"Replaced electrolytics..." -- A newb question...

[jasonbrent]

Folk,

When folk are repairing things, I frequently see that they "replace the electrolytic" capacitors as a near-first step.

Does this mean that polycarbonates, mylar, mica, tantalum bead, and ceramic disc capacitors are generally not thought to fail as easily and so they replace the electrolytics and ignore the rest unless there is an obvious problem with the other capacitors?

I'm pretty sure the answer is yes, just want the sanity check please.

-j

[Paul Price]

No.

All depends on the type and age of the equipment in question, but as a general rule, it completely foolish to waste time money and effort to replace any capacitor that doesn't test bad.

Don't fix anything that ain't broke yet.

If a piece of equipment hasn't been used much or at high temperatures, it is probably going to work ok. Make a visual inspection if you have doubt, look for stand-up capacitors that have raised domes or show signs of leaking. If all looks ok, it will probably work fine. If it doesn't you can just troubleshoot the suspected circuit sections of equipment to isolate and replace any defective components.

The practical approach with very old equipment that hasn't been turned on for many years to attach a low voltage (<30V) to the power supply filter electrolytic capacitors to reform them before plugging them into the mains. Takes a few hours.

Old paper capacitors are commonly leaky, but just using a multimeter resistance measurement can show even a fractional uF capacitor taking charge, but noticing any high steady-state resistance indicates a defective capacitor, that is, if it is not a large uF valued electrolytic.

[Bassman59]

I'm pretty sure the answer is yes, just want the sanity check please.

Yes.

Because the electrolyte in aluminum caps especially has a lifetime.

Look at cap data sheets. You will see that they are rated for so-many-thousand hours of life at some given temperature. That is, lifetime is a function of temperature. Older consumer products were built using the available caps of their day, so expect that their lifetimes aren't as good as caps we can buy today.

There are two aspects to temperature. One is the obvious, the ambient environment. Is it warmer than, say, room temperature? Elevated temperature in an enclosure shortens a cap's lifetime. If the cap is keep cool then it can last considerably longer than the lifetime in the data sheet.

Second is cap self-heating. This is caused by ESR (equivalent series resistance), and the higher the ESR the more heating. This is especially a problem in power-supply smoothing applications, where there is considerable ripple current that causes a cap to heat up.

[jasonbrent]

I'm pretty sure the answer is yes, just want the sanity check please.

Yes.

Because the electrolyte in aluminum caps especially has a lifetime.

Look at cap data sheets. You will see that they are rated for so-many-thousand hours of life at some given temperature. That is, lifetime is a function of temperature. Older consumer products were built using the available caps of their day, so expect that their lifetimes aren't as good as caps we can buy today.

There are two aspects to temperature. One is the obvious, the ambient environment. Is it warmer than, say, room temperature? Elevated temperature in an enclosure shortens a cap's lifetime. If the cap is keep cool then it can last considerably longer than the lifetime in the data sheet.

Second is cap self-heating. This is caused by ESR (equivalent series resistance), and the higher the ESR the more heating. This is especially a problem in power-supply smoothing applications, where there is considerable ripple current that causes a cap to heat up.

Thank you... I've been attempting a repair on a DC  Voltage Standard/Calibrator and I'm to the point of replacing caps after going through the manual's instructions of checking (entirely too expensive) relays, mosfets, and such.

About to swap out out $60 bucks in electrolytic capacitors and second guessed myself about the other 70ish capacitors in this unit that aren't electrolytic.

Appreciate the response!

-j

[jasonbrent]

No.

All depends on the type and age of the equipment in question, but as a general rule, it completely foolish to waste time money and effort to replace any capacitor that doesn't test bad.

Don't fix anything that ain't broke yet.

If a piece of equipment hasn't been used much or at high temperatures, it is probably going to work ok. Make a visual inspection if you have doubt, look for stand-up capacitors that have raised domes or show signs of leaking. If all looks ok, it will probably work fine. If it doesn't you can just troubleshoot the suspected circuit sections of equipment to isolate and replace any defective components.

The practical approach with very old equipment that hasn't been turned on for many years to attach a low voltage (<30V) to the power supply filter electrolytic capacitors to reform them before plugging them into the mains. Takes a few hours.

Old paper capacitors are commonly leaky, but just using a multimeter resistance measurement can show even a fractional uF capacitor taking charge, but noticing any high steady-state resistance indicates a defective capacitor, that is, if it is not a large uF valued electrolytic.

Paul,

This logic makes sense. In this specific case, I'm (attempting to...) repairing a device that is a couple of decades or so old and I've already been through the digital circuitry and a lot of the analogue circuitry. This unit has a few large caps in it that were designed to buffer and sometimes drain >1000V quickly. Those are the expensive ones and a challenge to test in circuit.

They do appear visually fine, and seem to test more or less correct as best I can tell with resistance, etc.

Beyond the large electrolytics, it has 8 smaller caps spread around the board that are filter caps.

I intend to pull the large caps intact and test them  out of the board and swap in replacements as I go. If the pulled ones are good, I'll put them back into my parts bin.

The smaller ones are suspect as I have a couple with unexpected readings.

I just wanted to ensure I wasn't missing the concept of electrolytics implying something more than that specific type of cap since the device under repair is smothered with all of the other types I listed.

I wouldn't generally start with "replace all caps" and in this case it's the Nth step that I'm calling a "hail mary". If it isn't fixed after this, I have a couple of zeners en route for some questionable ones on the board.

Device has voltage potential on the outputs when it shouldn't, such means something is shorted somewhere.

-j

[Paul Price]

It is quite possible to mis-install a replacement capacitor or cause other damage in attempting to replace so many capacitors.

You might easily find that replacing all suspected components might just be guesswork that results in the instrument being in the same repair state as before your efforts.

Your best bet would be to identify the defect to a module, circuit and then to a component level. This would require a schematic and some knowledge about how the circuits inside this instrument work. Of course, if you can identify the defect in operation and present the schematic on this forum you might easily find the help you need to effect a quick repair.

You state that there is a voltage present at the output when it shouldn't be present. This sounds like a shorted active device, a BJT or other active device, not a capacitor or zener problem. From my experience, connecting the output to a short circuit at a high voltage setting or even just to a circuit having a HV charged capacitor could easily cause a BJT or MOSFET to short.

Even without a schematic, just by using an ohm meter at  x1 ohm setting, you might easily find a shorted junction of a BJT or MOSFET or even a welded relay contact. If the output current of the device is high when set to zero output, this is a sure sign of an active device failure.

[katzohki]

Just wanted to chime in and say that there was a known problem with many many capacitors during a certain time frame. Known as the "capacitor plague" it relates to bad chemistry having been used in capacitors placed in many consumer electronics of the era. I bet people who replace caps habitually lived through this...

https://en.wikipedia.org/wiki/Capacitor_plague