Always trust your meter

[aron]

I had an amp that had an unusual weird tone and the bias seemed funky. Occasionally one of the tubes would be really bright and the other not at all. I thought it was mis-matched tubes.

However, it even persisted with matched tubes and it sounded crappy.

I had 10 ohm resistors from cathode to ground for current sensing. One resistor would "make sense" and the other would be way, way off.

When I measured the resistors, I got 10ohms on one side and 1.5xx MEG? on the other. I couldn't believe it.

I desoldered both and measured and sure enough, one was "open" or a very, very high resistance. No burned markings or anything.

Weird. Both were 1% resistors.

[davebungo]

This is one of the problems with resistors used in power circuits.  They can go open circuit like this after a time or sometimes they go low resistance if in a high enough voltage circuit which causes over-heating and then eventual further failure and a circle of decline.  I had a power resistor in an MXR limiter in series with the mains transformer (no less) which overheated to the point when it actually de-soldered itself from the PCB and dropped out!  BTW I believe this was MXRs cheap-skate way of avoiding the use of a 220V transformer for the European market by dropping half the mains supply across a simple resistor.  I know this because as the resistor had gone low resistance, the transformer secondary was supplying too much voltage to the DC regulator which was also getting a little sweaty.

If I had to use resistors in a circuit like this I would always put 2 or more in series to reduce the risk of failure.  In your case, I am guessing that the 10ohm R probably wouldn't see much voltage anyway as one side is to ground so using 2 resistors probably wouldn't help.

[MR COFFEE]

Aron,
The cathode circuit of a tube amp is one of the few places where a carbon composition resistor makes sense. They take time-limited overloads fairly gracefully without opening.

What probably happened is that the old tube did a brief short and mostly fused the precision resistor (which is made from a thin resistive film - carbon or metal alloy - with a spiral cut in it). That construction doesn't tolerate overloads well at all, because the conductor has a small cross-section. Carbon Comp resistors (no MOJO bullshit please :lol: ) are a slug of material with leads stuck into the bulk material, and the large cross-section tolerate up to 5 seconds or so of 1000% overload.

I can't tell you why tubes do those brief shorts, but they do. I suspect it has to do with the getters they put into tubes to try and soak up the stray residual gases the vacuum pump doesn't pull out, but I can't tell you the technical details of it. Maybe an Ampager can kick in here on that. I just know that tubes do briefly short and then clear the short, and cathode resistors have to suck it up or die.

You may want to consider replacing both resistors with carbon composition ones if it's in a circuit/rig you really like. Just a thought. :idea:

[nelson]

I have an amp that is sounding misbiased, I think given the fact there are 50yo carbon comps in there, they are now out of tolerance and need replacing. I am going to go through it replace all the 50yo electrolytics(for obvious reasons) and resistors that are out of tolerance.....

I have still to go through the amp and measure with my DMM, it could be the 40 year old tubes......... I may replace them too. They are mullards......would be a shame if it were the tubes....

didn't want to hijack, just thought I would mention that i am having a similar experience.

[aron]

You may want to consider replacing both resistors with carbon composition ones if it's in a circuit/rig you really like. Just a thought.

Thank you. That is exactly what I did.



Aron

[aron]

Code Select Expand

I have an amp that is sounding misbiased, I think given the fact there are 50yo carbon comps in there, they are now out of tolerance and need replacing. I am going to go through it replace all the 50yo electrolytics(for obvious reasons) and resistors that are out of tolerance.....

Good luck! Let us know how it goes.

I had Mullards in my amp but amazingly it sounds ok with EH!

[R.G.]

The cathode circuit of a tube amp is one of the few places where a carbon composition resistor makes sense. They take time-limited overloads fairly gracefully without opening.

Big overloads are indeed a serious sickness, but carbon composition resistors constitute an ugly tasting cure.

The biggest problem with using them where aron is trying to use them is that they don't have enough precision, and can't. If you had 1% CC resistors, they'd drift enough to make them not 1% any more with as little as soldering heat.

IMHO, the right answer here is to use a bigger non-CC resistor. Ceramic composition resistors were developed specifically for their overload capacity, and wirewounds offer nearly the overload ability that CCs do. 1% WW is easy enough to find.  

If it weren't for the precision, I'd be inclined to agree with you.

Carbon Comp resistors (no MOJO bullshit please...

The "mojo" in CC's is a known quantity. It's not the magic bullet that a lot of people want to promote, but they do have a slight even order distortion for big (over maybe 50V peak) signals that can be calculated and measured. There is a grain of truth in CC mojo - but only 1.000 grains, not a ton as we get asked to accept.

