Demystifying Tube Amps---The Standby Switch

[okcrum]

How about we combine the two and make it a burrito dog?

Damn, I sure know how to torture myself when I'm trying to lose weight!

I just got home from the gig, and I'm gonna fire up my SVT, er, microwave and have a cheese steak burrito! You can leave Philly, but the Philly never leaves you...

Ok, how many of you guys have cooked something on the exhaust manifold on the way to of from a gig? There are sound guy recipes for cookin things on the way to a gig.

Heck, out here, we can just cook 'em up on the fender, at least between spring and fall.

Actually, I have done the manifold cuisine before, but on the way to a gig when I was in college. I'd have stopped at IHOP on the way home, no longer being broke afterwards.

Or are we talking about voltage dividers, and not burritos?

 

[ThisBass]

Once the amp was pushed hard on stage and if the show has finished I give the tubes a chance to slowly cool down for about 5..10 minutes before I switch to stand by and finally power off the amp.
This means that I don't switch to stand by mode between two sets of the show respectively for having a short break.

 

[okcrum]

I guess I should put a disclaimer here for all the traditionalists who look askance at some of my tube posts.

If I were in @agedhorse's position, I'd be designing my amps like he does, to accepted industry specs. But I'm not.

When I worked at pppl.gov, it was our job definition to do things that, on the surface, appeared impossible. Like what?

1) getting 250 kw (pulsed) out of a VA955 TV klystron rated for 50 kw @800 MHz. We used 6 of them @200 kw each in a 1.2MW system. MIT just retired it earlier this year. That system was a gypsy that travelled to several different labs over its lifetime.
2) getting 2.5 MW @55 MHz (pulsed) out of X2159/8974, when the tube was rated for 2 MW max @30MHz max.
3) using X2170/8973s (little brother of 8974) as series regulator elements in 40kv/100A gated power supplies, just like we use gated LM317 regulators

etc. etc...I learned (from the guys who designed these tubes, like Sterling McNees) what you can actually get away with in vacuum tube technology if all the regulatory shackles are removed. It also helped that Eimac & Varian gave us tubes to test to destruction, because they couldn't create the test conditions we could. Most people never get a chance to work in FantasyLand for engineers.

So when I play with a homebrew design like the amp I've described here a bit, I take a different approach than folks familiar only with guitar amps and receiving tubes, or grizzled industry leaders like Andy, who's seen some of my "work".

One last note: Kids, don't try stuff like this at home unless you have a healthy budget for experimentation. And some plexiglass shielding for when/if the thing explodes.

 

[agedhorse]

I do color outside the lines, but am careful to quantify the remaining margins to be sure I don't "cross the line".

Pulsed operation is quite different, even semiconductors have pulsed limits and duty cycle limits that are often greater (and sometimes much greater) than the steady state conditions. If you know your margins, it's all good.

In mass market products, where many thousands are being built and sold, the margin is not a solid line but a shaded line with width. The width represents the statistical uncertainty of those margins due to manufacturing spread, assembly spread and surrounding circuit component tolerances. These can all contribute to uncertainty that even though we (or at least some of us) test for, it's an imperfect science in practice.

This is why I suggest avoiding intentional mis-use as it does increase the risk.

 

[logdrum]

I have a few amps with some relays but they are not time controlled.

As far as the soft start, I don't think you necessarily need a relay. I have a couple of amps with soft start as well. With one, the schematic shows a fairly high value resistor in the bias supply is shorted when the amp is switched from standby to operate. I am not 100% sure because the schematic is hand drawn and pretty rough, but I think the switching is done with a relay. In standby the bias is high enough to cut off tube current. In operate, bias ramps down with the RC time constant of the circuit. So basically standby turns on the DC voltages and simultaneously gently drops the bias.

I have another amp that also does a soft start, but I am not sure this one is intentional. It has a huge amount of resistance in the grid and bias netork (too much). When you switch the amp to operate, the bias slowly drops until it stabilizes. It's possible this amp has a resistor that is brought into the bias circuit as well, but I haven't found it. I don't have a complete schematic this amp as I have only traced out the signal path.

This is the manual: https://mesa-boogie.imgix.net/media/User Manuals/baron manual.pdf

I did not explain but the soft start is on the power on. It's a user initiated soft start before power on. Spring loaded. Helps with other things like preventing the circuit breaker to trip. It does in my old Adobe house.

Standby has 3 positions - off, warm. operate.

The amp can be in standby-warm for days.. I think this standby warm is what amps should have if you are so concerned about cathode stripping and other issues the no-standby crowd are all about.

 

[Wasnex]

This is the manual: https://mesa-boogie.imgix.net/media/User Manuals/baron manual.pdf

I did not explain but the soft start is on the power on. It's a user initiated soft start before power on. Spring loaded. Helps with other things like preventing the circuit breaker to trip. It does in my old Adobe house.

Standby has 3 positions - off, warm. operate.

The amp can be in standby-warm for days.. I think this standby warm is what amps should have if you are so concerned about cathode stripping and other issues the no-standby crowd are all about.

