[balou]

I was looking at some boutique parts website, and yeah, boutique parts are bad for your wallet.

so - why not build your own cap? all you need is
- nonconductive material
- conductive material

so, aluminium foil and cling wrap would theoretically be sufficient to diy your cap. but we're of course in crazy audiophile country, so something other is needed.

Paper in oil capacitors are highly regarded for their good sound quality. The oil seems to be castor oil, as it has a very high dielectric constant. And the paper itself should be as thin as possible to provide high capacitance with small size. And tissue paper seems to be the thinnest commonly available paper.

The conductive material poses another problem. Household aluminium foil is quite thick - not very good if you want a high energy density. Composition gold (a gold leaf imitation) could probably be used - the castor oil should prevent oxidation.

And of course - one could also try to obtain some other material like ultra-thin teflon foil. But good luck acquiring that in DIY quantities.

Anybody got some other ideas for the conductive material? Or comments on my castor oil/tissue paper idea?

...
did some more reasearch: castor oil has a dielectric constant of 2.6 at room temperature, teflon 2.0, and paper also 2.0.... why is the oil needed at all in paper/oil? the paper alone has a better dielectric constant

 

[Clutz]

Price/performance ratio is one. I doubt you could make a DIY capacitor that would function as well as a commercial one for the same price.

 

[balou]

Well, that's not really the point - it's the fun of building your own stuff.
And if you look at prices of 40$ for a single cap, I think you can build quite a bit for that money

Smallparts.com actually has ptfe sheets in stock for a good price - see http://www.smallparts.com/products/d...vt%2Dvte%2Ecfm
thickness 0.005"

Maybe I'll go to my drug store tomorrow and annoy the people there once more with my strange product requests

they should be able to get some castor oil for me.

I'll also have a look in the local crafts store for tissue paper and composition gold. and its also time for some calculations - how big would be the required surface for a 1uF teflon cap.. bbl searching the appropriate formula and constants

edit: for approx 1uF, 0.7 square meter would be needed - 84x84cm (very rough values). you could easily stack that a bit and fit it in a small tube

 

[cotdt]

you can make parallel plate air capacitors. it would be superior to film caps but they would be quite big.

 

[balou]

the biiiig problem with air caps is that they need to be extremely close together - meaning that the plates must be rigid, and they also need to be extremely even (if the plate distance is 0.01mm, even small deviations may lead to a direct contact). some foil definitely eases the whole process

oh yeah another problem - I don't have a tool to measure capacitance. seems like I need to buy a new multimeter

 

[cotdt]

Quote:

Originally Posted by balou /img/forum/go_quote.gif the biiiig problem with air caps is that they need to be extremely close together - meaning that the plates must be rigid, and they also need to be extremely even (if the plate distance is 0.01mm, even small deviations may lead to a direct contact). some foil definitely eases the whole process oh yeah another problem - I don't have a tool to measure capacitance. seems like I need to buy a new multimeter

ok try the next best thing, flurinert (liquid teflon). the dielectric constant is a lot lower than teflon.

 

[balou]

that wouldn't solve the problem of the plates touching each other, and also not the problem of having to rely on stiff plates instead of flexible material.

does anybody knows how the capacitance changes when stacking layers? is it (n-1) times the capacity with n layers compared to just two layers? are there significant effects caused by the plates not adjacent to the current plate? (meaning the same-potential plate two plates away, and again the different-potential plate three plates away and so forth) or do these next-plate effects cancel themselves out?

 

[cotdt]

well i know that the common way to do it is to put oil-soaked paper between the plates. you get a low-inductance capacitor for not much money. they are better than the rolled ones.

 

[Roam]

Sure you could make one, but I wouldn't bet on its long-term reliability nor stability. Capacitors need to be hermetically sealed or moisture from the air is going to be adsorbed into the constituent parts which then causes oxidation of the conductors, failure of the insulators, and all sorts of other problems. In short, there's a good chance the capacitor is going to drift wildly from its initial value, and there's a decent chance of shorts and arcing which will most likely destroy whatever circuit the capacitor's a part of.

If you want to make capacitors, stick to the pF values where you can use a length of coax cable. Tie the core to one lead and the shield to another, presto, instant capacitor.

 

[balou]

it wouldn't be much of a problem to stick it in a pvc pipe and fill it with epoxy

Quote:

In short, there's a good chance the capacitor is going to drift wildly from its initial value, and there's a decent chance of shorts and arcing which will most likely destroy whatever circuit the capacitor's a part of.

it's gonna be a input dc blocking cap for an amp - so voltages are low, and shorts wouldn't be instantly deadly to the circuit and headphones, only if the input dc offset would go unexpectedly high at the same moment (this of course is source specific)

 

[ericj]

Air is a crappy dielectric.

There are low-power HAM DIYers who make their own capacitors.

price/performance isn't the issue. It's size.

Due to the limited availability of materials, and limited tooling, you end up rolling capacitors that are quite huge.