oscillations at the output of NI9237

[krupa]

Hello,

I am using Ni9237 to which 3 force transducers are connected.I am measuring the Mod output directly to check if the values are in range but i see 2 things

1.The output is FXP.Why is the output of ADC a FXP??

2.Too much oscillations at the output.For this i tried calibrating with the exapmle vi provided by NI but the example vi does not serve the purpose.I have no clue what that does.

My main point here is to understand this fast oscillations,i must know how,where,why is it generated so i can compensate it.

Pls help me on this.

[Jacobson-ni]

First of all, what chassis do you have the 9237 in, and if it's in a RIO chassis, are you using FPGA mode or scan engine? If you are getting FXP results then I'm going to assume you are in FPGA mode in which case you can right click the module in the LabVIEW project and select properties where you can change the input from calibrated to raq to return the ADC value instead of FXP. The FXP that is returned is the value in engineering units. It's a lot easier to do all of your logic in engineering units which is why you have the option to just return that value.

I'm not sure what is happening with the oscillations but if you could tell us exactly what example you are running and show us what you mean by oscillating we might be able to help.

[Corey.Rotunno]

As Matt said, if you are using the FPGA module to program a FXP type value is prefferable. Especially if you are doing any kind of division or other types of math functions in the FPGA, floating point is very resource intensive. What I did for my cRIO application is work with FXP right until the point I convert everything to SGL for transmitting to RT via a DMA channel type FIFO. That way once your values get into the RT side of things, they are SGL floating points which will be much more familiar to work with.

Also, the oscillations.... to me that sounds like electrical noise, like perhaps something is floating. However, I cant offer any advice on that end as I have never used an NI bridge completion module, I have always used external transmitters that give a simple +/-10V output that I then measure using a traditional ADC module.

[krupa]

Thank you both for your kind help.

I am using CRIO-9073.I have connected 3 sensors,the sensors have a calibrated output value ranging from 0.012 till 1.99 mv/V.

The FXP values from the ADC under no  load condition:the values keep changing(that i mentioned as oscillating).It is expected to throw some value in the calibrated range with some error but since the value keeps on changing,i do not know what is the actuall no load value.

The same when it is loaded also.I expect it to output 1 or 2 values,so an averge value can be considered.But always there is continous and fast change in value in both no load and load casse.

All ia m looking for is that,it must provide output as per the sensor calibrated value may be with some errors.

[Corey.Rotunno]

How big are the oscillations? 

[krupa]

sensor 1 : 0.xxxE-5 - 9.yyyE-5

sensor 2 : 3.xxxE-5 - 8.yyyyE-5

sensor 3 : -0.xxxE-5 till +0.yyyyE-5

all are FXP outputs,since the sensor outputs in mv/v,i assume these values in mv/v.

When  converted to SGL,the 3rd sensor displays 0 most of the time,only at max load displays some osc. values

[Corey.Rotunno]

You are correct, the value output from the NI9237 should be in terms of mV/V. Unless you created a "Force (Bridge)" task. In that case it will be in engineering units, in which case your noise would possibly be acceptable. The below exerpt is taken from http://www.ni.com/tutorial/52778/en/

Like I said I have never used this type of module, so I am of little help....but ~.1mV/V noise does seem a bit high when your FS value is 2mV/V if you are truly reading mV/V. The data sheet indicates this module has internal filtering that I presume can be manually configured, perhaps that can take care of it for you?.

Good luck, hope you get it figured out.