Resistors losing resistance over time

[Spencervs]

Something I have noticed when checking some old resistors is that while I am used to the idea of them increasing in resistance over the years, do they decrease too?

So the resistors I am testing are in a valve radio and look like those standard pale green ones but with a roughish surface. They are in tolerance but I worry they may keep decreasing until perhaps a bang or busted valve or two. At least the other way they will eventually just starve the circuit.

Is this possible that they decrease and should I replace them on sight? Or is it the rest of the circuit causing the readings to be off?

[Nuvistor]

Can’t say what happens after 40 odd years but a rule of thumb I used was that high values rose in value while low values went lower. Many values could go low if they were over run.

[The Philpott]

As regards the last line of Post No.1- As you have already sussed out, unless you can see all of the relevant circuit the best way to be sure is to test 'out of circuit'..avoiding current finding an alternative path and skewing results

Dave

[Skywave]

Originally Posted by Spencervs

The resistors are in tolerance but I worry they may keep decreasing . . .

Your use of the present tense - "they may keep decreasing" - implies that in the past you have measured their ohmic values at least once & find that now those values are less. So, how long ago in the past was the last set of measurements made and by what %-age or ratio have they changed?

Al.

[Tractionist]

Why do they [even] go high - I've never contemplated the reason, and just accepted that they do?

[Sideband]

Originally Posted by Spencervs

They are in tolerance but I worry they may keep decreasing until perhaps a bang or busted valve or two. At least the other way they will eventually just starve the circuit.

Is this possible that they decrease and should I replace them on sight? Or is it the rest of the circuit causing the readings to be off?

You are worrying FAR TOO MUCH! Resistors can and do decrease in value BUT usually because they are overrun by a fault condition or underrated in the first place. As has been said previously, to check resistors accurately you need to test out of circuit by unsoldering one end. You can often check resistors in-circuit (in valve radio's at least) without disconnecting but you need to study the circuit to see if any possible parallel paths might reduce the apparent resistance. This is certainly true of transistor circuits and with televisions. Always best to unsolder one end of a resistor to check its value if there is any doubt.

As a general rule, resistors will increase in value and you tend to notice it far more with high value types. It's not uncommon for (say) a 220k anode load resistor to creep up to several megohms or even go open-circuit. As previously stated, lower value types (in the ohms range) can often go lower in value but because they are overrun.

As for causing a bang and 'busting a valve'..........very unlikely! A faulty valve is more likely to cause a low value resistor (like cathode bias) to fail rather than the other way around.

I have many sets with original resistors that are still within tolerance and even if they have drifted slightly above or below, that is no reason to change them. Most radio's will work quite happily with resistors way out of spec.

Stop worrying and enjoy the radio's!