Resistors invisibly failing

[Cursitor Doom]

Gentlemen,

Whilst trying to discover why there was no audio output from a 4W amp
board, I noticed there was only 0.6V on the supply pin of the audio
output IC concerned. There should have been 13.6V. This pin was fed
via a 1 ohm 1/4W resistor. DC voltage measurements showed there was
13.6V on one side of this resistor and 0.6V on the other! Clearly
something amiss here, I thought. So it must have gone open circuit.
However, there's not the slghtest sign of any phsyical damage even
under high magnification, whatsoever to it: none at all. Prior
experience has always taught me such resistors burn out in an obvious
way which is dead easy to spot. Not this one. I'm just wondering how
unusual this is and if anyone else has encountered such an issue with
a resistor. Here's a photo. The resistor in question is the one right
next to the largest electrolytic. There it is, looking all innocent
like butter wouldn't melt, yet it's caused me a massively
disproportionate amount of head-scratching, the little shit.

https://disk.yandex.com/i/dE9o0lh937qdrg

When I get that out of circuit.... just you wait....

[Liz Tuddenham]

Micro crack in the solder?


--
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Cursitor Doom]

It'd have to be a pico crack given the hi-mag examination I gave it
and found nothing suspicious.

[Trevor Wilson]

**Many years of experience has taught me to suspect low value (<47 Ohms)
resistors and high value (>100k) resistors. Both types can fail with
little or no visible evidence. Cracked carbon types are the ones that
are problematic, as metal film types seem to be far more reliable.

--
This email has been checked for viruses by Avast antivirus software.
www.avast.com

[Dave Platt]

In article <ujpfdilc78ahgovbs...@4ax.com>,

Cursitor Doom <c...@notformail.com> wrote:

>>Micro crack in the solder?
>
>It'd have to be a pico crack given the hi-mag examination I gave it
>and found nothing suspicious.

Possibly a bad wave-soldering job? I've heard of cases where
bad/incorrect solder or flux caused a mostly-failed joint... the
solder flowed up over the component lead, formed a decent-looking
meniscus, but didn't actually bond to the lead properly. The
lead could eventually crack away from the solder _inside_ the
joint, leading to an open circuit which is invisible to the eye.

You might want to try solder-sucking away the existing joints,
re-fluxing, and re-soldering, and see if that fixes it... although
I'm not sure if this would teach you anything more than you'd
learn by just unsoldering the resistor and measuring it out of
circuit.

An internal crack in the resistor is probably more likely, though.

[Liz Tuddenham]

If it were a large heavy component, I would look for circular cracks in
the rings of solder around the terminals; the solder blob in two parts
with one attached to the pin and the other as a circle surrounding it.
This used to happen mainly where the dip-soldering conveyor was vibrated
to shake off solder splashes and mass of the large component stressed
the cooling solder blob. The centre of the blob was cooled by the pin,
the periphery was cooled by the track and the last bit to solidify was
stressed into a ring of 'dry joint', which later failed almost
invisibly.

I can't imagine that happening to anything as small as a surface-mount
resistor, but include this historic information for the benefit of
anyone repairing through-hole boards with large heavy components.

[Cursitor Doom]

No, this is through-hole and about 35 years old. As a mattter of fact,
it also suffers from the issue you mentioned above, in the PSU
section, where the heat from a couple of 5W resistors exacerbates the
problem. In fact the original PSU only lasted about 10 years before
requiring a rebuild as a result of this particular design flaw.

[legg]

On Sat, 12 Aug 2023 19:36:04 +0100, Cursitor Doom <c...@notformail.com>
wrote:

Googling F2887103 gives you all the info you'll ever want about R30.

RL