Pressure signal from DAQ keeps changing daily.

[encore]

I would start by reviewing the specs of the pressure transducer.

http://www.meas-spec.com/myMeas/download/pdf/english/msp/msp_300.pdf

This indicates accuracy is <1% fullscale at 25C, or 2% over temperature.  The 'Electrical' section of the spec on page 2 says the amplified transducers have a zero offset of +- 2%, and a span tolerance of +- 2%. Your output range is 1-5V, so 2% FS is 80 millivolts.

You're concerned over a drift of 2 millivolts, yet this is within the transducer's accuracy specification.

[Xiaoting]

Hi John

Thank you for your reply. I'm sorry for this late response, but I spent quite a while communicating with the others. The manufacturer hasn't come back to me  for my enquiry. I had a few discussions with the person who sold the pressure transducer to us. They are saying that they are not convinced that the pressure sensor would be a problem. They are going to replace MSP300 with a MSP600 with a better accuracy for us, but there is no guarantee this one would work, since they think the problem is caused by somthing else.

They said the electrical ZERO OFFSET should be an absolute value which shouldn't change daily.

I'm wondering whether you or your colleagues have any expereince in using a pressure transducer to get a Load Reading from a Hydraulic Tensile Tester and what types of pressure transducers they used? The load range is 0-300kN and the pressure range is 3500psi. I would like to compare their specifications with ours.

Thank you and look forward to your reply.

Kind regards,

Xiao-ting

[hoffiz]

Hi Xiao-ting,

I’m going to be taking over John since he is out of the office. The first thing I did is to read the entire history and I can see that all the necessary troubleshooting steps have been taken so let’s take a step back and analyze everything from the beginning.

I would like to go over this statement: “We've been told by our sales engineer that this is caused by the pressure loss in the hydraulic system due to daily use, which is tolerable, and can be eliminated by set up a reference zero value when the machine ram is floating and ready to load.” If you take a look at all your readings: the first one is 1.604, 1.061, 1.059, 1.053, 1.026 does this seems to be a due to what you sales engineer said? If this is the case the calibration you apply to get the zero offset needs to be change daily, right?

What do you think about “encore” post I did look at the specification and the 2mV offset seem to be expected on the FS range.  “They said the electrical ZERO OFFSET should be an absolute value which shouldn't change daily” but this would be the ZERO OFFSET of the sensor not the one for the Hydraulic Tensile, right?. Is there any chance you can use this sensor to read pressure from something else rather than the Hydraulic tensile? Can you read Standard Atmosphere?

I will wait for your response

[Xiaoting]

Hi Jamie

Thank you very much for your reply.

Yes, at the beginning, we were reading some document about the flaws in using a pressure transducer to reading load on tensile tester, which include 1) moving the cylinder to overcome friction between piston and cylinder; 2) raising piston off the cylinder bottom 3) weight of oil. The pressure from these sources creates errors in the indication of load.

Then we realized that the original dial gauge of this old machine is reading the pressure of oil. Therefore these must be compromised in a way.

However I'm not entirly sure about this argument. If you are right, we will never be able to calibrate the whole setup to ISO standard. They won't allow for the various zero offset. And this zero offset changes from morning to afternoon as well.

Yes, I realized the accuracy specified for the pressure transducer is not very good and here is what our sales person told me: '1) Although the MSP300 range is sold as 1% that covers all errors on every device over the full temperature range. In order to guarantee this any individual device has to be somewhat better specified. 2) the Electrical Zero Offset should be an absolute value which shouldn't change.' I'm not experienced in those equipments, therefore hard for me to judge whether this statement is right or not. But it is unclear to me.

I don't think I can use this sensor to read pressure from something else.

I realized the effect of atmosphere, but the oil is not sealed, thus the effect should be minimal, is it right?

Thank you and look forward to your reply.

Kind regards,

Xiao-ting

[encore]

You still need to go back to the transducer's fullscale range, and accuracy spec. You say it's 0-5000psi, with an output of 1-5V (4V range). That means you'll get 0.8mV per psi. If the transducer's accuracy is ±1%FS, its accuracy is ±50psi, or ±40mV.

Earlier you said it was specified at 0.1%, but then implied that 0.1% is 1mV, when it's actually 4mV:

I still can't find the reason for it. The machine has a original dia gauge which is reading the pressure and converts it to force, therefore any pressure friction loss within the hydaulic system should be compromised. The pressure transducer has a maximum error of 0.1%, therefore the 1.061-1.059=0.002>0.001.

In any case, it seems to me that the transducer and electronics are behaving within tolerance, but the tolerance you purchased isn't what you want.

[Xiaoting]

Thank you for your reply.

Can I ask another question about labview signalexpress for NI USB 6210 (DAQ)? Eventually we want to get both pressure and displacement readings simultenously and be able to plot a pressure-displacement curve. The pressure output is 1-5vdc, displacement output is counter output (we use magnetic strip and sensor for measuring movement).

We've been told that Labview Signalexpress cann't acquire both of them simultenously and we need to program a clock source. Is it true? Does anyone know how to do it? Thank you and look forward to your reply.

Kind regards,

Xiao-ting

[tnek]

Hi Xiao-ting,

I was wondering how you wanted to acquire the linear position data from the counter. Will the counter acquisition be a buffered acquisition or will it be a software timed acquisition? You should be able to acquire data from the pressure transducer as well as the linear position using two DAQmx Acquire steps. You should not need any external clock if you wanted to have a buffered analog input and a software timed counter input.  If there is no dependencies between the two, there should be a jagged line separating the steps. This designates that the steps will run in parallel. You can find some more information in this article. Try programming this acquisition in LabVIEW SignalExpress and let me know how it goes.

Regards,
Kent
Applications Engineer

[Xiaoting]

Hi Kent

Thank you for your reply. The output signal of Magnetic displacement sensor (MSK5000) are A: /A; B;/B; /I or R; /R. It is real-time signal processing. I have no idea whether it is a buttered signal or software timed signal. The Realtime requirement is 'real-time signal processing'.

Our Labview SignalExpress is only Limited Edition. I don't think we can add a Arithmetic Step to it. Is it true? Thank you and look forward to your reply.

Kind regards,

Xiao-ting

[tnek]

Hi Xiao-ting,

You are correct in saying that processing steps are not available in the limited edition. However, you should still be able to accomplish both acquisitions without the use of the Arithmetic step, it was just shown the article as an example. I would try creating the acquisition steps to see if the result meets your needs. Here is a screenshot of what the steps may look like.



Regards,
Kent
Applications Engineer


Message Edited by tnek on 04-23-2008 05:27 PM

[Xiaoting]

Hi Kent

Thank you for your reply. I can do that without any problem. But I'm just afraid that the readings are not simutenous. Like the displacement reading is taken at 8am, but the pressure is taken at 8:001am. Is there a way that I can find out check it out?

Thank you and look forward to your reply.

Kind regards,

Xiao-ting