Resistor in DC supply line

The aforementioned resistor failed open-circuit without any external sign of overheating at all. Bonus points will be awarded for anyone who can think of a reason why this might have occurred.

My guess would be to act as the R in a pi filter. Hard to say actually. Or maybe the idea is to provide current limiting on application of power. The downstream cap seems to be 2,000 uF which is a whopper. How much current does the output drive? What is the output load? The output cap is 1,000 uF.

I couldn't come up with anything better than the pi filter possibility either to be honest, but haven't done any calcs to see if that would make any sense or not. This IC directly drives a 2W 8 ohm speaker.

Have you replaced the resistor yet? I have no idea why it broke, but it's not like everything is made perfectly. Have you tried just reflowing the solder? I'm assuming it's surface mount.

** It provides peak current limiting at the moment the unit is connected to a DC supply - plus a measure of noise filtering thereafter in conjunction with the 2200uF electro. Should be a composition or WW type, not carbon film, to survive the energy in such surges. Nothing strange about the original film resistor failing open and invisibly when subjected to 13A spikes lasting a few milliseconds.

.... Phil

No, it's tru hole. Sorry I thought I'd linked to a picture as well but it hasn't shown up. Let's try that again:

It's the resistor right next to the largest electro.

I'll second that. British power plugs have a built-in fuse and R30 makes it less likely that this fuse would blow on switch-on.

Bil Sloman, Sydney

Film resistors are not good surge absorbers, Phil Allison has already given you the best answer. The only other reason I can imagine is resistor end cap failure, carbon film resistors without crimped end caps were made by a very few manufacturers and were very reliable but lost market share.

piglet

It's audio. Circuit randomness is high.

Protection mostly. Like you see with the old sg3525.

Cheers

Resistors are cheaper than fuses, I guess.

I was recently looking into using resistors, low ohms 0805's, as fuses on two D25 connectors, on the theory that I'd rather fry resistors than pcb traces. But it looks like we can squeeze in 48 leaded polyfuses.

We've found surface mount polyfuses to be flakey, and they don't save much PCB area.

The AoE supplement, the X-chapters, has a bunch of experimental data about exploding resistors. They used my DUT (device under torture) rig.

Okay, thanks. I know the photo could have been better, but it's good enough for this question: is this a MO resistor or a carbon one? The radio it comes from is about 35 years old and with my eyesight I struggle to tell the difference.

I just thought - we haven't heard a peep out of Win since he invited corrections to AoE III (or was it II?) Anyway, I hope he's okay. We didn't get on politically, but he was a hugely valuable contributor here.

But you say essentially the same thing about RF! Am I to assume you're only comfortable answering topics relating to the bit of the spectrum in between? :-D

Cheaper than a fuse.

RL

Yeah, but ONE OHM? How much current is that going to limit?

It's more a matter of how much energy is needed to blow it up.

I used to insert small resistors in the power supply connections of subcircuits because that makes it easy to measure how much current goes into each. That's a great help in fault finding.

Jeroen Belleman

Resistors can sometimes handle 10x rated current or power briefly. It can give out smoke signals and/or change requests (burnt out mark).

I also use halogen/incandescent light bulbs as current meters and fuses.