Polycarbonate Capacitors

[bostonman]

I have one of those Capacitance Decade Boxes that is identical to the resistor type, but obviously capacitors.

Recently I discovered the x00 nF (4th dial) has the same capacitors as the 3rd dial, therefore I can't select 100nF - 900nF. It's really odd because the board doesn't look like it was altered, yet, I can't believe it was manufactured wrong.

In any case, looking at the parts list, they use 10nF, 20nF, 30nF, and 40nF, 100V, polycabronate. I did a search on Mouser and Digikey, but I can't locate PC, or, if I do, they are not in stock. I'm uncertain which category PC falls under (I think film), so maybe I'm searching wrong.

I'd like to fix this box if possible because I'm planning to order some parts including fixing some burnt resistors on the resistor box.

If I can't buy PC, what else would be adequate?

[bdunham7]

PPS is supposedly the 'correct' replacement for polycarbonate, or so I'm told, but I personally would use any decent film capacitor like polypropylene.  You may find it harder to get those exact values and may have to use multiple 10nF caps in parallel to get 20,30 and 40.  What precision is required?  1% might also be tough to come by from normal sources.

[bostonman]

Looks like:

<10 μF: ±(4% + 3 pF)
>or equal 10 μF: ±7%

Repairing this isn't a big deal. I dislike owning something that doesn't work, however, it's rare I use a decade capacitor box (the long leads don't help with the precision). Replacing the capacitors isn't difficult, so thought to buy a few for the heck of it and correct the issue. If the solution is too difficult, no sense spending valuable time on it then.

Unfortunately space is a big problem. The capacitors on a good one are approximately 0.5" height, 0.5" wide, and about 0.26" thick.

[bdunham7]

Unfortunately space is a big problem. The capacitors on a good one are approximately 0.5" height, 0.5" wide, and about 0.26" thick.

Are they axial, radial or SMT?

[Kleinstein]

Polycarbonate caps are obsolete and AFAIK no longer made regularly.  AFAIR there main advantage was a slightly smaller temperature dependence.
Polyester, poloypropylene or polystyrene caps would be OK for a replacement. values of 20 nF , 30 nF and 40 nF are not common (39 nF may be close enough). One may have to use parallel combinations. For a fine trim C0G capacitors would also be OK and some 1 and 2.2 nF caps are available also as small THT parts.  A few parallel 10 nF SMD ones on a small piece of PCB would also be an option.

[bostonman]

Are they axial, radial or SMT?

Axial (rectangular in shape and through-hole).

Polycarbonate caps are obsolete and AFAIK no longer made regularly.

Makes sense why I struggled to find them. I thought they were categorized differently and I wasn't looking in the correct areas. I'll have to look at some other options. If I can get away with values and voltage ratings (along with accuracy) on a SMT, I'm sure I can piece together something to mount a board to the existing PCB.

[bdunham7]

Axial (rectangular in shape and through-hole).

You could probably piece together a 'pretty good' solution with these C0G MLCC 2% units from Kemet. 

https://www.mouser.com/ProductDetail/KEMET/C420C103G1G5TA?qs=sGAEpiMZZMs7ZEmUmaUL06SW6ZQcwDWy%252BmPcBuMuNCEifyGIvXUDpg%3D%3D

[Benta]

Polycarbonate caps are obsolete and AFAIK no longer made regularly.  AFAIR there main advantage was a slightly smaller temperature dependence.
Polyester, poloypropylene or polystyrene caps would be OK for a replacement. values of 20 nF , 30 nF and 40 nF are not common (39 nF may be close enough). One may have to use parallel combinations. For a fine trim C0G capacitors would also be OK and some 1 and 2.2 nF caps are available also as small THT parts.  A few parallel 10 nF SMD ones on a small piece of PCB would also be an option.

Polystyrene is also obsolete.

[Conrad Hoffman]

Polypropylene is nearly perfect, other than they tend to be physically large. Mylar will have higher dissipation but it usually doesn't matter. COG is excellent if you can get the values and voltage for a reasonable price. Prior to MLCC you couldn't get higher values.

[TimFox]

Mouser has only one polycarbonate capacitor in stock (a huge 3x 154.1 uF unit rated at 37 A).
However, they have 25 polystyrene capacitors (all Xicon brand) in stock, from 150 pF to 10 nF.

[Vovk_Z]

I would possibly use a bunch of any (I mean 5%-10%)  COG/NP0 0.1 uF (50V). They are available and usually quite accurate (+-2..3%), plus additional 1-2-3 nF NP0 caps in parallel.
I guess something like 20-50 pcs is needed. It depends on exact circuit how they are connected.

[Gyro]

I remember we used to use parallel combinations of Polycarbonate and either MKS or MKP (can't remember which now) to achieve a virtually zero temperature coefficient over an extended industrial temperature range.

[bostonman]

Thanks.

Think I'm going with Polypropylene through-hole and call it a day.

After all, this is only a capacitance box, so using it to 'select' a capacitor in a circuit isn't going to be very accurate; especially if it's sensitive RF.