50 mA to 1000 mA Current Measurement Circuit but with 1 voltage output

Thread Starter

Kulanib

Joined Jul 6, 2022
43
Hi

I am trying to measure current rated 50-mA to 1000-mA. I saw that the easiest method is to use a Shunt resistor.

My challenge is that my measuring device can only accept 1 analog input terminal reference to ground (0V). So I have to take my shunt voltage drop and convert that to a single signal (similar to what is shown below) and have that single signal connected to my Measuring Device.

How can I achieve this? How can I reduce the 2 probes to 1 probe?


Capture.JPG
 

Ian0

Joined Aug 7, 2020
5,504
You already have two probes - the pointy one and the earth clip.

Is this just something you are measuring, or part of a larger piece of equipment that you are designing/building?
 

BobTPH

Joined Jun 5, 2013
5,227
The probe on the right side of your diagram is exactly what you requested, a voltage with respect to ground.
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
You already have two probes - the pointy one and the earth clip.

Is this just something you are measuring, or part of a larger piece of equipment that you are designing/building?
I am designing a DAQ system that reads multiple sensors and logs them. My measuring device is the USB 6002 from National Instruments but it doesn't measure current, only voltage (with 16-bit resolution). It has 8 analog inputs so I am limited to using 1 input terminal per reading. It does have a reference ground that is common per 2 inputs.

The system I am measuring has multiple loads on the same wire loop hence I need a 2 probe deferential reading across the shunt resistor, that reduces to 1 signal into the input of the DAQ device - NOT reference to Earth Ground though to the point right after the shunt (which has a none zero potential).
 
Last edited:

Thread Starter

Kulanib

Joined Jul 6, 2022
43
The probe on the right side of your diagram is exactly what you requested, a voltage with respect to ground.
Yes exactly, but differential probes are to expensive.

I tried using this Arduino ACS712 which outputs a voltage but it is not sensitive enough for low current. But it has the 1 probe going into A0.

24eceab8db1601bf90be82c3736202b42f7a1156.jpeg
 
Last edited:

Ian0

Joined Aug 7, 2020
5,504
I am designing a DAQ system that reads multiple sensors and logs them. My measuring device is the USB 6002 from National Instruments but it doesn't measure current, only voltage (with 16-bit resolution). It has 8 analog inputs so I am limited to using 1 input terminal per reading. It does have a reference ground that is common per 2 inputs.

The system I am measuring has multiple loads on the same wire loop hence I need a 2 probe deferential reading across the shunt resistor, that reduces to 1 signal into the input of the DAQ device - NOT reference to Earth Ground though to the point right after the shunt (which has a none zero potential).
You need a current shunt amplifier, such as INA180, or, my current (no pun intended) favourite, TSC2010
They have a gain of 20 (other values are available) which would allow you to use a smaller (and cooler) value of shunt resistor.
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
You need a current shunt amplifier, such as INA180, or, my current (no pun intended) favourite, TSC2010
They have a gain of 20 (other values are available) which would allow you to use a smaller (and cooler) value of shunt resistor.
That's more like it. The TSC2010 looks very lovely indeed. If I can just figure out how all the pins work and how I can soldering everything into a broad.

I forgot to mention that my DAQ device can has a Max input voltage of 10 V.
 

Ian0

Joined Aug 7, 2020
5,504
Easy: IN+ and IN- go to either side of the shunt. Make sure that use use a Kelvin contact so that you are measuring the voltage across the just the shunt resistor and not the voltage across random bits of wire of pcb track.
Ignore SHDWN. (The AD8418 is the same thing but without SHDWN. That used to be my favourite - I thought that it must be good because Analog Devices made it, then when it was out of stock everywhere, I found the TSC2010 and its spec is actually better.
How you connect Vref1 and Vref2 depends on whether you want to measure current in just one direction or in both directions.
(And "output" is fairly obvious)
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
Easy: IN+ and IN- go to either side of the shunt. Make sure that use use a Kelvin contact so that you are measuring the voltage across the just the shunt resistor and not the voltage across random bits of wire of pcb track.
Ignore SHDWN. (The AD8418 is the same thing but without SHDWN. That used to be my favourite - I thought that it must be good because Analog Devices made it, then when it was out of stock everywhere, I found the TSC2010 and its spec is actually better.
How you connect Vref1 and Vref2 depends on whether you want to measure current in just one direction or in both directions.
(And "output" is fairly obvious)
So I couldn't find the exact AD8418, but I am convinced that the AD8204 will do the job. Funny enough, I have a solenoid in line with the loop I am trying to read, just like in the diagram. I will place my shunt on the high side as well for easier access.

Capture2.JPG

I will construct a circuit to mount the AD8204 with its components. The output should be the differential reading with its gain.

