

So in conclusion standby switches help disconnect tubes completely from the supply until the power supply can get its shit together and stabilize. It also helps remove even more stress at turn on than already exists. I rather help my PT and at the same time not have my expensive glass be electrically connected while all of that is going on at initial startup especially with how much tubes cost now

 

[7704A]

I have more questions, but first, where in the circuit is the standby switch you're talking about located? You mentioned completely disconnecting tubes from the supply, is it breaking the plate and screen supply connections?

 

[7704A]

Added clips to the original post.

 

[glpg80]

I have more questions, but first, where in the circuit is the standby switch you're talking about located? You mentioned completely disconnecting tubes from the supply, is it breaking the plate and screen supply connections?

Sure no problem. Standby locations vary by amp but in my example the standby switch would be after the LC filter. The power tubes would get their screen voltage via triode strapping (so they’d only get filtered H+ and nothing else)

 

[7704A]

Sure no problem. Standby locations vary by amp but in my example the standby switch would be after the LC filter. The power tubes would get their screen voltage via triode strapping (so they’d only get filtered H+ and nothing else)

Got it, thanks. So basically something like this? Tube sketch in the bottom right is to make sure I understand what you mean by triode strapping. Presumably just running the tube like a triode, with the screen tied to the plate?

 

[7704A]

Quick update: Drilled out the screw-holes for the front panel in the head-shell and installed t-nuts. Still need to drill out or drill new holes in the diamond plate though.

 

[7704A]

'nother quick update: drilled the diamond plate for mounting with the t-nuts. Still came out crooked like it was originally, but this was more about ease of service and durability than aesthetics. Worn black bolts on shiny diamond plate ain't a bad look though, I'll have to remember that.

 

[7704A]

Without the cathode sufficiently heated to promote electron flow, you can begin stripping electrons from the coating of the cathode.

Finally getting around to asking the questions I had about stand-by after you confirmed my doodle's setup. First up:

Why are electrons stripped? What's the physics behind that? Also, can you comment on this quote from the RCA Transmitting Tubes Technical Manual No. 4, p65, as referenced by Merlin Blencowe in this article which says: "Voltage should not be applied to the plates or anodes of vacuum, mercury-vapor, or inert-gas rectifier tubes (except receiving types) until the filaments or cathodes have reached normal operating temperature" which seems to imply that per the RCA guitar amps (using receiving tubes) don't need to worry about cathode stripping? I double-checked that the quote is correct using the copy available from http://www.tubebooks.org/tubedata/tt4.pdf (pdf warning), and I've screenshotted the relevant section to save a click:

Granted, it's referring to rectifier tubes, however given our amplifying tubes are kinda(?) rectifiers with more electrodes, I'm not sure it can immediately be ruled out. I've found precious little info on either side of the argument that doesn't require me to take things as articles of faith (even the RCA manual doesn't provide the quantitative physical model(s) behind the stripping, just a qualitative description). I would like to be able to reason from quantitative models of vacuum tubes whether or not cathode stripping is a concern in guitar amps, are you able to point me in the right direction to where I can learn this sort of stuff? Currently I'm working through Theory and Application of Electron Tubes by Reich in hopes of getting a better grounding. Thanks for any info, quantitative or not, that you're able to provide. Even if it's just anecdotal info from working as a tech that no standby correlates to less tube life, the info is valuable to me since I lack that experience.

 

[glpg80]

Funny you quote Blencowe and version 4 of the transmitting tube RCA book. I actually own an original RCA book (not a reissue or reprint) from 1953

Let’s start with the basics. RCA was a manufacturing company back in the day but they never made guitar amps.

The only difference between a diode rectifier and a class A biased triode is a negatively applied gate to control the flow of electrons from cathode to anode.

Electrons are stripped because that is the essence of AC current. Valves are voltage controlled voltage source devices (VCVS). There are also voltage controlled current sources, current controlled voltage sources, and current controlled current sources.

Electrons are heated to excite the valence level of the oxide layer that exists on the cathode. The negatively charged gate inhibits the flow of electrons to the plate by creating an opposition electric field. As the gate is modulated, the voltage is modulated, and thus the electric field on the gate is modulated. This acts as a gate (hence the name) to allow or deny the flow of electrons (negatively charged) to the plate (positively charged). Remove the AC modulation on the gate and it defaults to deny electron current flow from cathode to anode as quiescently designed.

Without proper heating of the cathode, you can physically ruin the cathode faster than with normal use.

 

[7704A]

Thanks for having the patience to start with explaining the basics, I appreciate it.

Funny you quote Blencowe and version 4 of the transmitting tube RCA book. I actually own an original RCA book (not a reissue or reprint) from 1953

(to the tune of JHS' "he has the box") : "He has the book!" Very nifty, a little jealous.

