USB-1208LS noisy signal

[cchiarello]

Hello.

 

We are using a USB-1208LS acquisition board to measure the signal of a pressure transducer. The signal, however, is far from being smooth (accuracy is +- 5 bar).

 

We are measuring it using differential setup and a 24V power supply to the pressure transducer. We have tried different ports, supplying 12V to the transducer, different transmitters, and different sensors. All of the setups exhibit the same noisy behavior.

 

When grounding common (V-), the voltage between signal (V+) and GND becomes a constant 1.415V, as does the voltage between V+ and V-. If ungrounded, the noisy signal is recovered, but the voltage between V+ and GND stays at 1.415V.

 

We have used the same board model before and the results were always outstanding, as it is with different models (USB-TEMP-AI). When replicating the setup, however, we are not able to have the same high standard measurements.

 

Would you help us smoothen the signal and troubleshoot the GND port voltage?

 

Thanks,

 

Conrado

[Fausto]

Hello @cchiarello.

What is the make and model of your pressure transducer?  Do you have a datasheet for the transducer?

How are you wiring the transducer to the USB-1208LS?

Are you using a standard desktop PC or laptop?  If laptop, is the power supply connected to the laptop?

https://www.mccdaq.com/pdfs/manuals/USB-1208ls.pdf

https://www.mccdaq.com/pdfs/manuals/USB-TEMP.pdf

The USB-TEMP has a 24-bit A/D, while the USB-1208LS has a 12-bit A/D, when the analog inputs are configured for differential mode.  Therefore, there is less resolution with the USB-1208LS module.

Regards,

Fausto

[cchiarello]

Hi Fausto,

 

Thanks for the quick response. We are using a PX-613 by Omega (https://assets.omega.com/manuals/test-and-measurement-equipment/pressure/pressure-transducers/M947.pdf)

 

We are using a standard desktop PC.

 

We are further investigating the issue. We have replaced the power supply with a 24V grounded, more stable one. It reduced the noise considerably, but we are still a bit short from what we would expect (+- 0.5 bar). Grounding V- does not seem to make a difference.

 

We are connecting V+ (red) from the sensor to V+ in the power supply; S+ (green) from the sensor to V+ in the board; and V- from sensor is connected both to V- in the power supply and the board.

 

Is that the limiting resolution for this model in our application case?

 

Thanks,

 

Conrado

[Fausto]

@cchiarello,

What is the complete model nomenclature of your Omega pressure transducer?  Please refer to the second page of the following PDF link, which will help determine the range of the transducer.

https://assets.omega.com/pdf/test-and-measurement-equipment/pressure/pressure-transducers/PX603.pdf

If the transducer's range is 0-1379 bar (0-20000 psig), then a 1 mV output equates to 0.345 bar.  (Calculations:  1379 bar / 4 V)

If the USB-1208LS analog input range is set to +/-5 V and differential mode, then the A/D resolution is 12-bit, with an accuracy of 2.44 mV/LSB.  (Calculations:  10 V / 2^12)

Therefore, for this scenario, each LSB on the USB-1208LS's A/D equates to 0.842 bar.  (Calculations:  0.345 bar/mV * 2.44 mV/LSB)

However, as a second scenario, if the transducer's range is 0-10.3 bar (0-150 psig), then each LSB on the A/D equates to 0.092 bar.

Which transducer model do you have?

Lastly, your wiring appears to be correct according to initial datasheet.

 

Regards,

Fausto

[cchiarello]

Hi Fausto,

 

Once again, thanks for the response. We are using a PX613-5KG5V (0-5000 psi).

 

If I'm following the calculations correctly, that would imply 1 mV = 0.086 bar, which, in turn, would give 0.21 bar/LSB.

 

This is in the same order of magnitude of our noise, but still lower than what we measure. Do you have any suggestions on how to minimize the fluctuations (aside from buying a new board?)

 

Thanks,

 

Conrado

[Fausto]

@cchiarello,

Those results are for an ideal, error-free USB-1208LS A/D transfer function.  Since the A/D transfer function is part of a system, then there is offset error and gain error to consider.  Table 4 in the USB-1208LS User's Guide (https://www.mccdaq.com/pdfs/manuals/USB-1208ls.pdf) shows the accuracy at Full-Scale, for a +/-5 V range to be 19.766 mV.  Therefore, using the conversion of 1 mv = 0.086 bar, a signal read by the USB-1208LS, with a +/-19.766 mV accuracy specification, gives you the accuracy of +/-1.70 bar.  

 

On 7/28/2023 at 12:23 PM, cchiarello said:

We have used the same board model before and the results were always outstanding, as it is with different models (USB-TEMP-AI). When replicating the setup, however, we are not able to have the same high standard measurements.

Have you used a different USB-1208LS device in the past, with this same transducer model, and achieved desirable results or were you referring to the USB-TEMP-AI results?

Regards,

Fausto

[cchiarello]

Fausto,

 

Thanks again for your support.

 

I have spoken to my supervisor and the magnitude of the error we are observing with the current setup (+- 2 bar) seems to be acceptable. The transducer giving +- 0.5 bar was not connected to the system, and I suspect that is why it had such a good accuracy.

 

We were able to achieve better results using a USB-1208L, but again, the current setup seems to be working just fine. I will let you know if any further support is needed!

 

Thank you,

 

Conrado