Question: Amp Damage from Bad Tube

[Zelmo]

In a related thread (http://www.thegearpage.net/board/showthread.php?t=202789) a question was asked about how to repair a resistor that blew because of a bad 6V6. Rather than hijack, I'll ask a related question here.

I had a similar experience not long ago where I bought some 6V6s from a reliable source and one of them took out some resistors in my amp (don't ask me which ones - I'm trying to educate myself, but I'm not there yet. I had it fixed by someone who knew what was going on).

The question is: Shouldn't this kind of thing be avoided if the tubes are burned in, or is burning in not something that everyone does? Again, I'm talking about reputable dealers here, not EBay potluck.

Thanks.

[jezzzz2003]

I havent read the thread but I take it youre reffering to screen grid resistors, these are designed to take the heat when an output tube burns out instead of more expensive and hard to replace parts.
they should always be checked and measured when replacing tubes that have blown.

[John Phillips]

No, it can't be avoided just by testing or burning-in tubes - tubes simply fail sometimes, even after having been thoroughly tested. Sometimes they do so in a way that puts a lot of stress on some components in the amp.

IMO the fault is with the amp designer, builder or whoever repaired it last. Components that are known to be stressed by tube failures need to be over-spec'ed enough to withstand the failure, for long enough for the fuse to blow. If that's not possible with a single primary fuse (the way many amps are still done) then you need to fit more fusing. IMO a properly designed and built amp should never blow anything other than tubes and fuses.

There is an idea that over-spec'ing resistors can lead to other parts failing instead (like transformers), so therefore you should use an under-rated resistor to act as a sort of fuse - but this is very bad practice because firstly resistors are not designed to fail safely like fuses and can actually burn, which is obviously not a good idea inside an amp; and secondly because it makes the amp unrepairable without a trip to the bench. If only tubes and fuses blow, you can get it going again on stage if you have to. If the transformers aren't capable of taking the current until the fuse blows, you need to fit more fusing, not use resistors as a substitute, IMO.

[Zelmo]

Interesting. Screen grid resistors...now I recall. Well, they did their job because I think that's all that blew.

I hear you about possibly over-speccing these things in the design. Suffice it to say that this was a tweed amp that was designed and built close to the original design, so in this case, the amp is not a 'modern design'.

Just like to say that I really appreciate all the info around here. I'm working my way through Dave Hunter's book right now and I'm beginning to learn something. I'd soon like to build a 5E3 kit. In the meantime, I'm sucking up all the info I can get.

[Blue Strat]

What John Philips said, plus the tubes have to be biased correctly in your amp. And don't assume that because an amp claims to be "self biasing" that the tubes are biased correctly. You have to check to be sure. Your equipment is your responsibility.

[unklmickey]

...And don't assume that because an amp claims to be "self biasing" that the tubes are biased correctly. ...

hi Blue Strat,

if an amp is cathode biased, it truly IS self biased.

it will tolerate wide variations in tube characteristics, and still idle near the current the circuit was designed for.

of course, you don't get somethin' fer nuthin'......the maximum power available is decreased by this configuration.


cheers,

unk

[Blue Strat]

hi Blue Strat,

if an amp is cathode biased, it truly IS self biased.

it will tolerate wide variations in tube characteristics, and still idle near the current the circuit was designed for.

of course, you don't get somethin' fer nuthin'......the maximum power available is decreased by this configuration.


cheers,

unk

In theory yes, in reality, not quite.

Check some Fender 5E3 Deluxes and clones or almost any cathode biased EL84 amp. They all run their tubes over 100% of the specified maximum idle power dissipation of the tubes chosen. Tubes that draw higher current will operate even further out of range.

[unklmickey]

In theory yes, in reality, not quite.

Check some Fender 5E3 Deluxes and clones or almost any cathode biased EL84 amp. They all run their tubes over 100% of the specified maximum idle power dissipation of the tubes chosen. Tubes that draw higher current will operate even further out of range.


i think "not quite" is a good choice of words here.

remember, tubes that draw more current with the same voltage between cathode and control grid, will only draw slightly more current in this configuration, because as the current increases, the voltage between the cathode and control grid also increases. so the voltage between the cathode and control grid is not the same.

AND, the power is not the plate voltage times current (as it is in a grounded cathode configuration).

it's ( plate voltage - cathode voltage ) times current.


so it isn't nearly the problem it might appear to be, on the surface. :AOK

[Zelmo]

Your equipment is your responsibility.

Thanks for the advice. I really appreciate it and I'll certainly keep that in mind.

[John Phillips]

i think "not quite" is a good choice of words here.

remember, tubes that draw more current with the same voltage between cathode and control grid, will only draw slightly more current in this configuration, because as the current increases, the voltage between the cathode and control grid also increases. so the voltage between the cathode and control grid is not the same.

AND, the power is not the plate voltage times current (as it is in a grounded cathode configuration).

it's ( plate voltage - cathode voltage ) times current.

so it isn't nearly the problem it might appear to be, on the surface.
It often is, because even taking those things into account, many cathode-biased amps are still intentionally run too hot, and without changing the cathode resistor there's nothing you can do about it - exactly because it's self-adjusting to the same point for almost all tubes. So it's actually more 'fixed' than even a non-adjustable fixed-bias amp.

I don't know whether it's because of the false belief that cathode bias is the same thing as Class A, and so the tubes can (or even "should") be run at 100% at idle, or what, but many amps do run the tubes this hot, and so - because they're actually Class AB - a lot hotter at high power. IMO this is also a design/build fault, and if I'm working on an amp like this which is going through tubes faster than the owner is happy with, I don't hesitate to raise the resistor value.

[Blue Strat]

i think "not quite" is a good choice of words here.

remember, tubes that draw more current with the same voltage between cathode and control grid, will only draw slightly more current in this configuration, because as the current increases, the voltage between the cathode and control grid also increases. so the voltage between the cathode and control grid is not the same.

AND, the power is not the plate voltage times current (as it is in a grounded cathode configuration).

it's ( plate voltage - cathode voltage ) times current.


so it isn't nearly the problem it might appear to be, on the surface. :AOK

Correct about idle dissipation ( plate voltage - cathode voltage ). Problem is, you're still relying too much on theory and not actual amps. Check some of the amps I mentioned and you'll note that the max idle dissipation spec is violated more often than not.

[unklmickey]

Correct about idle dissipation ( plate voltage - cathode voltage ). Problem is, you're still relying too much on theory and not actual amps. Check some of the amps I mentioned and you'll note that the max idle dissipation spec is violated more often than not.


although this is going a bit away from the issue we were discussing -- the effect of tube variations (now were talking about design center values), i still think that would be worth looking at.

other than the 5E3, you generalized nearly all cathode biased EL84 amp as being offenders. care to cite a few specific examples?


cheers,

unk

[Blue Strat]

although this is going a bit away from the issue we were discussing -- the effect of tube variations (now were talking about design center values), i still think that would be worth looking at.

other than the 5E3, you generalized nearly all cathode biased EL84 amp as being offenders. care to cite a few specific examples?


cheers,

unk

AC15, AC30, Matchless, Marshall 18 watt, Marshall 20 watt, etc...