Ultra Low Profile Capacitor? 1

[metalman89]

I'm working on an AC to DC power supply application that is very space limited in terms of board thickness. I am looking for a 400V ~10uF cap that is no thicker than 0.090". I have much more room in terms of length and width but the thickness of the cap itself cannot be greater than 0.090". Are there any off the shelf solutions that I could use to solve this problem? I've checked popular sites like Digikey with no such luck. If not, do I have any options in terms of custom cap design?

 

[itsmoked]

Look for 450V SMD caps and parallel a bunch?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[metalman89]

At that point I would need at least 100 parts in parralell which isnt an option due to size and cost. Any other ideas? What about a custom aluminum electrolytic capacitor that is large in the X and Y direction but very thin in the Z?

 

[ScottyUK]

If using 100 commercially-available parts isn't viable due to cost then you definitely can't afford custom work.

What ripple current are you expecting this capacitor / capacitor bank to handle, and at what frequency?

 

[metalman89]

Is there any reason why a radial aluminum electrolytic capacitor couldn't be custom made flat to a pancake shape. Essentially the total surface area of the internal plates would be the same but the overall shape of the capacitor would be much smaller. Is there anything that would theortically limit this design?

 

[IRstuff]

Consider the fact that the diameter of such a capacitor is proportional to the number of layers and layer thicknesses. That means that without some sort of radical change, the thickness is fixed.

TTFN
I can do absolutely anything. I'm an expert!
faq731-376 forum1529

 

[VEBill]

You could gain some headroom by cutting a window in the PCB and installing the component(s) within the opening (gaining at least the PCB thickness in the height). One sees this all the time in consumer electronics (e.g. often with very small transformers).

This won't solve the issue, but it might help towards the end.

 

[MikeHalloran]

The aluminum needs some finite thickness in order to conduct electrons toward and away from the terminations.
The dielectric film needs some finite thickness to store the charge.
Your university physics book should have an equation that tells you the square area you need for dielectric and plates. The voltage determines the dielectric thickness.
So figure how big a sandwich you need, and then fold it to fit the volume available.



Mike Halloran
Pembroke Pines, FL, USA

 

[VEBill]

"400V ~10uF"

To figure out the rough volume, you might try the following.

Interesting tidbit: 400v 12.5uF = 1 Joule

Ref.

http://www.electronics2000.co.uk/calc/capacitor-charge-calculator.php

"Highest Joule/cc J-CAP™ Conductive Polymer cathode solid electrolytic chip capacitors ... the highest energy per volume (Joules/cc) of any comparable SMT capacitor currently available on the market."

Ref.

http://www.avx.com/products/polymer/j-cap-series/

To start, ignore the practicalities of the required series/parallel design, just focus on determining how many cubic cm of their products (AVX as an example) would be required to reach 400v and ~10uF.

Series capacitors for voltage is obviously a bother, but might be feasible. Perhaps with equalizing resistors or similar.

At least you'll get a rough idea of the required volume.

Give some indication of the feasibility.

 

[sreid]

Maybe here.

http://www.avx.com/products/polymer/high-reliability-polymers/tch-hermetically-sealed-polymer/