Would this actually work in real life? It doesn't seem to in LTSpice. (or maybe it can somehow???)
Basically I have 2 power sources that can range from -75V to 75V.
I must keep each supply (specifically the returns) isolated from the other (hence the FWB)
I need to scale the voltage of each supply to 2.5V -/+2.5V to be read by a micro. (where 0V on A or Bout represents -75V supply voltage and 5V on A or Bout represents 75V supply, and 2.5V on A or Bout represents 0V supply)
The orD net will be used on another sub circuit (not shown)

If not can you give suggestions on would you accomplish the goals stated above.

 

It won't work because the bridge provides a path to ground through the diode. Or will the two voltage's commons never be more than a diode drop apart in voltage?

I don't understand your simulation circuit. You should be taking the signal from each bridge output. What is the connection orD?

If the output of each circuit needs to be proportional to the 75V value then you could use an isolation type op amp such as one of these or these.

 

How about this?
This actually works in LTspice

 

So you don't care that the two supplies are isolated by only a couple diode drops?

 

So you don't care that the two supplies are isolated by only a couple diode drops?

The requirement is to have "return path isolation"
I assumed that would meet that requirement.

If you have a better circuit to accomplish the task I'd love to see it..
I'm far from an electronics expert and am just stuck with this task and am obviously over my head here..

oh and by the way orD is used by another sub-circuit that I didn't show..That sub circuit basically just requires a positive voltage and is redundant in case one of the supplies is off.

 

The requirement is to have "return path isolation"
I assumed that would meet that requirement.
............................

That can be a dangerous assumption. You need to know what the reason is for the isolation, and how much isolation voltage is needed.

 

The diode drop isolation is suitable for this application.. The isolation is needed to keep the other supplies return path from becoming a fault path during a supply fault on the other supply.

Now last question.. The second schematic I posted works perfect on breadboard BUT how would I add gain and offset calibration to it?

 

The diode drop isolation is suitable for this application.. The isolation is needed to keep the other supplies return path from becoming a fault path during a supply fault on the other supply.

Now last question.. The second schematic I posted works perfect on breadboard BUT how would I add gain and offset calibration to it?

The fault current will only go to the supply with the fault whether you have the diodes or not. It just depends upon how the return lines are connected to the supplies as to what the current path is. If you run a separate return line from each supply to the load then the fault current will only go on the power supply return line with the fault. So I don't see that the diodes are needed. I need to see your complete circuit with the power supplies to understand.

You could add offset calibration by adding small value pots (connected as a rheostat) in series with R6 and R10 and change R6 and R10 to a smaller resistance (for example a 500Ω pot in series with 3KΩ). This also changes the gain so you want to do offset adjustment before adjusting the gain.

To avoid interaction with the offset you could add two more op amps with gain on the outputs of U1 and U1 with pots to adjust the gain.

 

Thanks crutschow.. I really appreciate the help.
I'll try to find out more about the reasoning for this isolation and post back if I hear something.. Basically I have been told by the end user that simply adding oring diodes on the supply and return lines is sufficient to meet their requirement.

 

Well I got an answer..sort of...They just want to have diodes in the power and return paths for each supply so that each supply is in effect isolated by the diodes being reverse biased when one supply is off or at a lower voltage. Then something about preventing one supply back-feeding the other.