I have a current loop (pls see attached). The loop runs a distance of about 500 meters. There's 20mA on the wire supplied by a current source at the sender off of an LM317 regulator. At the receiver I want to measure the current value on the loop wire. I put a 50ohm Rsense resistor at the receiver so I can read 1V when there is 20mA on loop. However this Rsense is floating (not ground referenced) because I need ground reference further down in wire (attached picture might make this clearer). So my question is: sinse Rsense is floating, how do I measure the voltage difference across it?

(I think an opamp comparator across the Rsense might do but I would have to single supply the opamp with one of the Vin's equal to Vs+. I think this makes getting Vout to = 1V (or less when there is < 20mA on the current loop) hard.)

Is there a good method to measure voltage across Rsense? Or can I use a different approach, everything can be substituted.

 

Why is there a problem measuring across a floating resistor?
That is what op amps do.

 

I did the subtractor opamp circuit (pls see attached) across Rsense and could not get the value I was expecting (It should be something like (14V - 13V = 1V at the Vout but instead I got some Vout in the 200 millivolts..) I'm using ad820 as my opamp which supports single supply. I'm thinking is it too hard for Vout to swing to value too close to Vs- (i.e ground) being that it's single supply and Vin+ is basically saturated at Vs+.. Note my whole opsamp is drawing less than 1mA, I put an ammeter between Vs- lead and ground.

 

Google high side monitor.

 

What is your signal output load?

1) The way it's configured, your output signal current and opamp power supply current become measurement errors. Can you power the circuit from another local source?

2) This configuration is not the best in terms of CMMR, (common mode rejection ratio) both inputs have a large common mode voltage. (the power supply voltage) Google "instrumentation amplifiers" - you can find other configurations that can handle the common mode voltage with less error.

 

Thanks for pointer to high side monitor. I'm still reading it..

My Rload is just a 12V battery (pls see attached). So this current loop functions as a charger to a remote battery. However voltages on either side of Rsense are higher than 12V (like between 12 and 14V). Could you elaborate about CMRR's effect here? Why can't the opamp just reject common mode voltages shared between 2 inputs, so it wouldn't matter if they are high or otherwise..

I'm beginning to think I should put Rsense on the low side to save total current consumption on wire (20mA in total) from which some will go into powering the extra opamp (or specialized high side monitor ic) needed when Rsense is on high side..

 

Could you not put the sense resistor on the ground side, and then use a buffer op amp like this?

 

Could you not put the sense resistor on the ground side, and then use a buffer op amp like this?

Sorry I'm unable to open this file.. Do I need to install a reader? Or could you send another format? Thanks!!

 

Why can't the opamp just reject common mode voltages shared between 2 inputs, so it wouldn't matter if they are high or otherwise..

Devil is in the details- this link has an in-depth analysis of CMMR as it relates to high side sensing.

http://www.eetimes.com/design/analog-design/4010355/Understand-low-side-vs-high-side-current-sensing

 

what did you use to draw the image, i used Ltspice?

Put the resistor on the ground feed and use a buffer op amp like this

 

thanks! I use tinycad. Yes, I rewired it with sense resistor at the ground (low side) and it worked - although I can't avoid the offset to ground everywhere else.. Would be a nice exercise to make it work with resistor on the high side though..

 

Try this see if it works...

 

Try this see if it works...

I did and it didn't work. A responder above stated CMRR and power supply and vin tied together being a problem which makes sense but I'm not grasping the wisdom behind it.

I should power the op amp with local power that is the battery but it's 12V while Vin's across Rsense are higher.

 

Check out this over-the-top current sense amplifier.

EDIT: What are you going to do with this measurement, i.e., why does it need to be ground-referenced?

 

Did you mean why does my load have to be ground referenced? I guess it didn't. My load is just a battery that would be charged by the 20mA current loop. I just assumed the battery(-) terminal ought to be tied to current loop return (ground) so I needed to put sense resistor on the hot or high side. But I experimented it with sense resistor on the low side and battery still charged although sitting on a virtual/offset ground so I'm happy with it.

So I guess current sensing on the high side requires specialized op amp but not a general one configured to do subtraction..

 

Did you mean why does my load have to be ground referenced? I guess it didn't. My load is just a battery that would be charged by the 20mA current loop. I just assumed the battery(-) terminal ought to be tied to current loop return (ground) so I needed to put sense resistor on the hot or high side. But I experimented it with sense resistor on the low side and battery still charged although sitting on a virtual/offset ground so I'm happy with it.

So I guess current sensing on the high side requires specialized op amp but not a general one configured to do subtraction..

That is not what i asked you. Where is the output of your current sense amp going to go? To an ADC? If so, that is a good reason to have the measurement referenced to ground. If not, then how will the measurement be used?

 

That is not what i asked you. Where is the output of your current sense amp going to go? To an ADC? If so, that is a good reason to have the measurement referenced to ground. If not, then how will the measurement be used?

Oh, the measurement will be fed to a comparator op amp comparing against a fixed reference, and the output of which will go high when the current loop has 20mA on it or low when there is 10mA on it. (I didn't detail it in the drawing but the sender will be able to switch between to current levels 20mA or 10mA off of a switch.) This information (Vout from comparator) will set off a relay powered by the 12V battery at the receiver.

 

Oh, the measurement will be fed to a comparator op amp comparing against a fixed reference, and the output of which will go high when the current loop has 20mA on it or low when there is 10mA on it. (I didn't detail it in the drawing but the sender will be able to switch between to current levels 20mA or 10mA off of a switch.) This information (Vout from comparator) will set off a relay powered by the 12V battery at the receiver.

With the sense resistor on the low side, you don't need an amplifier. You can go straight to the comparator.
If you want to put it on the high side, see the over-the-top currrent sense amp that I referenced previously.

 

I may be sounding like an idiot,as usual, but why you need a ground reference? place a multimeter and note the reading. multimeter has a very high impedance so only a negligible amount of current will flow through them. Or use a buffer.

 

I may be sounding like an idiot,as usual, but why you need a ground reference? place a multimeter and note the reading. multimeter has a very high impedance so only a negligible amount of current will flow through them. Or use a buffer.

That's what I was leading up to, but our OP told us why the measurement needs to be referenced to ground:

Oh, the measurement will be fed to a comparator op amp comparing against a fixed reference, and the output of which will go high when the current loop has 20mA on it or low when there is 10mA on it. (I didn't detail it in the drawing but the sender will be able to switch between to current levels 20mA or 10mA off of a switch.) This information (Vout from comparator) will set off a relay powered by the 12V battery at the receiver.