High quality, high voltage (~250V) reservoir caps for PSU

[thadman]

Could anyone recommend any high quality, high voltage, low impedance (low ESR) reservoir caps for a PSU?

I'd prefer to avoid capacitors marketed for audio use as I assume they are overpriced (snake oil). Industrial capacitors would be mass produced and have tight tolerances assuming their application.

The Panasonic FM series capacitors are low impedance ( ESR ~.015 ohms for the 16v model) and are reasonably priced ($433 for 1000). However, they are only available up to 50V.

My input stage will produce a minimum 200V full-wave rectified DC. As a result, the Panasonic FM series will not work for my application.

Could anyone recommend any low impedance 250V reservoir capacitors?

[thadman]

Is anyone familiar with the Panasonic ED series (type A)?

The 250V 220uF model looks interesting.

http://search.digikey.com/scripts/Dk...name=P13533-ND

[Eva]

If this is a SMPS, calculate required capacitance (for minimum input voltage and desired output holding time), then calculate required ripple current (at maximum expected load), then pick suitable capacitors according to these values. Parametric search should be quite useful, such as in digikey website.

Take a look at Panasonic ED TS series (part number starting with EETED...).

[DougL]

This may be way off base but a motor run is designed to be Very low ESR and can be found in those voltages. It is large and gennerally tops out at 100 uF or less.

HTH

Doug

[thadman]

Originally Posted by Eva

If this is a SMPS, calculate required capacitance (for minimum input voltage and desired output holding time), then calculate required ripple current (at maximum expected load), then pick suitable capacitors according to these values. Parametric search should be quite useful, such as in digikey website.

Take a look at Panasonic ED TS series (part number starting with EETED...).

Originally Posted by DougL

This may be way off base but a motor run is designed to be Very low ESR and can be found in those voltages. It is large and gennerally tops out at 100 uF or less.

HTH

Doug

Yes, the design is for a smps. btw, thanks for the feedback. I really appreciate the input

Instead of passing the rectified DC straight to the buck converter, I've decided a PFC input stage would be more advantageous. However, the PFC will produce a much higher DC voltage (450V vs 200V).

I'm interested in purchasing the High-Boost Harmonic Attenuation Module from Vicor. It possesses a power factor close to unity (.99) and is pretty efficient (91% @ 115VAC). I doubt my efforts would produce something with higher performance considering the research involved in developing this module.

http://cdn.vicorpower.com/documents/...ost_vi-ham.pdf

Vicor recommends an external hold-up capacitor of 1,000 µF with a minimum voltage rating of 450 Vdc for 16 ms ride through time at full power. They do not recommend external capacitance exceeding 3000 µF.

I would assume performance would only increase with additional capacitance (less than or equal to 3000 µF). However, the transient behavior of smps is non-trivial. Is my assumption wrong?

Would bulk electrolytic capacitors suffice for such transients or would you recommend foil/film capacitors?

Assuming electrolytic capacitors suffice, would this model work?
Digi-Key - P11691-ND (Manufacturer - EET-ED2W561EA)

(Panasonic TS-ED 450V 560 µF with an ESR of 236 milliohms)

I could parallel 5 of these capacitors for 2800 µF.

If this particular Panasonic capacitor is not optimal, could you recommend a different brand/model?

Thanks,
Thadman

[mag]

Why do you need low ESR at the input of your converter?
If your 400V rail comes from a PFC increasing the capacitance only increase
the hold-up time and nothing else.
If Vicor recommends maximum 3000uF it is probably because the PFC converter feedback loop will not stable with more than 3000uF.
BTW. 3000uF for 650W output sounds crazy... how much holdup time you need?

Transient behaviour:
A PFC is a slow system, it's bandwidth is usually 20Hz or less in order to be able to correct the power factor. Because of this the transient behavior of a PFC is always quite bad; it is up to your converter connected before the PFC to provide a stiff supply with good transient behavior. Adding a lot of capacitance at the PFC output does not help at all in having a better transient behavior. The output of the PFC will always have some volts of
100Hz/120Hz ripple that you can simply not filter using a huge capacitor, otherwise the PFC can not correct the PF anymore.

Don't care about ESR for this position, yust calculate your required total capacitance (for hold up time) and the ripple current in the capacitor at full load. Then choose a capacitor with the required capacitance and the required ripple current.

There are some capaictors specially designed for PFC rated 420V instead of the classical 400V or 450V. The allows to gain some PCB area when compared to their 450V counterparts.
Take a look at Nichicon PZ series : 150uF 420V in 18x40mm case and use
4-5 of them in parallel.