4-20mA receiver with current shunt

Hi.

I'm trying to built a receiver for 4-20mA signal. It is not the first time I try something but this time a use some existent stuff.

oups sorry.. press wrong button.. lol

so, I use the shematic of 4-20mA to SPI receicer from Mikroelectronika:
http://www.mikroe.com/downloads/get/1963/4-20ma-r-click-manual-v100.pdf

there's is so little strange thing about it... First if you take a look to the INA196 chip, Vin+ is connect to VCC-16.

Normally, 4-20mA use an external power supply for the loop. Is that VCC acting for it?

If that so, seems to me that only the 4.99 ohm of the shunt will provide only a small voltage into the loop render it most sensible to EMI. To solve that issue, can I add in series with the input a 250ohm (making input 1-5V in the same time, and with editing the code 0-5 capable as well...).

so here the shematic that I got so far.

4_20m_AR.png

Thank

Isn't is easier to just use a 51ohm resistor, and measure the voltage across with 1.1volt Aref enabled.
Unless you need the 12-bit resolution of that board.
Leo..

Yes I want to use the 12bit reso and all my analog input are already use by ma setup...

Also a little mistake was inside my picture. there a new one:

4_20m_AR_recheck.png

That ClickBoard schematic does look like it has a voltage booster to power the sensor. Depending on what sensor you choose, it may require the 4-20mA loop to be powered. If you have 12V on your board, you can probably use that.

The sense resistor (4.99R) should be chosen according to the INA196 datasheet. It should be located close to the chip and then there will be no EMI problems.

Hi MorganS.

That ClickBoard schematic does look like it has a voltage booster to power the sensor.

the second pic didn't show the part that I think provide ''vcc 16 v". this is it:

But, I don't think it can power a sensor, just generate the current for the loop. I think 2 wire 4-20 sensor cant be use with this type of input, and that's a problem. (I'm about to find something... keep you posted)

he sense resistor (4.99R) should be chosen according to the INA196 datasheet

I have been. 20mA @ 4.99 = 0.0998v, INA196 have 20v/v gain = 1.996V and the ADV is 2V range.

It should be located close to the chip and then there will be no EMI problems.

Is is as close as it can get... But is it prevent interference even for what can from a very long cable, even with shield twisted pair( many meters) and passing near by AC motor... That's my concern...

Thank.

I remove all the part of the vcc booster and reverse the polarity of the scew header. Now I think this will be compatible with any 4-20 sensor with and external source.

ho… need betther def… :stuck_out_tongue:

Seems an INA219 can do the job of both chips.
Current sensor and 12-bit A/D in one.
This board, with R100 changed (to 15ohm), could do what you want.

I2C though. Not sure what speed you need.
Leo..

Yes Wawa. I note the idea and investigate.

Not this is what I have done next, I add couple resistor in serial with the shunt and a jumper to select the type of input. Now, in position 1, it read 0-20mA like before.

in position 2, it put in serial a 246ohm for total of 250.99ohm. that will be the good scale for shunt for 0-5V signal.

in position 3, it add another 250ohm to scale the shunt for 0-10V signal.

also, like 4-20mA before, you can do the 1-15 and 1-10V signal trough the sketch.

You can't just make a voltage input from a current input.
A voltage input is high impedance, not 250ohm.

Still don't understand why you do all of this.
A 51ohm resistor (post#2) and a 10k protection resistor between that and the Arduino pin, and 1.1volt Aref, could give you 800 A/D values. Maybe 1000 with a bit of oversampling. 0.1-0.2% accuracy.
Maybe better than a current sensor chip and a 12-bit A/D.
Leo..

Well, maybe the rest of the circuit needs to measure 5V values so he can't just change Aref?

It's a good idea to have a dedicated 4-20mA chip standing between the Arduino and the outside world. Any wire leaving the Arduino box could potentially be dragged across a high voltage supply or hooked up backwards or whatever. Those chips intended for industrial applications may be able to protect against thousands of volts.

