[R-390] Crystal substitution puzzle

[James A. (Andy) Moorer]

Usually the problem is the series-versus-parallel rating of crystals - most 
of the R-390/R-390A crystals are specified in parallel mode, which means 
they normally have a capacitor in parallel with them which brings down the 
frequency. Parallel-mode crystals are ground with a free resonant frequency 
a bit high. If you put a series mode crystal in a socket that expects a 
parallel mode crystal, the frequency will be low. It can be as much as 
100-200 ppm low. For a 16 MHz crystal, 100 PPM would be 10e-4 * 16 * 10e6 = 
1600 Hz. We have a smoking gun.

When you specify a parallel-mode crystal, you have to tell the provider what 
the parallel capacitance will be. If it is a trimmer or other variable 
capacitor, you generally take the mid-point of the range to give you some 
flexibility to "pull" it up or down a bit.

I have been screwed by this particular gotcha more than once . . . and in 
both directions.

James A. (Andy) Moorer
www.jamminpower.com



----- Original Message ----- 

> My R390A's Y411 (16MHz) has died.  Since it was
> cheap to try, I bought all the wire-lead 16MHz crystals
> that Mouser stocks, five in all, and put them in.
> They run, but every last one of them is 1.5-2.5kHz low,
> which is 50% out of spec at best.
>
> This is five different parts, from Citizen, Fox, ECS,
> Abracon, and Vishay/Dale, so it's not a bad part or
> a bad lot.  The original crystal, though too weak to
> mix well, is right on.  It's a disappointment, as the
> new ones are around $.50 each and I hoped to open
> up an alternate supply for about half of our second-oscillator
> crystals.  I can live with the error, but still - what's up?
>
> Dave Wise