If I am reading the manual correctly, "warm" is not standby. It's basically and application of low operating voltages. So there is no worry of cathode stripping or cathode poisoning when the switch is in this position. Also of interest given my theory on why standby switches are used, the manual says: "Passing through the WARM setting softens the surge of current to the amplifier thus prolonging tube and capacitor life."

FYI early Ampeg SVTs had a Time Delay relay instead of a standby switch. I believe they were found to be unreliable though. @JimmyM did some research on this when he refreshed his 69.

I am not totally convinced that using the standby is not better for long term tube reliability. When the full operating voltages are applied, current flows through the internal tube elements and I would think this would result in rapid change in temperature. When metal changes temperature, it also changes size. So I believe it's actually better to give the tubes a few minutes to warm up rather than switch from standby to operate as soon as the cathode is hot enough to produce emissions. The idea is to let the internal structure of the tube warm up a bit more so there is less thermal shock.

Using a soft start, such as with the Baron, would probably be even more effective at reducing thermal shock in the tubes. But I would think a soft start circuit would add considerable cost and complexity to an amp, and it may not make enough difference to really justify the cost.

I probably should replace the EL34Bs in one of my amps. If I take the amp out of standby too soon, it sometimes makes a very sharp clack instead of just a soft thump. The amp has also pulled the HT fuse a couple of times. So I think the clack may be one or more of the tubes arcing on turn on. It could be some other component of course. Regardless, as long as I let the amp warm up for a few minutes before taking it out of standby, it seems fine.

 

[JimmyM]

If I am reading the manual correctly, "warm" is not standby. It's basically and application of low operating voltages. So there is no worry of cathode stripping or cathode poisoning when the switch is in this position. Also of interest given my theory on why standby switches are used, the manual says: "Passing through the WARM setting softens the surge of current to the amplifier thus prolonging tube and capacitor life."

FYI early Ampeg SVTs had a Time Delay relay instead of a standby switch. I believe they were found to be unreliable though. @JimmyM did some research on this when he refreshed his 69.

Yeah, it was controlled by a 30-second relay tube, and I ended up buying one NOS, but my tech left the shorting plug in the socket despite me wanting the tube installed because if the tube goes ka-flooey, the amp totally shuts down, and they're not as easy to get as they used to be. I couldn't argue.

I am not totally convinced that using the standby is not better for long term tube reliability. When the full operating voltages are applied, current flows through the internal tube elements and I would think this would result in rapid change in temperature. When metal changes temperature, it also changes size. So I believe it's actually better to give the tubes a few minutes to warm up rather than switch from standby to operate as soon as the cathode is hot enough to produce emissions. The idea is to let the internal structure of the tube warm up a bit more so there is less thermal shock.

Using a soft start, such as with the Baron, would probably be even more effective at reducing thermal shock in the tubes. But I would think a soft start circuit would add considerable cost and complexity to an amp, and it may not make enough difference to really justify the cost.

I probably should replace the EL34Bs in one of my amps. If I take the amp out of standby too soon, it sometimes makes a very sharp clack instead of just a soft thump. The amp has also pulled the HT fuse a couple of times. So I think the clack may be one or more of the tubes arcing on turn on. It could be some other component of course. Regardless, as long as I let the amp warm up for a few minutes before taking it out of standby, it seems fine.

Well the point of my post was to show that the standby isn't something to be feared by tube newbs, nor is it mission critical to use it. You may be right, and I'm not qualified to comment. But it sounds to me like the stress on the tubes by not using it on startup is extremely minimal at best. That said, I usually do use it on startup, mostly because I'm not done plugging in stuff when I start it.

 

[okcrum]

I do color outside the lines, but am careful to quantify the remaining margins to be sure I don't "cross the line".

Pulsed operation is quite different, even semiconductors have pulsed limits and duty cycle limits that are often greater (and sometimes much greater) than the steady state conditions. If you know your margins, it's all good.

In mass market products, where many thousands are being built and sold, the margin is not a solid line but a shaded line with width. The width represents the statistical uncertainty of those margins due to manufacturing spread, assembly spread and surrounding circuit component tolerances. These can all contribute to uncertainty that even though we (or at least some of us) test for, it's an imperfect science in practice.

This is why I suggest avoiding intentional mis-use as it does increase the risk.

Agreed, as I've done some production line work, too, both at the job I described, and more conventional outfits.

When you have to build even a short run of 100 units, and if one of them fails, you blow up a physicist, you design conservatively as heck.

You already know rule #1 of production engineering: Anyone can build just one.

FWIW, Varian had no idea how much power their tube could produce, and Eimac had no idea their tube could run at twice the frequency they thought possible. Pulsed operation is indeed different, but wasn't a direct factor for either. It turned out the output coupling structure was the limiting factor for the former, and considering only lumped element circuits was the factor in the latter. We used a large cavity to couple the 55 MHz output from the tube, running it in common (grounded)l grid configuration, to get around the limitations of the tube parasitic elements.