RCA was a manufacturing company back in the day but they never made guitar amps.

Right. But, why does that invalidate the application of the quote I mentioned to guitar amps? I get that guitar amps usually use tubes in a very different way than what the manufacturer intended and so specs kinda go out the window at times, however this wouldn't appear to be one of those cases since it's just to do with flipping the amp/equipment on. Unless I'm missing how the power up sequence of guitar amps is different than the power up sequence RCA was writing in the context of.

The only difference between a diode rectifier and a class A biased triode is a negatively applied gate to control the flow of electrons from cathode to anode.

Right, I follow along with that. The "(?)" in my previous reply was to deal with the fact that I don't know if there is some way additional electrode(s) would prevent the application of the RCA quote to non-rectifier tubes. I'd guess not, but then I've been wrong before...

Electrons are stripped because that is the essence of AC current.

Could you elaborate on that? Electron stripping being the essence of AC is a new one for me. At first glance, it would seem to contradict with the fact that I can have an AC current that passes through a capacitor where (ideally) no charge passes through the space between the plates.

Without proper heating of the cathode, you can physically ruin the cathode faster than with normal use.

I assume once I understand what you mean with respect to electron stripping and AC current, this will be more clear. At the moment I don't see how it follows. Plus, if electron stripping is the essence of AC current, why is it fine to strip electrons after heating the cathode but not before?

 

[glpg80]

Right. But, why does that invalidate the application of the quote I mentioned to guitar amps?

Because it’s historically incorrect. RCA was not a guitar amplifier manufacturer.

I get that guitar amps usually use tubes in a very different way than what the manufacturer intended and so specs kinda go out the window at times, however this wouldn't appear to be one of those cases since it's just to do with flipping the amp/equipment on. Unless I'm missing how the power up sequence of guitar amps is different than the power up sequence RCA was writing in the context of.

Guitar amps are very, very close cousins to AM radios. In fact they are so close that they do in fact act like AM radios simply by accident.

Could you elaborate on that?

I did in detail with the paragraph that followed.

Electron stripping being the essence of AC is a new one for me. At first glance, it would seem to contradict with the fact that I can have an AC current that passes through a capacitor where (ideally) no charge passes through the space between the plates.

Capacitors create an electron potential. AC is allowed to pass because electrons are knocked from one plate to the other when a sinusoidal waveform is present. Capacitors do not allow direct current to pass unless the two plates are shorted which makes it a wire now, not a capacitor.

I assume once I understand what you mean with respect to electron stripping and AC current, this will be more clear. At the moment I don't see how it follows. Plus, if electron stripping is the essence of AC current, why is it fine to strip electrons after heating the cathode but not before?

You should try to re-read what I originally wrote a few more times and think about it. I’ve explained this question in detail originally as well.

 

[7704A]

AC is allowed to pass because electrons are knocked from one plate to the other when a sinusoidal waveform is present.

Quick clarification, is this drawing an accurate representation of what you said?

 

[glpg80]

No

My terminology is too loose so I’ll try to be as precise as possible.

Let’s setup an example where there’s an excess of electrons on one side (Q-) and a lack of electrons on the other (Q+). Let the lack of electrons act as positive, and the abundance of electrons act as negative. When a sinusoidal waveform is presented to the positive terminal, it modulates the E field within the dielectric that’s created by the net difference of charge across the plates. This E field modulation causes a modulation of Q- electrons, thus AC current has just passed from the positive plate to the negative plate. It may seem like electrons physically flow from plate to plate but in actuality there’s no physical electron movement from one plate to the other plate.

When electrons physically flow across capacitor plates, it’s called dielectric breakdown and that’s when a capacitor has physically failed. Usually it’s caused by not having enough space between the plates for the amount of voltage potential between them. (Greater the distance between the plates, the greater the voltage rating, all things else considered)

 

[7704A]

When a sinusoidal waveform is presented to the positive terminal, it modulates the E field within the dielectric that’s created by the net difference of charge across the plates. This E field modulation causes a modulation of Q- electrons, thus AC current has just passed from the positive plate to the negative plate. It may seem like electrons physically flow from plate to plate but in actuality there’s no physical electron movement from one plate to the other plate.

When electrons physically flow across capacitor plates, it’s called dielectric breakdown and that’s when a capacitor has physically failed. Usually it’s caused by not having enough space between the plates for the amount of voltage potential between them. (Greater the distance between the plates, the greater the voltage rating, all things else considered)

Ah! Ok. In that case we're on the same page with respect to AC current passing through capacitors then. Thanks for being more precise. More follow-up questions when I have time.

 

[311splawndude]

No

My terminology is too loose so I’ll try to be as precise as possible.

