Burning 6L6 screen resistors

[EETStudent]

I've got a power amp running a quad of 6L6s, and one of these tubes/sockets has burned a couple 470 ohm resistors. The first of the sockets in line with the B+ is the only one that is giving me trouble. Is it because its the first one there at the B+?

[Enzo]

More likely it is the one with the bad tube in it.

The only thing that will burn out a screen resistor, or any other resistor, is too much current through it. And pretty much the only circuit path through that resistor is the screen of the tube. Certainly try a new set of power tubes in the amp - or at least a known good set from some other amp.


It is possible, though far less likely, that the downstream end of that resistor is arcing to ground at the socket. And I suppose I can;t rule out the wiring at the socket on pin 4 having a stray strand touching something.

[KevinOConnor]

Hi Guys

The socket should be replaced if it is charred in any way. Carbonization makes the socket conductive instead of insulative.

Apart from the that, the 470-1W carbon resistors were an economic choice that does not benefit the circuit very much. These are too low a value to fully protect the tube, as TUT points out, and should be 1k-5W each. Going to a 5W part assures you get a flame-proof component. See the other threads here where I have posted additional info regarding screen resistors.

Have fun
Kevin O'Connor

[Prairie Dawg]

Not sure about that, boss. A 470 ohm 2w carbon comp functions as a fuse of last resort. Not a bad thing considering.

[Chuck H]

And the OP never said it was a 1W resistor or even said what the amp was. So there's no way to determine the rating of that resistor as yet. And AFAIK there are other criteria for the actual resistor value (470 vs 1k) so how is 1k automatically best? Without knowing the voltages how can we know how much we want the screen to dissapate?

It's probably a bad tube or possibly a conductive socket (sometimes a socket can become conductive after such a failure). Probably a bad tube. You could check the bias condition of each tube. It may reveal a fault unless the tube still idles OK. If you don't have another set of tubes you could swap the two outers and see if the problem follows the tube. Of course doing it this way means another resistor is going to burn one way or another. As evidenced by the failure I'll back the 'screen resistor as fuse' observation too. Do we really need a more costly failure due to an over rated screen R? I'd rather replace a screen resistor than a power supply, socket, PT or just about anything else. That resistor blowing is usually indicative of a tube fault so it's cheap insurance and a good indicator not to run the screen R's at too high a rating. JMHO

[KevinOConnor]

Hi Guys

See the other thread for why "1k is best" - and for certain tubes, higher than 1k is better.

Generally, if it is 470R then it is likely a Fender amp or a copy. Mesa copied Fender's mistake and everyone who followed did so as well. The 470R became less damaging when safety approval processing required the use of a flame-proof part. At least collateral damage was reduced, but the tube is still better protected by 1k. More important - the OT and PT are better protected.

Carbon should never be used for power resistors. They may ultimately act like a fuse but only after they have burst into flames and caused more damage - and allowed more damage - than a fuse would have, or than the proper value of resistance would have.

You can choose to remain faithful to circuit values, or you can learn how the circuit operates and what makes the magic happen - and the flames from not happening.

Have fun
Kevin O'Connor

[Chuck H]

Thanks Kevin. Not that I have the EE to challenge your POV but I did qualify that operating conditions may be relevant. If your not running excessive voltages and you don't want the screens to sag much, what's wrong with 470 ohms. Many popular tube circuits (that run conservative voltages) don't even have screen grid resistors. There is a tonal difference so as long the right criteria are met...

I agree a fuse is better than using a resistor rating as a fuse. But if we tack every peripheral circuit in place that seems like a good idea that makes for it's own complications. I don't know why a blown screen resistor is more damaging to the rest of the amp than a failing screen through an over rated resistor. Not calling you out, I actually don't know what the reasons are. Unless your point is that we can prevent all screens from failing by using the 1k value? And therefore screen failures becomes a non issue?

I don't think the 470R typical value is a remnant left over from Fender designs, per se'. How the screen behaves greatly affects the response of the amp and more resistance means a different sound. It's all about the sound, right? I read somewhere that the lower value is typically best for beam tubes like the 6l6 because it allows the screen to track more accurately. That a 1k value reduces power and dynamics with 6L6's. To your credit many designers that seem particularly hedged on reliability seem to choose 1k as a blanket value for all tube types. But like you, most don't say specifically why.

I'm just saying that as long as the right criteria are met, like not overdissapating a tube in the first place, why not choose the value that sounds best.

As for knowing the circuits and finding out where the "magic" comes from... Well, there's room for gray area and opinion there. If you've spent any time at this forum you know that we don't generally go for mojo and tone lore like they do at other forums. While some of us are just 'guys building amps' and not EE's, some of us are EE's (though not myself). Certainly a lot of the lore floating around about what makes an amp sound good is pure BS. But your blanket criteria may not be the ONLY answer. Bigger screen grid resistors means more screen grid sag and lower screen grid current. This does effect the tone and feel of an amp. So how is this the BEST answer for EVERY amp REGARDLESS of other design considerations?

You know that no matter what you say we'll probably just go on CLONING the MISTAKES made in the golden age of tube technology with no understanding of why like a bunch of ignorant children.

[pdf64]

Agreed, I don't get how the tubes, PT or OT are better protected by a 1k 5W than a 470R 1W, as long as both are flame proof? When I've measured screen dissipation, either allow massive over dissipation under overdrive conditions; need to go >2k2 to bring it down to below 2 x max, as I remember. At least a 1W rating allows a fusing action when the screen eventually gives up.
Would be grateful for your thinking behind this. Pete.