Standby switching and amp football

[BackwardsBoB]

I've been sticking JJ's 6V6S tubes in several of my amps, and an old Carvin X30 sounds like angels with them, right up until a tube fails. The screen resistors are very low value (maybe 10 Ohms), and beefing them up to 1K should help, and B+ is around 460V, but I'm noting a current surge through the power tubes when I take the amp off standby, like some blocking cap is passing a 200V kick, and it sticks around for a while at the power tube grids. The failure seem to occur on turn-on or turn-off. The amp runs well when it runs. I see the failure coming out of standby or shortly thereafter.

This glitch looks hard to avoid. What's the downside to simply floating the cathodes of the power tubes in standby instead of exposing tubes that are all heated up and ready to go to to a HV turn-on transient?

[Enzo]

Explore what is happening. Got a scope?

What level of idle current is flowing?

What does the bias voltage do as the transistion occurs? Is it slow to arrive?

What happens at the screens?

[BackwardsBoB]

I must look into the problem further (I've got maybe 18 of these x-series amps), and I have a scope, but it's alot like destructive testing. I'm running the tubes at 20mA idle, and the bias is all set up with with a seperate PT winding when the B+ comes up.

I think that the best way to proceed is to beef up the screen resistance first to harden the amp a bit, and then track down the surge. It may save me a few tubes.

But still, can anyone think of a down side to floating the power tube cathodes in standby instead of switching the B+? I've found that they go up to about +15V on one amp I measured it on, which uses the technique to turn off half the tubes and halve the ouput power, and a zener across the switch could insure that it stays in that reasonable range. Another way to go would be to disconnect the ground leg of the bias divider.

[Enzo]

No, I see no problem, some commercial amps use that method.

You could do some of the testing I proposed with the power tubes removed.

[Merlinb]

but I'm noting a current surge through the power tubes when I take the amp off standby,

Perhaps the standby is arranged so that it also kills the bias, so that the bias has to climb up (ok down!) when you come off standby? That is no damaging to the tubes per se, unless you have insufficient screen stoppers (but you spotted that already).

like some blocking cap is passing a 200V kick,

When the B+ comes up tot full voltage the capling caps have to charge, which will cause a positive pulse to appear on the PA grids. Again, this is not damaging except to the screen grids- you definately need those 1k stoppers!

What's the downside to simply floating the cathodes of the power tubes in standby instead of exposing tubes that are all heated up and ready to go to to a HV turn-on transient?

The turn on transient is not damaging in itself, it is simply that the screens are getting red hot for a moment. Floating the cathodes will lead to cathode poisoning, which will reduce the useful lifespan of the valves. However, most people replace their valves at least every couple of years anyway, so you probably wouldn't notice the shortened life anyway. (We're talking like a lifespan of 10000 hours instead of 20000 hours- it's a long time either way!)


However, since your problem is the screens you may like to look into converting to a screen standby instead. The plates can be permenantly powered up, while the standby switch would be in parallel with a very high value resistor in series with the screen supply. Closing the switch would short the resistor, for normal operation.

[BackwardsBoB]

Thanks, folks. Merlinb is confirming my fears and pointing toward a solution.

The x-series from Carvin went through various changes in the decade that it was made, but in many ways, it stayed the same. One obvious change is a frequent increase in the value and power handling of the screen resistors, and the power rating of the resistor in the screen voltage filter. The amp I'm having a problem with is very early in the evolution. I have two later grey fuzzy models that don't share the problem, and sound really great - warm, compressed and chimey - with 6V6s. Several of the older x-series amps I bought had damaged screen supply filter resistors.

I took a pair of 6V6s out of a newer x-series amp I have. The tubes had several hours of use on them, sometime at very high output. I put them into the older amp and biased them, and one tube developed a screen short on maybe the fourth power-cycle.

[Steve Conner]

Can you see the screen grids of your 6V6s flash red or orange like lamp filaments when you power up? Even with tubes that have black coatings, you can look in the hole in the top.

Is it always the same tube that fails? Do both tubes flash at power-up or only one?

If they don't flash at power-up, do they light up when the amp is cranked?

I experimented with 6V6s in my Ninja-Deluxe amp and saw the screen grids glowing red at high power, however they didn't flash at power-up. I was using about 420V B+ and 1.5k of screen resistance, a shared 1k plus a 470 per tube. With 10 ohms I'm not surprised they disintegrate.

If it is a transient caused by the DC block capacitors charging up, then you can mitigate it by slowing the rate of rise of B+ to the PI stage at power-up. For example by using a bigger decoupling cap or a larger dropping resistor. (While we're brainstorming, is the existing filter cap maybe dried out?)

You could avoid it altogether by having the PI powered in standby, so that when the switch is flipped, the DC block caps are already charged.