Let’s setup an example where there’s an excess of electrons on one side (Q-) and a lack of electrons on the other (Q+). Let the lack of electrons act as positive, and the abundance of electrons act as negative. When a sinusoidal waveform is presented to the positive terminal, it modulates the E field within the dielectric that’s created by the net difference of charge across the plates. This E field modulation causes a modulation of Q- electrons, thus AC current has just passed from the positive plate to the negative plate. It may seem like electrons physically flow from plate to plate but in actuality there’s no physical electron movement from one plate to the other plate.

When electrons physically flow across capacitor plates, it’s called dielectric breakdown and that’s when a capacitor has physically failed. Usually it’s caused by not having enough space between the plates for the amount of voltage potential between them. (Greater the distance between the plates, the greater the voltage rating, all things else considered)

I know jack all about what you just said but I think I just learned something

 

[7704A]

Quick update: just placed an order for the parts to make the bias adjustable per tube, with individual current-sense resistors and meter jacks.

 

[7704A]

Quick update: Just installed net 200kΩ 4W flameproof resistors across the first filter caps to bleed off charge when the amp is off. I was looking for 220kΩ or 330kΩ but to get a high enough voltage rating I had to to put two 100kΩ in series. Some quick math shows that 220kΩ is close enough to 200kΩ with respect to power dissipation, time constants, and bleed current that I don't think it will matter, so in they went.

Also, it looks like someone modded the tone stack of the lead channel before I got it. R4 was replaced by a 68kΩ unit. The person didn't want to pull the board, so they just clipped the last one out and tack soldered in the new one. I have parts on hand to almost fully convert the tone stack to the london power standard's values (mid pot is the only one I don't have on hand) so at some point that will get done. I've also identified several places in the circuit I suspect are contributing to what I don't like about the sound, we'll see if I'm correct. In short, there are two interstage attenuators that roll off everything below 329 Hz and and 720 Hz respectively, which seem to suspiciously correspond to my dislike of certain mid characteristics of the amp. That 720 Hz roll-off in particular... it reminds me of tube screamers, and I'm not a fan of tube-screamers. The lead tone kinda reminds me of one too, so that seems like a good candidate for the snippers.

C103/R108 and C106/R112 below are the two filters mentioned. For the first one I'm thinking I'll just remove R108 and short C103. For the second I think I'll just remove C106 so that there is no high frequency emphasis.

 

[7704A]

Speaking of filter caps, at least one appears to be on the fritz. Some mystery white substance like battery acid around the edges of the insulating plastic wrapping on the top of the cap as well as a top that you can feel a noticeable dome on, even if it's hard to see. Another cap has a similar dome without the leakage. A third one is flat as Kansas on top (which is scientifically flatter than a pancake), so I was able to compare the other two to that one to see if they're curved. The curved ones are on the plate node, the flat one is on the screen node. Lifting a leg of R201 and measuring across C220 and C202 only gives about 100uF.

Measuring across C204 gives about 100uF. Removing C202 and measuring across C220 gives about 100uF. Oddly enough, once C202 is off the board it reads 100uF with my DMM. Pop it back in though and the readings of C220 and C202 in parallel stay around 100uF. Strange. I also lifted a leg on the previously installed bleeder resistors across C220 to make sure those weren't mucking with anything when I had C202 out of the board, and the result didn't change. Huh. At any rate C202 is highly suspect now, and I figure it's probably linked to the persistent hum I've been hearing in the amp even on the clean channel, though I haven't scoped it yet. Just placed an order for three new caps from digikey, should be here by Thursday, hopefully with enough time for me to install them before rehearsal.

 

[7704A]

Last thing for tonight: changed the slope resistor in the lead tone stack to 100k, up from 68k, and removed C106 like I mentioned doing a few posts ago. We'll see how that affects things. This board is stupidly easy to lift pads on though. More motivation to spin my own.

Tone stack calculator shows that the previous tone stack had a response like this:

And with the resistor swapped it should be something like this:

 

[7704A]

Another small update: swapped out the low-voltage supply diodes with some 5408G's, as the stock ones are running right at the edge of what they're rated for and seem to have a tendency to fail. Cleaned the preamp tube sockets with some IPA, the push switches with some control cleaner, applied some Dow 748 silicone to the main filter caps after re-seating one closer to the board, and put some spots of silicone on the various connectors I had cut the factory goop off of previously. Also took a sharpie and labeled the mods I did on the board, and noted one thing that should be done later (replace some stock caps with higher voltage ones). With all that done, it'll be passing on to the care of a friend who will gig it much more than I will. Gigging is the motivation for the diode swaps 'n silicone, to make the amp less failure prone. I might get some clips in before handing it off now that I have a decent mic setup, but we'll see.