High-Side current monitoring with an LM358

[neddie]

Hi to all.
If I want to monitor a high-side current for a power supply with a 358 opamp , I was wondering about the common mode
input voltage to the opamp. The 358's datasheet shows that the input must be about 1.5V below the supply voltage for
the opamp to function correctly. If the opamp supply is taken from the same supply rail that I'm trying to monitor , I wont get a valid
reading because the input voltage is the same as the supply , for 1 pin anyway. The other input pin(-) is on the other side of the current sensing
resistor and may be below the supply depending on current.
IF I connect the opamp as a single supply differential amplifier , will this allow me to measure the high side current correctly.
The reason I ask is that in this configuration , if all 4 resistors are equal then the opamp pins "see" half the supply voltage. (ie , below the supply)
Will this configuration work correctly? I can add amplification after the diff amp.
Before anyone asks I cant simulate this as the LM358 model does not accurately model this , I've tried. 3 models from 3 manufacturers
all behave incorrectly. :0(
Cheers
Neddie

 

[barry]

I think the diff-amp approach should work. If you can't simulate it, why not breadboard it? It would be a pretty simple test.

If the diff-amp approach doesn't work (and I can't think why it wouldn't) you could try putting a voltage divider on either side of your sense resistor and use a diff-amp to measure the difference across the two dividers; this would guarantee that you're working inside the CM range.

 

[neddie]

Thanks for the reply. My colleague built the cct and said it did not work. I've just looked at his schematic and he has made the diff amp into a gain stage of
about 47. 1K and 47K resistors. That means the opamp pins won't be 1.5V below Vcc and the cct won't work. I'll redo with better values to ensure the biasing
is correct.
Cheers
Neddie

- - - Updated - - -

Thanks for the reply. My colleague built the cct and said it did not work. I've just looked at his schematic and he has made the diff amp into a gain stage of
about 47. 1K and 47K resistors. That means the opamp pins won't be 1.5V below Vcc and the cct won't work. I'll redo with better values to ensure the biasing
is correct.
Cheers
Neddie

 

[FvM]

LM358 behaviour isn't completely represented in usual Spice models, but common mode range is (at least roughly, as far as I remember). Which problems do you particularly see?

G=1 differential amplifier will work, you'll usually want a higher gain for a current sense amplifier, however. Why don't you refer to a R2R OP or dedicated high side current sense amplifier. In some cases, a dual PNP transistor (e.g. BCV62) in degenerated current mirror configuration might give sufficicient accuracy.

 

[samson A.]

I think you will have to move your sensing point to the ground path, if you must use 358.
Else you can find opamp whose input can go all the way to their supply rail.

 

[neddie]

Thanks for all the help.
I've saved myself a lot of hassles and got me a MAX4080 to do the job :0)