[Ian Iveson]

Thanks Fred

Sounds like a very good current regulator to me,
capable of a 100:1 dynamic range. You, as the
designer, control how low the dynamic R goes in
your circuit by controlling how low the dynamic
voltage across the device goes during operation
- the lower the instantanious voltage, the lower
the instantanious R required to maintain the
design current.

The same would be true of a resistor. The greater the
voltage you have to play with, the greater you can make the
resistance for a given required current.

I hadn't read far enough into the datasheet to see under
what circumstances that dynamic R might vary, so thanks for
the info.

But is 100k OK? Fine if it were a resistor in most cases,
IMO, and for most ppl maybe fine even for a non-linear
device. But I would bet that some would argue that 100k is
not great enough to guarantee that the sound of silicon
can't seep in under the back door.

OTOH, I have seen many claims that simple discreet silicon
circuits achieve many Megohms, but with no mention of
frequency response, as if it didn't exist or didn't matter.

If the ixys device achieves 100k at all frequencies of
interest, it's just about good enough, IMHO. It's certainly
convenient.

cheers, Ian