USB1608GX - Differential Voltage Measurement Noise

[GSev]

Hello,

 

I ran some tests measuring two current output sensors (a 2-wire and 3-wire) using sense resistors using differential analog output and DaqAmi. Both connected on the same power supply, Keithley 2200.

The blue and orange lines represent two pressure sensors with an input of shop air supply. As the drops are so fast I do not believe that it is the input and must be the measurement. Does anyone know the cause of this and how to avoid?

Thanks!

 

 

[Fausto]

Hello @GSev.

Please provide datasheets for the pressure sensors, sense resistors, and Keithley 2200 power supply.

You mentioned "using differential analog output".  Is that correct or should it read "using differential analog input"?

The first half of your graph looks good.  Did you have other equipment, i.e. pumps or motors, running for the second half of the graph?

Do you have a diagram of your entire wiring from pressure sensors to power supplies, PC, and ground?

Are you using a laptop with an external power supply connected to it?

Regards,

Fausto

[GSev]

Hi Fausto,

Yes you are correct it should be using differential analog input.

I'm using our internal sensors which I do not want to post information on a public forum. Resolution of orange line is 12 bit with a full scale of 200 psi. It has some input protections. Blue line is Prosense EPS25-1450 psiprosenseeps1001dps.pdf. Sense resistors are just some 0.5% through hole resistors that were selected based on the output. What are you looking for specifically?

The blue line sensor has a PNP switch that commands a solenoid which either vents or pressurizes the manifold. The grey line indicates the change of state of the solenoid. The solenoid circuit has a flyback diode and a voltage divider to allow the DAQ to measure (grey vertical line). Sensors are parallel and connected to negative of the power supply. Solenoid negative is on the same line. Solenoid ground is connected to power bar.

Laptop uses an external power supply that is connected to the same power bar.

Thanks,

-Greg

Series2200_0.pdf

[Fausto]

@GSev,

Thank you for the response and attachments.

The EPS25-1450-1001 pressure sensor requires an operating voltage of 18 - 30 VDC.  Which Keithley 2200 series model are you using?

Are you using a 250 Ohm resistor across the analog input channels, on the USB-1608GX device, or a different resistor size?

With the analog input channels configured in differential mode, did you connect a wire from the CH#L terminal to the adjacent AGND terminal?  If you have not, please do so and test again.  Likewise, if you have a wire connected between the CH#L terminal to the adjacent AGND terminal, then remove it and retest.

Next, please run your test with only one sensor connected to the USB-1608GX and the other terminal shorted.  Keep both analog input channels enabled in your software application.  Afterwards, swap out the pressure sensor, while keeping the shorted second channel and repeat the test.  Run another test with both analog input channels shorted and the equipment running.  Do you still see a similar amount of noise with both shorted channels, while the equipment is running in the background?

1 hour ago, Fausto said:

The first half of your graph looks good.  Did you have other equipment, i.e. pumps or motors, running for the second half of the graph?

Were you able to identify the start of a piece of equipment for the second half of the graph?

Lastly, connect both pressure sensors back to the analog input channels on the USB-1608GX.  Disable the solenoid(s) and rerun your test.  It's fine if there is no pressure.  Does the data contain a similar amount of noise?

Hopefully, we get some clues from these tests.

Regards,

Fausto

[GSev]

Hi Fausto,

 

Keithley 2200-60-2. Set to 24V with about 250 mA to handle the solenoid demand.  Attached is diagram. Orange channel has a 390 ohm resistor (measured at 388.2). Blue has 220 ohm resistor (measure at 221). I'm not particularly concerned with inaccuracy of the conversion just changes in shape and those drops.

As I have three channels connected. Should I connect 1's CH#L to 2's AGND and 2's CH#L to 3's AGND?

1 hour ago, Fausto said:

Were you able to identify the start of a piece of equipment for the second half of the graph?

I think there is no other equipment connected to that fuse. It runs in a conduit that does have plenty of other wires.

I'll run the tests.

[GSev]

Hi Fausto,

Yes, I forgot to do the "floating" voltage source connection. If only I had read the manual before coming here!

Thanks,

-Greg