cathode resistor sudden rising value on 7591

[glassandlight]

I accdently posted in this in new members, I appologize.

Strange and unexplained event on a cathode resistor of 10 ohms, 1/2 watt, Xicon brand suddenly racing up to 250 to 350 ohms. They are supposed to be rated to 350V. This has happen three times now, different tubes, different sockets, different times. I caught the event twice after ruining one tube. The tubes are NOT shorting, the resistor is NOT popping. The tube which was damaged, now draws considerable more plate current than before, the other two are OK and still the same. Tubes tested on a restored and calibrated Triplett 3444.

The amp is a restored Sherwood 5500 II, hot-rodded with outstanding parts, except these Xicon resistors. Plate is 405, screen is 397, grid is pot adjustable to get 22ma across the 10 ohm resistor so the tubes run quite cool. The amp is never run more than 2/3 open driving 8 ohm speakers.

Everything checks out as OK circuit wise so I am stumped what is making these resistors fail unless it is simply the brand.

Any ideas????

[mastodon]

Sounds odd. How did you measure those resistors?

Oh and I might just be my own dumb self here, but how does a cathode resistor that apparently rises in value kill a tube? 'Faulty' cathode-R resistance rises --> voltage drop over the resistor becomes larger --> grid becomes more negative compared to cathode --> tube squeezes and current drawn decreases --> voltage over cathode resistor drops --> situation stabilizes a round a different, lower idle current than before. A cathode resistor shorting certainly would be more worrying. Am I missing something?

[glassandlight]

fluke 87 III to measure everything.

I have no idea if the resistor "killed the tube" or not, but it did draw 2x current than before on the Triplett so it could not be matched up again.

But the question is...why would these resistors jump up in value like that. This also happen in a Fisher 500C which I did not mention in the first post.

Has this happened to anyone else or are they a bad Chinese batch of junk resistors.

[kevinkr]

Usually they rise in value as a result of being massively overstressed due to a fault condition.

They could also be defective..

I've used this brand of resistors on and off, and have never had a problem with their products.

[Eli Duttman]

I'm guessing the 7591s used are current production. Current production 7591s are intolerant of the liberties 1960s manufacturers sometimes took with the published grid to ground resistance value limit. The tubes may have run away and overheated the 10 Ω "idle" current set resistors. It doesn't take much to "cook" 1/2 W. rated parts.

What's the grid to ground resistance value Sherwood used? Please post a link to the schematic.

BTW, what is the material construction of the troublesome parts?

[aardvarkash10]

any chance the bias circuit is going intermittantly open and therefore forcing the cathode resistor to become the last of the great white hope?

[dcgillespie]

G&L -- I'm guessing that since you said this amplifier has been hot rodded, there has been additional capacitance added to the power supply, and the screen grid take off point in particular. If this is the case, I am rather certain that this is the reason for your problems. You are describing the symptoms of screen instability to a "T", which can do light damage to cathode bias measuring resistors and tubes, making the resistors read way high and/or the tubes off spec on the one hand, or in the worst cases, cause an arc and take the resistor and tube out altogether. You say the tubes did not arc. But I bet more happened than you suspect.

If capacitance has been added to the screen grid circuit, it can cause screen instability in new production or NOS power tubes that are run in fixed bias at voltages that are near the max screen voltage rating, and also nearly equal to that of the plate voltage. It is particularly prone to happen in high value Gm tubes such as the 7591 under these conditions. The answer is to add screen stability resistors to each output tube, that will eliminate this kind of problem. 100 ohm 1/2 watt resistors of nearly any type construction will do. Add one at each tube between the screen terminal, and the source providing power for it.

If resistors such as these are not in place, it is likely that they will solve your problems.

Good luck!

Dave

[stratus46]

350 Volt rating is total nonsense for a 10 ohm resistor. 2.24 Volts puts it at full power. What is the normal operating Voltage? Is it simply getting fried?

BTW resistor Voltage ratings are for resistors below the dissipation limits. For a 1/2 Watt resistor the 350 Volt limit starts at 245 K Ohms and up. IOW a MAX of 350 Volts OR 1/2 Watt.

G²

[MattyAve]

Are you measuring the resistor with amp on?

[glassandlight]

as to the grid circuit going open, I don't think so the diodes are Fairchild Stealth. From there on each channel has its own pot to set tube pairs ( the original was common to all 4).
The capicitors in the entire amp are new, Atom TVA, F&T for the power supply, United Chemicon and Nichicon Silk caps in bias.

As to dcgillespie's reply, the screen has a wire wound 480 ohm/5 watt feeding all 4 tube screens (this is stock), power taken from the plate, extra capacitance was added in the power supply from the F&T being 50uf instead of 40uf, but it is fed into a 33 ohm 10 watt resistor before distribution into the general circuits...I can't see that as the culpret. Since this has now happen 3 times, you may have a point add in a discrete resistor to each tube. As for the tubes, they are near new EH and test fine, with current draw within 1% each. The tubes did not arc because I watch and listen like a hawk, anything that changes(sound wise) has me on top of it asap. This last episode did have the tube giving a yellow ( I think) glow at the top, but I had the amp off in a heart beat so I didn't let the occurance go further, and the tube tests fine.

The BIG question is that this has also happen in a Fisher 500c that I restored for a friend last year.