0
\$\begingroup\$

I've been reading up on the concepts of properly isolating sensitive signals in applications like analog front ends.

For small signals, it is appropriate to use an in-amp. I understand that. But what about isolation?

It seems (see figure below) most flow-charts and lectures show the in-amp being the first stage of input, cascading into an isolation amplifier or barrier of some type.

enter image description here

Practically speaking, looking at isolation amplifiers it seems their topology outputs a differential signal to mitigate noise. However, what if you are only interested in observing the positive wave cycle? In this instance is it appropriate to leave the naught output floating? Pull down resistor? Or cascade into a differential amp?

TI Isolation Amp

enter image description here

From this line of thought, why is it important to first feed the signal to an inamp before isolation? If an isolation amp outputs a fully differential signal wouldn't it make sense to first isolate the signal, feeding the differential output to the difference amp and reduce the component count?

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Isolation means isolate from noise, so signal can be transfered for turner implantation. Instrumental amp normalize signal to measureable form \$\endgroup\$
    – user263983
    Jan 13 at 14:52
0
\$\begingroup\$

In general, when amplifying small signals, you but the main gain stage first to get the signal up to a usable level. This makes the system more noise immune, and also (if the source is high impedance) performs impedance conversion.

In this case, reversing the order would mean connecting the source (looks like the human body in this case) directly to the isolation amp, which probably wouldn't work at all. If the preamp here is amplifying electrodes on the human body, the preamp needs to be first.

\$\endgroup\$
3
  • \$\begingroup\$ What about using the output from an isolation amplifier? If its not feeding an ADC, but let's say some other downstream analog filtering or device, is it suitable to simply observe the potential at the "P" output? Or would the proper way to use the signal be to first use a differential (to single) amplifier? \$\endgroup\$
    – Matt001
    Jan 13 at 14:41
  • \$\begingroup\$ yes you should be able to look at the output from the phototransistor with a scope, or connect it to some other device. Output impedance may not be that low, so a buffer would be a good idea. But yoi can use it unbalanced if you want, you'll just have 6dB less signal. \$\endgroup\$
    – danmcb
    Jan 13 at 14:52
  • 1
    \$\begingroup\$ Not sure if this is conceptual or building something, but with clean enough power for the INA on the front end (i.e. required), I think it would now be more common to just convert the INA output to a frequency (or even more likely ADC) and use a digital isolator chip for the signal to the rest of the system. \$\endgroup\$
    – Pete W
    Jan 13 at 15:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.