Resistor tolerances and power ratings

bandmaster188

Couple of questions, iv'e just bought 30 1ohm resistors to pick some out to put on the cathodes of my output valves for measuring bias. I bought 2 watt wire wound with a 5% tolerance. They all measure between 1.3 & 1.4 ohms which by my calculation is 30 to 40% off. Is there a particular brand that will have better accuracy? The others I bought were carbon film 100k for plate loads rated at 1 watt. The fender schematic states 1/4 watt. Will the higher rating cause any adverse effects?

ICBM

I'll take a bet that your meter has a zero error of between 0.3 and 0.4 ohms . Touch the probes together and see what it reads. In any case, you won't get anywhere near a precise enough reading to check whether the resistors are within tolerance or not, you'd be looking for +/- .05 ohms at most, and probably much less than that.

I would not worry too much about a 5% tolerance - it isn't ideal, but in practice is close enough since there is already about that much - or more - error caused by the unknown value of the screen current, which is included along with the plate current when measuring the cathode current. It's true that you don't want two very different resistors in the same amp since that will make you think the valves are mismatched, but you could actually test a couple in the circuit by hooking them up in series and measuring the voltage across each one - if they're different, try a different pair.

Re the 100Ks - resistors actually have two ratings, not only power but also voltage - generally, high-power ones also have higher voltage ratings. This matters in valve amps and especially in plate resistor positions, because smaller resistors can have their voltage rating exceeded even though they're within the power rating. Fender are notorious for this problem on many of their modern amps! Blown 1/4W plate resistors are one of the most frequent faults, and the reason is the voltage and not the power - although 1/4W is cutting it a bit fine even for that as well. Using higher-power resistors has no disadvantage - in fact, Fender used 1W plate resistors on many of their old amps, and Matchless use 2W (I think) throughout.

There is a theory that because carbon-comp resistors produce more self-distortion - which is a desirable thing in guitar amps (but nowhere else!) - when they're operated close to their voltage rating, hence it's undesirable to use overly large ones, but Matchless amps seem quite highly regarded...

bandmaster188

cheers for that ICBM. yep i get a .3 reading when i touch the probes together!

ICBM

bandmaster188 said:

cheers for that ICBM. yep i get a .3 reading when i touch the probes together!

And that could mean anywhere between 0.25 and 0.35 ohms, since the meter can't display a more precise reading than that…. which means that the possible error in the meter display is the same - 0.1 ohm - as the absolute maximum difference in two resistors at opposite ends of the tolerance range! (I would guess in the upper half of that if some of the resistors read 1.4 ohms.) So there's no way of checking their values directly.

But run a current through them and produce a measurable voltage and it becomes easy.