[sdiy] cap acronyms and picking caps

[Martin Czech]

Today picking resistors is quite simple (for our purposes): look for
the tempco, wattage, size, metal film only and there you are. Not with
capacitors.  I know "picking capacitors" by March et. al. In practise
however there are lots of acronyms, very confusing. So I tryed to figure
it out.  Under "Name" we find the acronyms often observed at German
manufacturers, I don't know if there is any significance outside German
speaking regions.


Prefix MK: metal vaporised on foil
       PK: metal vaporised on paper
       FK: metal foil
       
>From that I'd say: for low voltage, precision work you can go until MKP or
MKY, depends also on manufacturer.  Best seems polystyrol, but the larger
values can not be obtained with that, so one has to rely on polypropylen.
Of course, 'lytics have their application, and paper foils have self
healing. High loss factor may damp power rail ringing better then low loss
factor. That is not the point here.  This is for filters and timing caps.


   
Name   Name2     Insulator                  loss factor 1e-4
MKS    Styroflex Polystyrol                 <2
       PE        Polyethylen                2
       Teflon    Polytetrafluoroethylen     <2
MKY    Styroflex Polystyrol                 5
MKP    PP        Polypropylen               8



MKC              Polycarbonat               10
PKP              Paper                      10
PKT              Paper                      45
MKT    Mylar     Polyethylenterephtalat     50
FKC              Polycarbonat               50
       Tantal                               50 
FKS              Polyester                  55
MKS              Polyester                  65
       PC        Polycarbonat               100
       Alu Elko                             >500
       

Very often one gets a bag full of unknown stuff, eg. the electronic
shop round the corner. Or you might buy a few samples, just to be sure.
So it would be interesting to test these components DIY.

I could think of a dc test (droop due to leakage or "hysteresis")
with noninverting MOSFET opamp buffer. Marsh descirbes 5min load, 5 sec.
discharge, then 1 min observe. 

Loss factor is often speced @
1kHz. Now, how to measure that without expensive bridge circuit?
Marsh describes a passive LP filter driven by sine sweep.
The voltage minimum over the cap is the point of serial resonance,
where the impedance get's real, thus showing series resistance.

Any usefull comments?



m.c.