I have read about the Kelvin Method but I'm not sure what they are on about. The 4-term shunt depicted in figure 2, is just me with my through-hole resistor (shunt) connected inline with the main current from the battery, then probing a wire to +IN and then another wire to -IN of my D8203 (meaning Shunt connected the same as the original differential probes I posted at the top)?
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
4-terminal shunt IS the Kelvin method!
You can also get devices that work with a shunt in the positive supply
https://www.diodes.com/assets/Datasheets/ZXCT1021.pdf
It seems like a neater solution because all the signals are not zooming up and down with your switching waveform.
Thanks for clearing the Kelvin Method. I still like the AD8203 because of it's while voltage input range, and I don't mind the 20 V gain because I will use a 0.1 Ohm shunt or so and my DAQ device has 16-bit resolution. The zooming up and down is exactly what I want to measure because I am testing the current through a solenoid and voltage as well - reading at 1000 Hz with 1 sample per read, I just want to see how the solenoid responds from a supply voltage (PMW) and the actuations as well.
 

Ian0

Joined Aug 7, 2020
5,504
Thanks for clearing the Kelvin Method. I still like the AD8203 because of it's while voltage input range, and I don't mind the 20 V gain because I will use a 0.1 Ohm shunt or so and my DAQ device has 16-bit resolution. The zooming up and down is exactly what I want to measure because I am testing the current through a solenoid and voltage as well - reading at 1000 Hz with 1 sample per read, I just want to see how the solenoid responds from a supply voltage (PMW) and the actuations as well.
With the high-side monitor you are measuring the same current, but without the huge common-mode voltage changes.
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
With the high-side monitor you are measuring the same current, but without the huge common-mode voltage changes.
Oh, So to clarify: the https://www.diodes.com/assets/Datasheets/ZXCT1021.pdf will work better for addressing the Common Mode Voltage (What is common mode voltage)? And oh, just remembered that my line voltage to solenoid is 5 V so it will work (20V max input range).

And sorry for spelling mistakes- PWM, AD8203.
 

Ian0

Joined Aug 7, 2020
5,504
Suppose your solenoid is being driven with a 20kHz PWM waveform at 14V, and your shunt is between the MOSFET and the solenoid.
You are trying to measure a voltage across the shunt of a few millivolts, BUT the voltage on each side of the shunt is a 14V squarewave. The 14V is the common mode voltage because it is COMMON to both sides of the shunt. You are trying to extract the DIFFERENTIAL voltage (the difference between the voltage at either side of the shunt).
Imaging trying to do that manually on a scope - not easy, is it? It's also not so easy for the AD8203, but it makes a pretty good job of it. Now see what happens if you put the shunt in the positive supply - same voltage across the shunt, but both terminals are now at 14V. Much easier for the electronics to measure the millivolts across the shunt, because there isn't a high freqeuncy large voltage in the way.
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
Thank you for the explanations @Ian0.

I see that most of these chips, e.g. AD8203, the TSC2010, max4080 have a very tiny package size. e.g. SOT-23.

Are there any high side current sense IC that are DIP package? Or maybe How would I solder such small package sizes?
 

BobTPH

Joined Jun 5, 2013
5,227
You need a current shunt amplifier, such as INA180, or, my current (no pun intended) favourite, TSC2010
They have a gain of 20 (other values are available) which would allow you to use a smaller (and cooler) value of shunt resistor.
Why does he need that when one side of the resistor is grounded?
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
Why does he need that when one side of the resistor is grounded?
I had a low side current measure solution, but it turns out that I can't splice the ground cables because they are submerged inside the valve pan (valve body) of a Transmission Gear box.
Valve Pan.JPGValve body.JPG
Solenoid.JPG

I now only have access to the Connector on the harness, which has individual solenoid Vcc lines, but a common ground (it seems).

Harness.jpg

So I will be splicing the "female" connector that plugs into this 19 Pin plug. Meaning I will only have access to the high side.
 

Thread Starter

Kulanib

Joined Jul 6, 2022
43
Hi @Ian0,

  1. I Managed to model the PCB and have all 8 current sensors in 1 PCB for a compact solution.
  2. Went with the Max4080T (20 gain) SOIC8. It only draws 75µA.
  3. I used the kelvin Contact method on the Shunt Resistors of 0.1 Ohms and 4W. I also gave them thicker 1mm copper in order to allow 1 A max Current flow.
  4. I Should have gone for 5.08 PCB Terminal Connector Blocks but took the 2.54 mm because I was stupid to not measure the pin pitch - which is really small but just enough to fill the 1mm cables.
  5. I also inserted some straight-relief holes for use with a zip tie.
  6. I am also proud of my copper paths as well, I am still learning some tricks here and there.

just wanted to update everyone.

Thanks.

PCB Layout Current Sensor.JPG
REV001 Schematic_Current Sensor PCB_x8_2022-07-23 PCB.png
 

Attachments

Top