Date: Wed, 06 Mar 2002 10:13:41 -0800

Author: "Shang, Yalan"

Subject: RE: F.G. with floating voltage


Thanks all for your suggestions and ideas. Yes, we are trying to avoid
introducing a second ground in the circuts when using an oscilloscope.
Probebilly it's a good way to use a differential amplifier on the
oscilloscope. It would be impossible to find a good
answer without your input. I appritiate your help very much.


-----Original Message-----
From: Dennis C. Henry []
Sent: Sunday, March 03, 2002 8:38 AM
Subject: Re: F.G. with floating voltage

At 11:20 AM 3/3/02 -0500, you wrote:
>David Kardelis wrote:
> >
> > I have been using standard FG but put it on a adaptor that is used for
> > attaching 3 prong connectors to older 2 prong outlets. This leave the
> > floating. Not elegant but it works.
>Please don't do this. Not only inelegant, but rather unsafe.
>Presumably Yalan's problem is that an oscilloscope or something
>introduces a second ground somewhere in the circuit? If that is the
>case, then this fix, or using an isolation transformer (which amounts to
>the same thing and is offering no safety when used like this) is sort of
>but only when the demonstrator is the Prof. who has been (we hope!)
>fully informed of the situation.
>Do it for a lab set up and you're just asking for trouble.
>Leave the signal generator grounded and use a differential amplifier on
>the 'scope input to prevent it from introducing a second ground. You
>can put an Analog Devices AD620 (or similar) into a box and power it
>with a couple of batteries for just a few dollars a pop.
>If the output of the signal generator MUST float, then put a
>small-signal isolation tranformer on the _output_ of the signal
>generator. The bad news is that these transformers belong to the same
>era as those floating output signal generators, and are hard to find.
>You can get them from Jensen Audio, since they are still used in
>high-end audio recording equipment, but they are painfully expensive. I
>have not found a cheaper source.
>-Geoff Nunes

Isolation transformers for the output signal also don't handle
square-triangle-sawtooth waveforms well. They tend to "round and
ring." If only sine waves need to be handled, then they are O.K., within
their stated bandwidths.

Military surplus radio equipment was once a good source of audio signal
isolation transformers, as was Western Electric academic giveaways. If I
had to start looking now, I would dissect some old modems.

Dennis C. Henry Office Location: 213 Olin Hall
Professor of Physics office phone: (507) 933-7314
Gustavus Adolphus College office fax: (507) 933-6104
St. Peter, MN 56082-1498 home phone: (507) 931-2784