Current sensing with OP AMP

Thread Starter

LewisMF

Joined Nov 15, 2014
100
I'm currently designing a short circuit detection circuit for a bench lab PSU.
When the PSU's output goes above 1.5A the short circuit detection circuit will disconnect the load from the output and latch so until the load has been removed.

This is my current circuit:

Sin título.png

Basically I pass the output through a 1Ohm power resistor and feed the voltage on both sides of the sense resistor into an op amp in differential configuration. After that the output of op amp 1 goes into the non-inverting input of a second op amp that is configured as a comparator against a reference voltage adjusted by a pot .

The problem is that op amp 1 is not working as it should and I think the issue might have something to do with the grounding.

The op amps are supplied by a separate 12V supply but the two voltages going into the differential op amp are referenced to the PSU ground.

Is it ok to feed in voltages that are referenced to a separate ground?

My question might sound a bit obvious to some people but I'm really not sure what the issue is.

All help is welcome :)
 

wayneh

Joined Sep 9, 2010
16,164
Is Rsense in the drawing supposed to be your 1 ohm shunt? How do you expect to drive 1.5A through it?

You could use a single comparator to trip based on the current through your sense resistor.
 

mvas

Joined Jun 19, 2017
538
Last edited:

Thread Starter

LewisMF

Joined Nov 15, 2014
100
No, the circuit will not function properly unless the psu ground and 12V ground are connected together. If they are not, the voltages on the inputs of the op amp are indeterminate.
Thank you OBW0549 - this was the answer I was looking for.

I'm going to build a small circuit to get 12V from the input stage of the PSU (Before the main regulator), this way they will share the same ground.

I'll come back with the answer :)
 

Thread Starter

LewisMF

Joined Nov 15, 2014
100
Is Rsense in the drawing supposed to be your 1 ohm shunt? How do you expect to drive 1.5A through it?

You could use a single comparator to trip based on the current through your sense resistor.
1.5A x 1ohm = 1.5V drop.
1.5V x 1.5A = 2,25W
With a 10W resistor I can drive it with no problem at all.

First I need a differential op amp to read the voltage drop across the resistor to know how much current is flowing. The differential op amp outputs a voltage proportional to the current flowing (as I know the resistance is 1ohm) and this is fed into a comparator and compared against a reference voltage, in my case 1.5V. If the voltage coming out of the differential op amp is higher than 1.5V, then I know that more than 1.5A is flowing through the circuit and can trigger the next stage of the circuit to disconnect the load.

Hope this makes sense.
 

Thread Starter

LewisMF

Joined Nov 15, 2014
100
How is the Load connected ?
The "R-Sense" resistor must be in series with Load
It can be either on the High Side or Low Side.

A simple High-Side Current Sense ...
https://www.electronicdesign.com/sites/electronicdesign.com/files/uploads/2015/12/0316_CTE_Maxim_F1.gif

A simple Low-Side Current Sense ...
https://e2e.ti.com/blogs_/b/analogwire/archive/2018/01/22/low-side-current-sensing-for-high-performance-cost-sensitive-applications
This is how the load is connected:

Sin título2.png

I'm measuring on the high side.
 

ian field

Joined Oct 27, 2012
6,539
I'm currently designing a short circuit detection circuit for a bench lab PSU.
When the PSU's output goes above 1.5A the short circuit detection circuit will disconnect the load from the output and latch so until the load has been removed.

This is my current circuit:

View attachment 173502

Basically I pass the output through a 1Ohm power resistor and feed the voltage on both sides of the sense resistor into an op amp in differential configuration. After that the output of op amp 1 goes into the non-inverting input of a second op amp that is configured as a comparator against a reference voltage adjusted by a pot .

The problem is that op amp 1 is not working as it should and I think the issue might have something to do with the grounding.

The op amps are supplied by a separate 12V supply but the two voltages going into the differential op amp are referenced to the PSU ground.

Is it ok to feed in voltages that are referenced to a separate ground?

My question might sound a bit obvious to some people but I'm really not sure what the issue is.

All help is welcome :)
Vaguely recall seeing an application bulletin for an off the shelf op amp designed for that very purpose. Unfortunately - my memory isn't getting any younger.
 

wayneh

Joined Sep 9, 2010
16,164
Vaguely recall seeing an application bulletin for an off the shelf op amp designed for that very purpose. Unfortunately - my memory isn't getting any younger.
There is such a thing as a “fully differential opamp”. Is that it? I can’t understand the data sheets but it sounds like it might be.
 

Thread Starter

LewisMF

Joined Nov 15, 2014
100
It was left as an exercise for the student - curiosity didn't motivate me enough to do their research for them.
I already did the research beforehand and replied - I'm measuring on the high side :)

The reason for using conventional op amps and not a 'ready made' chip is that you have more control and understanding of the circuit using the old school op amps.

Nevertheless, thank you for the reply. It could be useful for future projects!!
 

crutschow

Joined Mar 14, 2008
24,125
Here's a high-side current monitor using a rail-rail op amp that doesn't depend upon resistor matching to minimize common-mode rejection error.

upload_2019-4-1_12-16-57.png
 

wayneh

Joined Sep 9, 2010
16,164
I already did the research beforehand and replied - I'm measuring on the high side :)
Unless you're resolving to connect the grounds in some manner, I'm not sure you're measuring "high side", you're measuring a differential voltage with neither side defined relative to opamp ground. If you can connect the grounds, then the problem is no longer a problem.
 

Thread Starter

LewisMF

Joined Nov 15, 2014
100
Unless you're resolving to connect the grounds in some manner, I'm not sure you're measuring "high side", you're measuring a differential voltage with neither side defined relative to opamp ground. If you can connect the grounds, then the problem is no longer a problem.
Yes I connected the grounds and the circuit is now working correctly.

This is the final schematic:
Sin título3.png
 
Top