Pressure Transducer Signal Jumping

[kinghuds]

I have 2 Omega pressure transducers wired up through a PXI, both are reading somewhat accurately but while reading the data the signal fluctuates greatly, not sure if this is due to noise or something else. They are both measuring the pressure in a small duct but the readings are jumping around too much compared to a manometer taking the same readings.

[Ben]

Are you using proper signal conditioning?

Ben

[aeastet]

---

@kinghuds wrote:

I have 2 Omega pressure transducers wired up through a PXI, both are reading somewhat accurately but while reading the data the signal fluctuates greatly, not sure if this is due to noise or something else. They are both measuring the pressure in a small duct but the readings are jumping around too much compared to a manometer taking the same readings.

---

This could be a lot of thing....Where to start?

- The manometer is probably doing some sort of averaging. I would check this.

- You may not have the shield hooked up on the sensors.

- You could have a ground loop

- You could have sensors that are susceptible to noise.

- This could be real data that the other system is not picking up.

- Could be some artifact that you have because of programming

- You could be sampling too fast for the sensors capabilities.

If you know that the data is good. I would suggest sampling faster and averaging some data points to get cleaner data. You could add a filter. Take your pick of any of these ideas.

[kinghuds]

I don't have any conditioning in my program, just scaling the voltage in MAX and then reading/displaying/writing it in my vi.

---

- The manometer is probably doing some sort of averaging. I would check this.

- You may not have the shield hooked up on the sensors.

- You could have a ground loop

- You could have sensors that are susceptible to noise.

- This could be real data that the other system is not picking up.

- Could be some artifact that you have because of programming

- You could be sampling too fast for the sensors capabilities.

If you know that the data is good. I would suggest sampling faster and averaging some data points to get cleaner data. You could add a filter. Take your pick of any of these ideas.

---

I'm sure the manometer is averaging, it refreshes around once a second, so maybe a simple averaging of samples over a second would help to smooth out the data? I'm not sure how to accomplish this.

I have the shield on the sensors wired into the ground, so hopefully that isn't an issue. How would I go about checking for a ground loop?

I wouldnt think it was something caused by my programming, as I said above the program is just reading, displaying and writing to a file.

I'll play with the sampling rate and try to isolate the sensors to reduce noise and see if it helps. I'm not sure that the pressure would fluctuate as much as it shows in my readings, but the pressure system isn't my design so I will look down that avenue as well.

I really appreciate the input and will post back if I figure out any specific cause.

Thanks

[aeastet]

How fast are you currently sampling?

Ground loop would be if you have a ground at both ends of your sensor and they go to two different places.

[kinghuds]

Ive varied the sampling rate from between .5 and 100 per second. But the samples per channel on the Timing Vi is set to 10,000. Maybe thats where my problem is?

[johnsold]

You mentioned a small duct.  How fast is the air moving in the duct? Can you hear it "hissing" or "rushing" if you put your ear close to the duct? How is the air forced through the duct? What type of fan or blower?  What is the bandwidth of the pressure transducer?

I am thinking that your fluctuating readings may be due to acustic noise or buffeting from fan blades.  Both of these sources might be too fast to be recorded by the manometer but could be picked up by the pressure transducers.

If the transducer bandwidth is greater than the Nyquist criterion for your sample rate, you will need an anti-aliasing filter.  This is an analog filter between the transducer and the digitizer.

Lynn

[Henrik_Volkers]

What PXI DAQs are you using?

Take a close look on your signal: Sample with 10kS/s (and more) at least 100ms (1000points)  and about 1s (10000points).

Make .png of the graphs and post it here.

[kinghuds]

Ill find out about the pressure system, I have only been working on the data acquisition side so I don't know what the set up is, they are a bit isolated from the system as they are attached to tubing, but I will check on the noise and fan levels.

As far as the bandwidth, all I could find for the sensors is

Bandwidth: DC to 1 kHz typical

So not exactly sure how to interpret that as far as my use goes.

I am using the sensors wired into an SCB-68 which connects to a PXIe-6341, with the sensors both getting 10V from a seperate power supply. I just got back in to work on it, so I will try out those sampling rates.

[Ben]

is your sensor a strain guage absed sensor? These normally have a spec that reads "mV/V".

Ben