Unmarked mystery cap can

[R.G.]

AC rated caps have to withstand the full peak voltage of the AC wave, which is 1.414 times the AC rms value, plus some extra fudge for AC lines not being well regulated, +10% being common, as well as line transients and lightning strikes.

They often have different AC and DC ratings. It is *critical* that you use motor RUN caps of polypropylene film and oil and not NP electrolytic motor START caps if you go this way.

An AC motor run cap rated at 240Vac would be good to 240Vac*1.414 = 339Vdc, even including a 10% high line and such. You might find a dual-marked film cap with an even higher DC rating, but unless you get the maker's assurance that this is OK, don't trust it.

For a 500V supply a 370Vac rated cap should get you there (370*1.414 = 523Vdc). 440Vac and 450Vac are common ratings as well, and they are available in up to 100uF. Note that this kind of cap is bigger, heavier, and lower ESR than electros of similar rating, and will produce cleaner DC but higher transients on the rectifiers, so it may be a bit harder on tube rectifiers.

I looked on ebay for the motor run caps listed. Just typing in "motor run capacitor" turns up a lot of them, some of which look to be good deals. I like the metal can type, as it helps in getting the heat out that is caused by the high RMS currents caused by rectifier use.

Note that these are my personal preferences and you should satisfy yourself as to the effectiveness and safety of any non-labeled use of components. Your mistakes must be your own, not mine.

[Rhodesplyr]

Just a guess: as motor caps they are not rated for rectifiers and filter caps, they are for AC use and so would be rated for peak to peak voltage if the AC mains.

As the audiophile craze for building all-polypropylene power supplies blossomed about 10 years ago, some guys checked with the companies making motor run capacitors and discovered that they did have DC ratings for them. They just weren't published since they weren't marketed for that use. The DC ratings were pretty high, as I recall, but at least the information should be out there somewhere.

More specifically, the AC ratings of these caps take into account the heating effects of being charged and discharged almost completely on each AC cycle. Working as DC filter capacitors with a few volts of AC ripple at most is comparatively easy on them.

[R.G.]

More specifically, the AC ratings of these caps take into account the heating effects of being charged and discharged almost completely on each AC cycle. Working as DC filter capacitors with a few volts of AC ripple at most is comparatively easy on them.

Yep. The thing that ages caps of all kinds is internal heating from current flowing through the ESR. In motor run caps, the full current for a motor winding flows through them. This amounts to a fraction of an amp to many amperes depending on the size of the motor.

In power supply use, the RMS current in the ESR is best conceived of as a multiple of the DC load current. For a tube amplifier, the DC B+ current is probably under an amp. The RMS heating current is on the order of 1.4 to 1.8 times the DC load current.

The heating RMS current in normal power supplies with small ripple is probably very light duty indeed to a polypro motor run cap.

And it adds even more to that "immortal" thing.

[nashvillebill]

I've looked into motor run caps from places like Burden Surplus Sales but a lot of times it's not stated whether the cap is poly film/oil or NP electrolytic. Any tips to tell them apart if it's not stated in the ad or catalog?

edit: Here's their catalog page. http://www.surpluscenter.com/Catalog/Catalog287-149.pdf

Is it safe to assume their motor START caps are NP electrolytics but the RUN caps are film/oil?

[R.G.]

Yep. If the material is not stated, run caps are film/oil and start caps are NP electro. It is entirely possible to use a film/oil cap for starting, but it's a bit more expensive, so it's not usually done that way.

Film/oil caps have lower ESR to start with, and don't degrade with heating [as badly, at least] so they're sturdier in the continuous-current path for motor run applications.

Single phase induction motors will run on a single phase once started, but have zero starting torque. The techniques of using a second phase winding with its phase shifted by either a high resistance ("split phase motor") or a capacitor are what we do to get them started.

Start caps are generally switched out of the second winding after start. That's what happens in a motor that "clicks" after it starts. Run caps are left in circuit and provide both a start up kick by phase shifting the current into a second winding, but also improve efficiency and running torque by staying in circuit and making a second "phase" from the incoming AC line.

In this application, the cap carries a substantial current into the second phase winding, and so has to be able to withstand any heating the current causes. So we can use lighter-duty electros for starting because they don't stay in-circuit all the time, and get the chance to cool down between times. Film/oil caps generate less heat and withstand it better, so they are chosen for always-in-circuit run applications.

... er, generally...

[tedmich]

As RG so aptly states these motor run PP film caps have higher DC rated voltages. The ratio is far from constant; its contingent on the caps physical dimensions, thermal behavior and a number of internal characteristics. A check of caps at mouser rated for AC and DC shows DC ratings = 1.414x AC ratings as the minimum with many approaching 2-4 fold higher. The Mfg. will have the real specs, without them it is best to design conservatively.

Axial motor run caps are made, but hard to source. They probably aren't much bigger (per uF) than huge old Atoms but will undoubtedly work better and longer. Now if Italian motor cap Mfg. Ducati just made gloss black ones with a red "Ducati Monster" graphic...

[Mick Bailey]

Never tried it, but post #4 looks like it would work to give a rough idea of the capacitance;

How to measure capacitor with a basic digital multimeter? - diyAudio