• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Micro-Current Measurement Circuit Problems

Visitor

Active Member
I developed this circuit based on TI's SBOA336 app note: Low(Microamp), High-Side, Current-Sensing Circuit with Current-Sensing Amplifier... Instead of theINA190A1 I used a INA225 as it was available at JCLPCB, and an ADS1114 ADC for evaluating changes to the temperature monitor I've been working on. This works, with a few unexpected features I can't explain;

1. Large offset voltage for the input of INA225 (it follows the gain setting of the INA225. I have an offset of about 37mV, which translate into 0.97V at g = 25, 1.87V at g = 50 and 3.75 at g = 100. Initially, I thought this might be due to the input offset of the voltage follower for the analog output (poor op-amp choice), so I cut the trace to this stage without any change.

2. I've done testing with a 1k shunt, and loads of 4.7M, 2.7M and 1M with a 5 volt supply. If I subtract the open-circuit output from the load current measurement, the ratios for current for the different resistors track, and the ratio for the same resistors at the different gain settings track. For example, I get a reading of ~16.6mV/ µA, where I expect close to 25mV/ µA with a 1k shunt resistor.

I did look at the output with a cheapie 'scope. No evidence of anything oscillating.

Any suggestions as to what I'm missing?

MC Schematic.jpg
 

KeepItSimpleStupid

Well-Known Member
Most Helpful Member
You should take look at Dave Jones device, the uCurrent https://eevblog.com/files/uCurrentArticle.pdf

I made a 4-terminal I-V converter for in-house use using mostly the traditional approach but with a number of twists. 4 _-10V bipolar output ranges of 100, 10, 1 0.1 mA FS with 1s LED activation when the bipolar output exceeds 10V. Modes are 2T/4T, Suppress, Voc/Measure, Zero check/Zero correct/ The last function didn't get completed. AC performance was the design goal which it met. Offset was 40 pA DC uncorrected in any way.

With the traditional approach, Ib and Vos are the most important parameters. Vos varies with temperature.
 

Nigel Goodwin

Super Moderator
Most Helpful Member
You might also look at the 'Current Ranger' (which I've recently bought for work), and details (including the software) are freely available to make your own.
 

Visitor

Active Member
Here's a simplified schematic to make the situation more clear.

The shunt resistor is 1k, the load resistor is 4.7M. With a supply voltage 5 volts, the current through the shunt and load resistor is 1.1μA.

The INA225 has gains of 25v/v, 50v/v, 100v/v and 200v/v. With the load resistor open, I get 35mV offset at the input, resulting in an out of:

0.94v @ g = 25

1.87v @ g = 50

3.7v @ g = 100

If I subtract the offset from the readings, I get realistic readings that track with gain and with different values of load resistors, but the magnitude is off. With g=25, I get a result of 16.6mV/μA.

For 1μA across the shunt:

V = I R * G

V = 1uA * 1000 * 25 * 1000mV/V = 25mV/μA.

So my questions are:

o What's the cause of the offset?

o Why are my results so far off of the calculated results?


Simplied Schematic.jpg
 

rjenkinsgb

Well-Known Member
Most Helpful Member
The only thing I can see is that the data sheet has VREF at half supply.

Try putting a bias, eg. 0.6V using a diode and resistor on that, and also connecting the ADC AIN1 to the same point to get rid of the bias offset.
 

Visitor

Active Member
Thanks for taking a look. Below is a circuit diagram from the INA225 data sheet, which shows Vref connected to ground.

For uni-directional current measurement:

8.2.2.1 Design Requirements The device can be configured to monitor current flowing in one direction or in both directions, depending on how the REF terminal is configured. For measuring current in one direction, only the REF terminal is typically connected to ground as shown in Figure 41. With the REF terminal connected to ground, the output is low with no differential input signal applied. When the input signal increases, the output voltage at the OUT terminal increases above ground based on the device gain setting.
I think REF is ok for my application.
INA225 Circuit.jpg
 

ronsimpson

Well-Known Member
Most Helpful Member
What's the cause of the offset?
Please check my idea. You are trying to measure 1uA in the load while far more than that is going into the inputs of the INA225.
I think the input current into the IC exceeds the current to be measured. The IC was not designed to do this job.
Look at "Rint" and "Input bias current"

1601343208569.png
I could be wrong and often am but I think the INA190A1 is a very different part.
 
Last edited:

Visitor

Active Member
Good catch. Analog design is not my forte, and I was trying to hammer this into something JLC could assemble.

This is probably the explanation for lower-than-expected mV/μA output.

It will be usable for what I'm trying to do, with some odd scaling.
 

Latest threads

EE World Online Articles

Loading
Top