slightly OT: troubleshooting a device that blows fuses?

[ovnifx]

Hi, I just picked up (for cheap) a rackmount solid state audio limiter that is known to blow its fuse after being on for a couple of hours.  Supposedly it only just developed this habit recently.  What are some recommendations for where in the device to start looking and measuring?  I assume I'll need to start with the DC power at its first stages, and check the connections at the heat sink(s).  Any other ideas?
Thanks!

[Paul Perry (Frostwave)]

I'd let it run with a steady tone running thru it for a couple of hours (if you can stand it!) to see whether the failure is suden, or whether the sound changes before it goes bad.
And when it does go bad I'd quickly try to see if any component has overheated. This kind of job is very frustrating.
Maybe a cap is heating up, going leaky, screwing the bias on a transistor somewhere? The current that blows the fuse must be coming from somewhere... if all else fails yu could cut the power rail & stick a few more fuses in temporarily to try to localise the trouble...

[ovnifx]

Thanks Paul!  Anybody else want to weigh in?  Or any links to related sites, or a forum in which this Q would be more on-topic?

[gez]

Replace all the electrolytics and see if that cures it.  Got nothing to lose.

[Sir H C]

I would have it set and going and when it pops the fuse take the top off quickly (and unplug first) and feel for what seems to be hot.  Could be an electrolytic (even if it is relatively new as there are a bunch of bad ones out there because of some industrial spying gone bad), could be something else.  Usually the culprit will be hot.

[brett]

Hi
Although I might try this myself:

when it pops the fuse take the top off quickly (and unplug first) and feel for what seems to be hot.

I also think that if there are high voltages in there, it is very VERY dangerous, even after the power goes off.  High-voltage electrolytic caps can hurt or kill long after the power goes off.

There's a few other things you can do.  The valve heater fuses in some Marshall amps sag at switch on, but usually fail during use due to cumulative loss of strength and vibration.  So check for sag.

Even simpler than Paul's excellent and professional idea about inserting extra fuses is to just put a tiny cut the power rail close to the supply.  Give it a thorough test (e.g. 72 hours straight).  Re-join the cut and put a new cut a bit further away.  Give it a thorough test again.  Eventually, you'll either find the region of the fault or realise that the fuse was too small for the total load.

cheers

[R.G.]

The classic answer to fuse popping questions is to build a fuse protector.

This is any device that limits current but does not cause a fuse to actually blow. There are fancy ones, but the simplest is an AC light bulb socket wired up in series with the AC line going into the device. A light bulb, all on its own, lets a big pulse of current pass as it heats up, but once it heats up it limits the AC line current all on its own. a 100W bulb limits AC line current to under 1A (100w/120V), and you get succeedingly smaller limits as you use smaller bulbs.

If something in series with a light bulb pulls little current, the bulb may hardly glow at all, but when the thing in series looks "shorted" the bulb glows brightly and the fuses do not blow.  This holds the widget in the failed condition so you can get in there and find what's wrong.

Of course, wiring up such an adapter is deadly dangerous because of the necessity to do mains AC electrical wiring so do not do this without carefully assessing your abilities to do it safely. No piece of equipment is worth dying for. Worse yet, you still have to poke around inside live equipment once the power adapter is done, which is just as dangerous. Be really sure you can do this safely before trying it.

If it were me and I had a slow-fail piece of old equipment, I'd replace every electrolytic cap in it, just to start.

[DuncanM]

First, make sure that the fuse is the correct one - I've seen all sorts of incorrect fuses in equipment over the years... Replacing with the same value as the blown one is not as good as replacing with the correct one.!!! 

It sort of sounds like a heat fault - something is getting hot and passing too much current. Judicious use of a hairdryer or heat gun (carefully) to warm up small sections of circuit board can help to identify the physical area where the problem is located. Alternatively use freezer to chill areas to see if the problem takes longer to manifest. Slower but safer. You can kill dying components with the hot air  method, but at least then it becomes a hard problem.
Cutting tracks is good but you have to be methodical. Also, it might be misleading in this situation - you might reduce the current draw on the PSU so the fuse doesn't blow, but that doesn't neccessarily mean that the problem is "upstream" from the cut... Tricky. 

My guess would be something within the PSU - electrolyics, regulators, bridge or (hopefully not) transformer. If you could disconnect it from the rest of the circuit and load it with an appropriate resistor, you might be able to eliminate the rest of the circuit at a stroke.
Or you could run the rest of the circuit from an external supply and see what happens.

Good luck.

[ovnifx]

Okey-doke!  First off, thanks to everyone who gave advice, it all helped.
I posted this query before even receiving the unit in question (an Aphex Dominator 720), just so I could know where to start once it arrived.  It showed up, and I popped the lid and powered it up.  The toroidal transformer immediately started making a horrible racket, a loud clattering buzzing noise.  After about an hour, the fuse blew.  I immediately held my hand over various components (not actually touching anything) to test the temperature, but nothing was obviously very hot; the general area of the PSU was somewhat warmer than the other areas.  Based on the suggestions that the problem lay within the PSU, I decided to remove it and replace parts to see what helped.  Fortunately the PSU was on its own small board well segregated from the rest of the unit, so it was relatively easy.  I didn't feel like going one part at a time, so I replaced the electrolytics (two large and one small) and the bridge rectifier.  I also noticed that the thermal compound under the aluminum block of the heatsink had mostly dried up.  I took care of all that (it was hella hard finding 35V electrolytics in the same dimensions), reassembled everything, and powered it up.  Ouallah!  It has now been powered on for 24 hours straight with no noises, smells, or interruptions!

Anyway, just wanted to post a follow-up, and express my thanks to you guys.  Cheers!

[Connoisseur of Distortion]

that's satisfying. congratulations!