metallized polypropylene capacitors for DC supply 1

[fuseshut]

Hi I am looking for a metal pol capacitors for DC supply, I am rectifying 28VAC with a bridge rectifier (more efficient than single diode) and will attempt to "smooth" the rectified voltage with a capacitor bank. I have looked at some metal poly capacitors. I would like 60-70WVDC at 1000uF, but in a small form factor. for example, something smaller than

http://www.icwltd.co.uk/smf.html#

(click on technical details).
Any help on vendors would be greatly appreciated.

 

[ScottyUK]

This sounds like a peculiar application, or maybe an ordinary application with a confused designer? Why use a polysomething^* capacitor for what sounds like a reservoir application? An electrolytic would be the usual choice, with a low ESR film capacitor in parallel if required, especially if size is at a premium. Why is this unsuitable?

^*Polysomething: Polycarbonate? Polystyrene? Polyester? Polypropylene?

Do you need very low ESR and ESL? Very high ripple current rating? Very high ripple frequency? Elevated temperature? There is a lot of detail which is missing - it would be easier to help if we knew what was going on!

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[ScottyUK]

Apologies for the polysomething comment: just noticed it in the subject line .

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[fuseshut]

Well I am not a confused designer!!! Maybe need to give more detail. The AC-DC rectification will take place in a junction box that is filled with oil and box is mounted on a underwater ROV,that goes 15000feet down in the ocean. Electrolytic capacitors are best for this AC-DC application but when dealing with air cavities under pressure, electrolytic caps FAIL!! That's why I am looking for metallized polypropylene capacitors, which are wrapped or epoxied in a moisture resistant epoxy resin

 

[nbucska]

I suggest making an air filled custom case for the electrolitic cap and/or using higher ferquency e.g.
400 Hz, reducing cap-s and transformers etc.



Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[itsmoked]

Air filled?? This is for use in a liquid filled 5000PSI environment.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

An air filled container will probably collapse - that seems to be what the capacitors do. It will probably be tough enough to bring 50 Hz down a 15000 feet cable. 400 Hz next to impossible. If you have a possibility to use three phases, you will have a lot less ripple if you use a six pulse rectifier. Needs very little capacitance for smoothing.

Gunnar Englund

http://www.gke.org

 

[itsmoked]

Good points skogs. (Did you say you were a cable salesman?)lololo

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[nbucska]

yes,Keith, air filled. A 1000uF 63V ( i looked up this one)
is 10x25 mm. A solid iron cylinder of this size would be
less than 50 gram.

It must be air filled to remove the pressure from the C.
I didn't calcilate it but I am willing to bet that less
than 4 mm. thick wall will be enough.






Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[nbucska]

I think the rest of the electronics won't be able to
survive this pressure either e.g. transistors and IC-s.

Why not use DC ?



Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[Skogsgurra]

There are comparatively thick walls and very little, if any, hollow room in those other components. They will probably surwive.

DC sounds good. But there is probably a voltage drop issue involved.

What about supercaps? They are compact, no voids. Need to use several in series and a potential equalizer, though.

Gunnar Englund

http://www.gke.org

 

[nbucska]

About the pressure:
There are reflectors on the marker guaranted to 6000 m
with glass lens - see at <

http://www.deepsea.com/sa_hmi.html>

-- so it is possible to build a case for 5000m.




Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[OperaHouse]

Wirs insulation wise, you are much better bringing down a higher voltage DC and using a switcher to get the voltage you want. There are small epoxy switcher modules that operate at 1MHz that should have suitable internal components that can take pressure. This would be a good quistion to ask a manufacturer like VICOR.

 

[BrianR]

There is some similarity here to the electronics contained in the Crash Survivable Memory Module in a black box flight recorder which under the ED-112 specification must survive a month at 20,000ft (60 MPa) water pressure.

I agree with nbucksa that the rest of the electronics will need some help to survive this pressure. One technique used in CSMU's is to contain the electronics in a titanium container filled with silicon gel. I don't know of any small metal pol capacitors though and so the suggestion of a 3 phase supply has a lot of merit as the filtering requirements are a lot less. 400Hz would simplify things even more but you will need to treat the 15,000 ft cable as a transmission line to control the losses which would otherwise be large. There are a few challenges here.

 

[zeitghost]

I worked for a semiconductor company that supplied ceramic DIL ICs with a metal lid.

For use under pressure (downhole), the lid had a hole in it and the cavity was filled with silicone oil of some kind.

 

[fuseshut]

Many IC components have no issue in surviving very great depths. It's only the electromechanical relays, electrolytic capacitrs, crystal oscillators, anything with an air void in it.

 

[ScottyUK]

Icar have a really good range of polypropylene caps, mainly for the power electronics market. I'm not sure if they make anything at the low voltage rating you're after, and the volume of the HV stuff is likely to be significantly greater. Probably worth a look, or an e-mail to their technical people. Evox-Rifa are or were a partner with Icar: I have bought Icar caps through Evox-Rifa before.

Are you tied to 28V AC? If this is for a production part have you considered a hybrid as a power converter? They can be packaged to suit just about any environment.


----------------------------------

I don't suffer from insanity. I enjoy it...

 

[nbucska]

1.) If the input is high VTG DC, you don't need large cap.

2.) At 400 Hz the 15,000 ft is .006 lambda - no reason to
handle it as tr. line.

3.) BRIAN: if the Ti container carries the pressure the Si gel is only for heat xfer.

4.) What about the communication signals, connectors, cables? Fiber optics or Cu ?

5) I submit that each component must be evaluated for the
pressure and the design is far from trivial...



Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[BrianR]

nbucksa,

Have to agree with all you say although wrt the CSMU the other specs it must pass include 2 hr @ 1,100 degC + 9 hr @ 260 degC + absolutely huge shock and vib so the silicon in that case is not for heat transfer but impact protection and heat blocking. Virtually no internal heat is generated anyway.

 

[nbucska]

Whoa ! this is a horse of different color !
at 1100 C Au,Al,Cu are molten and there is no polypropylene







Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>