Personally, I like the MAX14626 current loop protector which protects the Arduino analog input from many common wiring mistakes. The INA196 is able to operate with -16V to +80V on the pins, which is more than just protection.

@nitrof - the EMI interference is why you use a current loop in the first place. With twisted-pair cable, you can run it through anything short of an atomic bomb without concern.

MorganS:
Well, maybe the rest of the circuit needs to measure 5V values so he can’t just change Aref?

Those chips intended for industrial applications may be able to protect against thousands of volts.

  1. If he is serious about measuring, he must change to a different/stable Aref.
    Default 5volt Aref is potentially unstable, and therefore not good for measuring the voltage across a shunt resistor.

  2. The datasheet says +/- 40volt.
    Easilly done with for the Arduino pin with a 10k resistor and, if needed, two schottky clamping diodes.
    As far as I can see, all that chip does is protect the shunt resistor by disconnecting it with mosfets.
    Leo…

Thank allz.

@MorganS I believe you understand well the approach I want to do. with remote sensor, long distance and chance of mistake.

@Wawa, I've done my homework yesterday... and you're right... I have to have hi impedance on voltage input... I propose something shortly. And another reason I didn't take 1.1 Aref and 51ohm is I already use all my analog input...

Same thing for the I2C chip... need an analog. I've looked a little for SPI version but there is no equivalent in INA. maybe with another manufacturer... keep searching.

Again thank a lot for you're help. always kind of you !.

Keep you posted with developpement.

Regards.

Nitrof

Ok. now I calculate value of all resistor to do 0-5 and 0-10. Found on some sensor manual that expected impedance for current input is 250ohm and voltage input, 1Mohm.

that's what I get, but I have concern about using a switch with the shunt resistor... not sure if it will cause problem with resistance of the nets and the switch...

??
For current, lower resistance is better. It has the lowest volt-drop, so more left for the sensor.
A 250ohm resistor will "rob" 5volt from the supply.

I don't see how you get a 1Megohm impedance with a 280k series resistor and a 20k shunt resistor.

Why change voltage to current, and then back to voltage again.
Better to connect voltage directly to the A/D.
Leo..

oups typing mistake.. it is 988 and 976k... sorry.

For 250 ohm.. I thing it is common use in industrial because they use 24vdc supply...

for voltage... 2 reason:

-protection
-I une 3.3volt board so 5 volt is an issue.

nitrof:
For 250 ohm.. I thing it is common use in industrial because they use 24vdc supply...

for voltage... 2 reason:

-protection
-I une 3.3volt board so 5 volt is an issue.

No, a 250ohm resistor is used to generate 5volt (with 20mA sensor current) for a 5volt A/D.

If you want 3.3volt across the sense resistor with 20mA sensor current, you would pick a 3.3 / 0.02 = 165ohm resistor.

Only the correct value shunt resistor uses the full range of the A/D.

Protection is a different story.
It has to co-exist with the shunt resistor without influencing anything.
The chip mentioned by MorganS does that by disconnecting the shunt resistor.
But there are many other ways.
Leo..

Ok.

Have you some reference or idea on what to do… because the the max14626 do only 4-20 and I want full scale 0-20 and 0-5… If I want some acceptable more protection with simple desing, explosion proof is not necessary . :wink:

And i remove the 250 ohm in my circuit. But I notice that max14626 have 250 Rload. Just saying… :stuck_out_tongue:

Never used that chip, but it just seems to be a switch that disconnects an external load/sense resistor in a fault situation. I didn't see a minimum current. There is a 0-20mA graph in the datasheet.
The "switch" has a ~25ohm resistance, but that does not matter in a current loop.
You can use any value load/measuring resistor in a current loop.
Not sure if you should use the chip for voltage input.
It has a reverse voltage, but not an over voltage protection "switch".
Leo..