Basics of using the isolating transformer

[meereck]

Hello, I have a differential signal going into an instrumentation amplifier (AD620).
I would like to employ galvanic isolation on the inputs of AD620.
Would anyone be so kind and tell me how it shall be wired?
I don't know much about signal isolation. Shall 1:1 or 1:1:1 transformer be used?
Can I also employ transils for protection - where to place them?

Thanks in advance,
regards Meereck

 

[dlgoff]

We need to know a little more about what the input signal looks like. What's it amplitude and frequency, etc.

 

[meereck]

hello, differentially, it is basically a sine wave in the range between 50-100kHz
Amplitude about 0.5 Volts. I use AD620 for amplifying the signal.
thanks, M.

 

[berkeman]

hello, differentially, it is basically a sine wave in the range between 50-100kHz
thanks, M.

50Hz-100kHz, or 50kHz-100kHz? It makes a big difference. If b), then you can use a 1:1 transformer designed to operate in that frequency band. If a), then you will not be able to use a transformer for such a wide range of frequencies. Instead, you would probably be better off isolating the full input amp & ADC circuit, and optically communicating with the ADC to get the data.

 

[Pumblechook]

How about a linear optoisolator?

 

[berkeman]

How about a linear optoisolator?

Optos aren't linear, generally aren't used for analog coupling. Digitizing first and sending the digital date through an opto would be fairly common.

There were some linear optos a while back (maybe from Linear Technology?), but I don't think they found any volume applications. Pretty pricey chips, as I recall.

 

[Pumblechook]

Agilent make some.

http://dev.emcelettronica.com/voltage-conversions-galvanic-isolation

3.7 Euros in France..

HCNR201

 

[meereck]

thanks guys for ideas,
the frequency will be 50kHz-100kHz.
I read something about using 1:1:1 transformer for differential signals. Does it make any difference?
Should some decoupling capacitors be used?
have a nice weekend.
M.