🚀 go-pugleaf

In article <4893@news.duke.edu> fang@phy.duke.edu (Fang Zhong) writes:

>	I work on small signals of less than micro volts.  I use a computer
>to control my experiment and take data via IEEE bus. 
>This equipment was designed so poorly that a ten-times larger noise was
>added to my circuit if I added it to my IEEE bus.  

	This is (alas) not unheard-of.  The standard solutions are:
isolation amplifiers (these allow a floating measurement, so you
can disconnect common grounds); true differential amplification
(rejection of common-mode voltage is pretty good if you take care);
and preamplification (so that the silly microvolt signals
are decently large..like a volt or ten... before they get computerized).

	Fiber-optic IEEE-488 links are available, but 
are expensive (and only solve half the problem: you still have
the noise in two IEEE-488 boxes to contend with on the 'sensitive'
side of the application).

	You should try isolation transformers for the AC power
on one or two of your boxes, and analyze the ground configuration
for loops, and look for unintended capacitive and inductive coupling.
It might be possible to rewire the connections to the power lines
(and the data cables) to get your noise floor down to tolerable
levels.

	John Whitmore