rin67630 wrote: ↑
Tue Aug 06, 2019 11:03 am
For industrial 4..20mA you need galvanic insulation
, since you may have got other instruments in the loop.
If you tie one pole to GND you might shortcut them.
Generally I agree that Optical isolation is most often useful, still some implementations will need it, others won't.
if there is just one 4-20mA source on the other end of the loop, which is powered by the loop, then you might not need optical isolation.
Also, the 4-20mA analog loop doesn't have to be where the isolation happens.
In my example the optical isolation is not on the 3-20mA side, but on the digital interface side, using a digital isolator ADuM5401 which also provide enough power over the isolation to power the ADC and analog logic.
But no, I'm not employed in this industry, I only have encountered a 4-20mA loop once in my career, working on a digital sunlight sensor.
For an analog isolator I would prefer to use an IL300 opto isolator that has a LED irradiating two photo transistors, one phototransistor used on the input side as a feedback for the LED current, with both transistors getting the same amount of light, so drawing an equal amount of current. This way you can control the current draw of the isolated phototransistor. Such an opto isolator doesn't cost much, and I think is cheaper than two dual pole reeds plus the vibrator logic etc.