analog isolation

[carmusic]

does it exist an IC that would isolate an analog signal (voltage input) to a voltage ouput
i know that there are isolations cards, but they are too expensive and i need to make my own pcb with isolate analog inputs (about 6 per pcb). an isolation of 30 volt would be enough (i dont not need 2000v isolation).

could i do that with opto transistor? i need a high precision (+/- 5 mV) over a 10volt range

 

[ronsimpson]

https://www.electro-tech-online.com/custompdfs/2008/04/iso122.pdf
https://www.electro-tech-online.com/custompdfs/2008/04/an116.pdf
**broken link removed**
https://www.4qdtec.com/aoi.html

Go to fig. 4 for diff amp
**broken link removed**

 

[Papabravo]

Get ready to unlimber your wallet for an ISO122. They are about $16.00 each, so 6 channels wil cost you $96.00. I'd be willing to bet modest sums that you don't really need isolation on your analog inputs. Somebody just convinced you that you did and nobody ever got fired for using it. I'd reevaluate the requirement at those prices.

 

[Hero999]

What bandwidth do you require?

If it doesn't include any DC, then use a transformer.

 

[Styx]

hcpl7800

...

 

[dknguyen]

Agilent has an analog opto isolator that does that. Or search for "analog opto isolator" on Google. Otherwise use a transformer if it's just AC.

 

[crutschow]

What is the isolation requirement? A difference amp such as a TI 1NA146 can withstand a common-mode voltage of 100V with gain adjustable from 1 to 100 and costs a few $. That may work if you don't need absolute isolation.

 

[Hero999]

He said it's 30V, which is nothing.

You're right, some op-amps will do.

 

[Papabravo]

Like I said -- It's all in the requirements.

 

[Dean Huster]

Clairex was the manufacturer of analog optoisolators usually based around LEDs "shining" on LDRs. Clairex was sold, I believe, to Perkin-Elmer. They really had a fascinating product line according to the old databook I had. Remember that you don't have to rely upon a commercial product for isolation. You can roll your own and before the advent of actual optoisolators, manufacturers did just that. There were some electronic organ manufacturers that used an incandescent lamp lighting an LDR (shielded in a light-tight tube, of course), not for isolation, but as the "swell pedal" volume control. The swell pedal turned a potentiometer which in turn varied the lamp brightness. By using the lamp/LDR combo, noisy pots did not transfer their noise into the amplifier. Pretty cute, I thought.

Dean

 

[Styx]

He said it's 30V, which is nothing.

You're right, some op-amps will do.

yes for 30V the correct opamp will surfice BUT isolation amps arn't just used for HV stuff

WE use them when we need to send an analogue signal and we do not want to connect the grounds of two systems together, even if they only differ by 2V (differentially when separate) usually due to one system being noisy

yes we also use analogue isolators (and a hell of alot of 6n137's) for HV stuff

 

[Hero999]

But if the input impedance of the op-amps are high (>10M), you've only be effectively connecting the grounds together with a >10M resistor.

 

[jbeng]

Clairex was the manufacturer of analog optoisolators usually based around LEDs "shining" on LDRs. Clairex was sold, I believe, to Perkin-Elmer. They really had a fascinating product line according to the old databook I had...

They call their devices a "Vactrol". **broken link removed**
Jeff