 

[JW56789]

I'd remembered reading this before, a 'white paper' from Hartley Peavey about standby switch why's and how's:

https://peavey.com/support/technotes/hartley/Chapter_6.pdf

 

[agedhorse]

Grounded grid and grounded base configurations are something that's fairly unique to VHF/UHF and higher applications. I tended to limit my work to 20kHz

 

[okcrum]

Grounded grid and grounded base configurations are something that's fairly unique to VHF/UHF and higher applications. I tended to limit my work to 20kHz

Pretty much the only place you see it outside of RF applications is in cascode amplifier stages.

 

[JimmyM]

I'd remembered reading this before, a 'white paper' from Hartley Peavey about standby switch why's and how's:

https://peavey.com/support/technotes/hartley/Chapter_6.pdf

Long way to go for basically the same conclusion but I did enjoy what I understood of the in-depth part of tube construction and application.

 

[okcrum]

Riddle me this: How is it that tube rectifiers, like 5AR4 (indirectly heated cathode), 5U4 (oxide coated filament), etc. are not prone to "cathode stripping"? They have HV applied immediately to the plates, whether or not there's a standby switch. Take a look at a 5E6 Bassman, '57 Twin, or '65 Super reverb schematic. Or a Mesa Dual Rectifier.

 

[JimmyM]

Riddle me this: How is it that tube rectifiers, like 5AR4 (indirectly heated cathode), 5U4 (oxide coated filament), etc. are not prone to "cathode stripping"? They have HV applied immediately to the plates, whether or not there's a standby switch. Take a look at a 5E6 Bassman, '57 Twin, or '65 Super reverb schematic. Or a Mesa Dual Rectifier.

Me, I'm still trying to figure out why the rec tubes in my amps with them run the hottest, yet techs tell me they last by far the longest of any tubes.

 

[Wasnex]

Riddle me this: How is it that tube rectifiers, like 5AR4 (indirectly heated cathode), 5U4 (oxide coated filament), etc. are not prone to "cathode stripping"? They have HV applied immediately to the plates, whether or not there's a standby switch. Take a look at a 5E6 Bassman, '57 Twin, or '65 Super reverb schematic. Or a Mesa Dual Rectifier.

I thought you already explained that?

My understanding was that cathode stripping actually occurs in the tubes we use, but it is so insignificant that it's not really something to worry about. The articles I have read do suggest that it's a bad idea to leave an amp in standby for hours at time, but the 30 seconds or so it takes for a rectifier to start conducting is not an issue. Per the articles, even setting an amp to standby for a 20 minutes break is supposedly nothing to worry about.

But didn't you assert that cathode stripping cannot even occur because the operating voltages in guitar amps are not sufficient?

Still cathode poisoning is a concern we have mentioned, although I don't know how significant it is.

Vintage SVTs do not apply any DC voltages to the tubes before the standby is set to operate, so perhaps cathode poisoning is a concern here. Only the heater transformer is energized in standby mode. The standby switch is on the primary of the main PT, so none of the DC voltages develop until the amp is taken out of standby. So is it a problem to leave a SVT in standby for a long time?

How about amps that vary max power levels by lifting the cathode of some of the power tubes from ground. For example, the Trace Elliot V8 has a half power switch. In half power mode, a switch lifts the cathode on half of the output tubes. There is pair a 47k resistors bypassing the two poles of the switch. Maybe the resistors are intended to produce just enough tube current to prevent cathode poisoning.

For practical purposes the tubes would be biased off, but there will be enough current to power a LED. Each pair of tubes in the V8 is fused at their cathodes. The fuses are bypassed with a (different) 47K resistor in series with a red 3mm LED. So if the fuse blows, the LED lights.

 

[jnewmark]

Not a tube owner, but, wow, 175 posts in a thread about a standby switch ! I am impressed!

 

[Bertr]

Not a tube owner, but, wow, 175 posts in a thread about a standby switch ! I am impressed!

We hope to reach 350 for the thread about the 'On-Off' switch.

 

[beans-on-toast]

Not a tube owner, but, wow, 175 posts in a thread about a standby switch ! I am impressed!

No more tube amps? Say it ain’t so!

You can hear the sound of the room in each and every STAX record – deep and powerful in the bass but raw and almost noisy in the mids.
([Invalid or Expired Link Removed])

 

[jnewmark]

No more tube amps? Say it ain’t so!

You can hear the sound of the room in each and every STAX record – deep and powerful in the bass but raw and almost noisy in the mids.
([Invalid or Expired Link Removed])

View attachment 3640008

No no, just kind of amazed that a stand by switch could cause so much intrigue ! I don't think Cropper/Dunn worried too much about it !

 

[bobyoung53]

When using tube amps, I've always used the standby switch when going on break. It prevents the instrument from spontaneously beginning to feed back. And, as one who is a little compulsive about turning off lights, it saves a little energy.

I always do too, also gives the amp a chance to cool down a little.