Just like in any electronic device, capacitors (especially e-cap) have always been one of the weakest link in reliability .

Most so for capacitors directly across mains AC line as they are subject to much continuous voltage spikes.

anything can happen in the tx summer . . . was 105 when i was there last month . . . please provide voltage of capacitor (from label) and system voltage (as supplied by electricity supply company) . . . and farad rating if available

The poor quality of the capacitors. You should use a capacitor of a reputable manufacturer, not just cheap capacitors.

Zlatkodo

web page

If nothing works ask him to pull long wires with the cooling coils, and hang all capacitors indoors in the air-conditioned environment.

But then when the cap blows it will send the smell into the conditioned space.

This does not sound like much of an issue to me.

Switch to a capacitors over to ones with a far higher voltage rating but of the same uF rating it will have a larger body of course. That will greatly reduce the applied internal thermal load placed on the capacitor thus greatly increasing the working life of them in high ambient temps.

15 months in Texas heat is probably 15 years anywhere else.

Elevated temperatures can drastically reduce component life. It seems clear that the operating temperature is higher than expected for much longer periods than expected and this is causing the capacitor failure. I'd contact the manufacturer and ask if they can recommend/supply a capacitor designed to handle abnormally high temperatures. You could try to find one yourself, but getting the manufacturer to do the work and supply a better cap for free would be preferred.

The post #4 suggestion to get the capacitor out of the "heat" could work, but you will need to calculate the cost/benefit ratio of running extra wiring vs. the maintenance/repair trips. If you do attempt to relocate the capacitor, follow all applicable codes to prevent possible injury and/or fire damage.

The main reason really why these starting capacitors give up at high ambient temperature is that, these capacitors are coupled to supply 3phase motor of the compressor. At high ambient temperatures, condeser(located outside) pressure is significantly above normal, making the compressor motor stressful thus to the current and phasing of this capacitors. That's is why these capacitors often give up at hot season, much more if there are voltage drop/variation of the entire power supply system

where is baffled? if baffled answers some queries raised in this thread then a specific solution can be offered instead of generalizations and conjectures . . .

I give mjb a GA. He didn't suggest that the problem could be a function of capacitance and pointed out that the presence of a capacitor indicates single phase. The start capacitor is not located on the fan motor in the hot slipsteam of the condenser but rather in a small enclosure within the pump enclosure.

I didn't know that these capacitors came in different temperatire ratings - I will check it out.

I wonder if, as the temperature rises and with it, the demand, as the grid reaches maximum capacity, the product gets dirtier. Is there any reason that that might happen? If that were so then the problem would be only indirectly related to temperature.

There should be two capacitors, one for the compressor and one for the axial fan motor. Which one is actually failing?

I suspect these capacitors are more likely to be motor start+run capacitors. That means the internal temperature of the capacitor will be even higher than the case temperature. Where the excessive heat comes from, internal or external to the capacitor, doesn't really matter in this case. I still think your problem boils (pun intended) down to excessive temperature. The links below may help.

http://www.illinoiscapacitor.com/tech-center/life-calculators.aspx

http://www.aerovox.com/Portals/0/PDFs/Hid-Lighting-Apps.pdf

While a higher voltage rated capacitor may last longer, the main problem is still temperature. A higher temperature rating, like the HID caps in the above link, or moving the capacitor to a cooler location are your best options for solving the problem.

You could mount it / thermally couple it to the compressor intake line - which should be sufficiently cool to make a difference.

Test -

Bad idea.

If the capacitor, for any reason, blows up (likely possibility), you will also have a damaged copper inlet pipe and gas los etc. Leave it simple and put it anywhere less exposed to extra heat from the hot air being moved out by the fan. That should be enough (otherwise the problem will probably be in the capacitor selection...).

Chances are low that a cap (paper & foil) blowing would penetrate the pipe. Also could put cap in an aluminium clip which would guarantee no pipe damage.

Sounds like the computer caps problem all over again.

Yes, It is easy to copy something. They just reproduce the original component in all its details. They later on try to cut cost. Since they don't understand the reasons why it is done that way, they eventually cut into something important and the failures start popping up. That is my simple explanation for this dropping quality.