[aron]

Ahhh sorry, I used carbon film and measured until I found ones as close as posslbe to 1ohm.

[brett]

and wirewounds offer nearly the overload ability that CCs do. 1% WW is easy enough to find

Gotta agree with RG.  Wirewound works and they're mighty tough.
(I accidentally did something really stupid a while ago with a 5W wirewound and gave it about 120W to dissipate.  Although the internals glowed bright orange for 10 seconds, it didn't break either physically or electronically.  Amazing!)

[aron]

I think we want them to burn out, they are supposed to provide protection; at least that's what I have read.

[brett]

I like the idea in amps of rugged components protected by well-placed fuses.  Unfortunately, in the case of mild or very short-term overloads, fuses take a second to blow, hence the need for over-rating.

[R.G.]

I think we want them to burn out, they are supposed to provide protection; at least that's what I have read.

Protection is a tricky subject. Using low value resistors as fuses is at least adventurous.

The problem you always run into is ensuring whether the protection element holds up to where you want it to stop holding, then opens cleanly.

Resistors are NOT designed with a fusing current in mind. Fuses are. But even with fuses it's hard to ensure that a fuse will pop when it should and hold when it should. The actual thing that pops a fuse is the temperature of the link, and that's determined for all fast conditions by the energy it absorbs, proportional to the integral of the square of the current through it and the time that current acts. The I-squared-t rating is classic fuse specification stuff. Waveform matters - how fast does the current spike, how high, and how does it drop? Resistors are not designed to drop out at any particular I squared t.

If it were my amp, I'd put 3W to 5W 1% wire wounds in there and fuse the cathode with a real, no fooling fuse. This is one place where the designers played clever with a 1/4W resistor taking the place of a fuse. The resistor costs maybe a penny, versus the buck or so of a power resistor and fuse. Good for company, bad for reliability. "Works as good as new" is not much of a recommendation in this case.

If I were really incented to do something like this, I'd put in high power sampling resistors as above, watch the voltage across them, and turn the B+ off when the circuit decided it had been too high too long. A resistor/cap does a nice job of integrating the cathode voltage, and a comparator does a nice job of deciding when enough's enough.

[aron]

Thanks R.G. I guess there is the problem of how to mount the fuse holders....

[MR COFFEE]

RG,

I'm afraid I, too, tend to go for the "brick shit house" approach to building things :wink:

Were you really serious about the comparators and all or was that at least partly tongue-in-cheek? Sounds like the Rolls Royce engineering approach to a tube amp (why use two a two-part thermostat in the radiator when a six-part does the job a bit better?)

'Course with the price of tubes these days, maybe that comparator idea isn't all THAT extreme :lol:

Would you use a mosfet or a relay to cut off the B+?

[Paul Perry (Frostwave)]

I've seen more fuses fail (without reason, that is) than correctly rated cathode resistors. If you are worried about a cathode resistor overloading and failing, use a Welwyn wire wound one!
Sure, put a fuse in the HT line.. but sprinkling them throughout isn't going to increase reliability.

[Paul Marossy]

I should check out the 1 ohm cathode resistors on my '74 Twin Reverb, I think something quirky might be going on with the measurements I took last...

[Doug_H]

Aron FWIW I like the big cement wirewounds for cathode circuits in the output stage.  I usually go high on the wattage here, 5W where 1W will do, etc.

Doug

[puretube]

inductance-free!

[R.G.]

Were you really serious about the comparators and all or was that at least partly tongue-in-cheek? Sounds like the Rolls Royce engineering approach to a tube amp (why use two a two-part thermostat in the radiator when a six-part does the job a bit better?)

No, actually I was serious. I designed up a circuit with sampling resistors, comparators and a power MOSFET to open up the cathodes. You watch for current that's too high for too long, and when it happens, you fire a latch with a comparator and turn off the tubes by opening the cathode circuit.

I called it a tube saver but it actually saves either the power or output trannies. You can make it as IN-sensitive as you like.

Fuses do get old. There is a nuisance popping mechanism if you set them too close to their threshold, and they don't protect fast if they're too far away. Watching the actual current and stopping things before they get out of hand is a reasonable thing when you look at (a) the reliability of solid state stuff - if it's well designed - and (b) the cost of a MOSFET and some comparators and a latch compared to the cost of a replacement transformer.

Sure, put a fuse in the HT line.. but sprinkling them throughout isn't going to increase reliability.

Never send something with a high intrinsic failure rate (IFR) to protect something with a low IFR.

[aron]

Well the interesting thing is I think I have read about the "fuse" current sensing resistors in at least 2 well known amp books.

One of them did have the fuse in series from cathode to ground.