Current Sensing for dual rail PSU

Thread Starter

RobinGriffiths

Joined Aug 16, 2011
20
Hi,

I've designed and built a dual rail PSU. I now want to use an Arduino to monitor and control it. The most important feature is to display current supplied by the positive and negative rails. I was planning to use Maxim 437x high side current sense ICs.

Can I use them for both positive and negative rails by simply reversing the shunt inputs R- and R+? Since the Arduino/Maxim grounds are floating with respect to the PSU ground, I guess that would just work? Or would this mean that I'd need a low side current sense IC for the negative rail? This is the type of thing I'm thinking of.

img.png
 

andrewmm

Joined Feb 25, 2011
1,128
check the data sheet for what voltage is allowed on the Rs pins with respect to the Vcc.
I'd not expect it to like a negative voltage on the Rs pin.
 

Thread Starter

RobinGriffiths

Joined Aug 16, 2011
20
check the data sheet for what voltage is allowed on the Rs pins with respect to the Vcc.
I'd not expect it to like a negative voltage on the Rs pin.
Hi,

I think the way I've done it, RS+ is positive with respect to RS- for both rails. 0V at the top is floating with repect to Gnd at the bottom, the only 'coupling' between the two cicuits should be via the differentional current sense amplifier, which should have a high input impedance.
 
Last edited:

schmitt trigger

Joined Jul 12, 2010
414
You did not attach the datasheet for the proposed component, but what the other poster is attempting to tell you, is that the monitored current will be at a voltage below the (floating) ground. Might not be possible but..... Hard to say without the datasheet.
 
Last edited:

Thread Starter

RobinGriffiths

Joined Aug 16, 2011
20
Thanks for that. Have attached data sheet. Not quite sure what I'm looking for on data sheet, but these three under 'Absolute Maximum Ratings' maybe ?

VCC, RS+, RS- to GND-0.3V to +30V
OUT to GND-0.3V to (VCC + 0.3V)
Differential Input Voltage (VRS+ - VRS- )±8V

Not sure if these are 'the chip won't work but won't blow up' rather than operational range values.
 

Attachments

Thread Starter

RobinGriffiths

Joined Aug 16, 2011
20
Revisiting this again. Looks like the other posters are correct, RS- cannot go below -0.3V:

VCC, RS+, RS- to GND..........................................-0.3V to +30V

And indeed, when I tested the circuit tonight, it did not work and gave a worrying 8V output. Interestingly though, the chip did not blow up as when I reconfigured it for the 'normal' positive rail sense, it worked.

So, I need to rethink negative rail sense.
 
Last edited:

Ian0

Joined Aug 7, 2020
1,561
I'm assuming your supply is less than +/-18V. If so an ordinary op-amp and four resistors in the standard diff-amp configuration will do the job. If you give it a lot of gain, make sure that the common-mode input voltage is within spec.
Most of the specialist current sense amplifier ICs don't like inputs below their negative supply.
 

Thread Starter

RobinGriffiths

Joined Aug 16, 2011
20
I'm assuming your supply is less than +/-18V. If so an ordinary op-amp and four resistors in the standard diff-amp configuration will do the job. If you give it a lot of gain, make sure that the common-mode input voltage is within spec.
Most of the specialist current sense amplifier ICs don't like inputs below their negative supply.
Thanks Ian0. max voltage is +/-15V. So you're suggesting replacing the Maxim component with a differential amplifier which won't be worried about a negative voltage on the RS-. I'll give that a go tomorrow.
 

crutschow

Joined Mar 14, 2008
26,728
Here's a circuit that will accurately monitor a negative supply.
It does not suffer from the common-mode voltage errors that a 4-transistor differential circuit has (if built with 1% resistors).
 
Top