Current Shunt Monitor

[2camjohn]

In order to monitor the current flowing through a motor in one of my circuits, I got some current shunt measuring ICs.

The chip i have chosen is the INA139

These chips are supposed to give a voltage output, relative to the current flowing through a shunt resistor.




However Im having trouble getting any kind of sensible output.



I copied the typical circuit as shown in the schematic and connected my oscilloscope between VCURRENT and GND.

Please see the attached schematic.
The load is turned on when CB1 is shorted high.

Also there is a 100uf capacitor in parallel with the battery which is not shown on the schematic.


After trying a wide selection of shunt resistors, changing the gain, using a resistor instead of the motor etc etc I cant get anything close to an indication of the current flowing through the load.

At best I get some transient noise when the load is switched on, it looks very much like the signals you get by touching the oscilloscope probes but with less amplitude.


So before I start again from scratch, I was wondering if anyone could spot anything obvious in my schematic?
Or if anyone can recommend an alternative high sided current monitor which is straight forward to get working(must be small like sot23)?
Or any other methods of monitoring current?
Or if anyone knows any good websites where I can read up on this sort of thing, i feel like ive already read every post in this forum related to current monitoring, seems like a fairly simple concept so i dont have a clue where the problem lies.

 

[ericgibbs]

hi john,

Have you tried connecting the motor suppresion diode at the Rsense resistor, so the diode is actually across the motor?
[not across the mtr and resistor]

As you may know you will get a back emf pulse from the motor at power off, which should be absorbed by the diode, but depending what value resistors you have been trying, this back emf could appear across the 1NA139 inputs.

Also as a test, link the motor ground side directly to 0V [bypass the FET], to ensure that the low side of the motor is actually at ground ref for the 1NA139

 

[Roff]

Do you have some idea of the peak current of the motor?

 

[2camjohn]

hi john,

Have you tried connecting the motor suppresion diode at the Rsense resistor, so the diode is actually across the motor?
[not across the mtr and resistor]

As you may know you will get a back emf pulse from the motor at power off, which should be absorbed by the diode, but depending what value resistors you have been trying, this back emf could appear across the 1NA139 inputs.

Also as a test, link the motor ground side directly to 0V [bypass the FET], to ensure that the low side of the motor is actually at ground ref for the 1NA139

Hi Eric,
Thanks for the reply.

Yes that is what I tried initially. I switched to this way as I read on a website it was a better way of doing it when your current sensor only reads in one direction.
But it makes no difference to the output.
The output isnt degraded or noisy, its just noise. It oscillates around 0V.


Thanks for the Idea of bypassing the FET, I didnt think of that.
Ill try that right away.

 

[2camjohn]

Do you have some idea of the peak current of the motor?

Hi Ron,


Its a tiny motor but often under high load.


From memory:

It can go as high as at 10A under fault conditions, but in my circuit i plan to limit it to 5A maximum. (thats the plan right now anyway)
Normal use would be around 0.5-1.5A



One aim of this circuit is to limit the current to the motor.

 

[ericgibbs]

Hi Ron,


Its a tiny motor but often under high load.

Perhaps I'm stating the obvious, [but I have to ask] when you do your Imtr tests, is the motor physically loaded or is it just running free?

As its a tiny motor the off load current is probably very small.

 

[2camjohn]

Eric,
Yeah I am comparing the unimpeded and impeded motors as part of my test.





Minor update.

I now seem to be getting an output but it disappears after a few ms of the motor being switched on.
Even with the motor impeded the signal falls just as quickly.

I can pulse it and get a pretty pattern on the scope See attached pic.



Is that supposed to happen? I was going to drive the motor by PWM anyway so I can measure it after the pulse starts I guess?

What I was expecting was as long as the motor is running it draws current which creates an output voltage from the current monitoring IC?

Based on the voltage levels I was getting from the scope, I am going to make another prototype with slightly different component values and a beefier mosfet (as this one is getting hot and I think limiting the current and screwing up my results).

I think im making progress but I want to be sure im understanding things right.



Thanks alot for the help so far.

 

[Hero999]

Connect a 100nF capacitor from either side of the motor's terminals to its frame.

Connect a 100nF capacitor across the current sense resistor.

You should also consider clamping a ferrite bead over both of the supply lines to the motor.