Mylar caps

[Paul Schreiber]

Mylar suffers from the following:

1) they have whats called a "memory function". This is similar to nicad batteries. What happens is that if the cap
sees a small constant charge, it "holds" that charge, sort of like a DC offset. Then when a large charge goes in
or out, the voltage is not correct.

2) they do not age gracefully. they tend to absorb moisture. the values drift all over the map.

The reason you see so many is that the technology did not exist to make high volume, low cost non-Mylar film
caps until aout 1979. Back then, a good (say 5%) 50V film cap cost about $1.50 in 100,000 pieces. Now,
you can get Panasonic metallized polyproplyene caps from Digikey for $0.50 for 1.

Funny (true) story follows=========================================================

I was at Tandy in 1980 designing a 300 baud (ROTFL) modem. I needed 12 1% 4700pf film caps per modem.
That was about 450,000 a year total. I called all of the US suppliers at the time (ElectroCube, Sprague,
Mallory) to get quotes. The lowest was $.78 for a 2% and $1.01 for a 1%. A 2% Mylar at the time was about
$0.32.

I called Panasonic in Japan. They had just come on-line with the ECQP series (Digikey has theses as well).
I asked for "a few samples". They sent me 500! I asked for a price. It came back $0.16! I then sent a fax
saying please requote, there must be an error (I had $0.28 in the price target). They requoted at $0.14!

Over the next 4 years, Tandy bought 5 million of these caps.

Paul Schreiber
Synthesis Technology


----------
From: 	David Halliday (Volt Computer)[SMTP:a-davidh at microsoft.com]
Sent: 	Monday, August 25, 1997 4:56 PM
To: 	'analogue at hyperreal.org'
Subject: 	RE: Concertmate internal trimpots

Bonehead question here 

I have seen this before about avoiding Mylar film caps - why???


> -----Original Message-----
> From:	Paul Schreiber [SMTP:synth1 at airmail.net]
> Sent:	Sunday, August 24, 1997 3:34 PM
> To:	analogue at hyperreal.org; 'Joseph Hughes'
> Subject:	RE: Concertmate internal trimpots
> 
> 0.1mfd is a tad on the low side. The low frequency response will
> suffer (say middle C and below). A 0.47mfd is
> just as easy to get and has almost 5 times lower frequncy cutoff.
> Virtually the same price, too.
> 
> Just be sure to get a film cap. Avois non-polar electrolytics and
> ceramic. Avoid Mylar film like the plague. Use
> a polycarbonate or polyproplyene film. Both from Mouser or Digikey.
> 
> Paul Schreiber
> Synthesis Technology
> 
>