[time-nuts] Frequency counter recommendation

[Magnus Danielson]

On 12/19/2010 09:57 AM, Poul-Henning Kamp wrote:
> In message<28E8C13C-3227-4CD6-BBEA-902F03C81243 at gmail.com>, bownes writes:
>
> One of the big problems with "PC-instruments" is isolation: most of them
> havn't any.
>
> If you are going to design one such, remember to isolate the USB
> port galvanically from the actual instrument...
>

To illustrate the problem:

A friend of mine do electronic designs, among those power supplies, some 
of those has relatively high voltages.

When measuring one he had his oscilloscope hooked up, which was bad 
since he also had the RS-232 hooked into his computer... which meant 
that he ended up toasting the IO-board of the scope as well as his 
computer... with lots of lots of work on it.

He has recovered, but it was unwanted setback of operations.

To achieve meaningful isolation, it should be optical. Transformer 
isolation has a few issues to get right...

But then again, for this kind of signal, the voltage levels involved is 
usually low, except for some few voltages in cesium-tubes and mains.

Cheers,
Magnus