PCI-4472 self calibration error under MAX

[Ron Norton]

Hi,

Our PCI-4472 appeared to have offsets on the order of several hundred microvolts, so I tried a self calibration under MAX (version 4.1.0.3001) to try to reduce them. The self calibration gave me the error message:

"Measurement taken during calibration produced an invalid AI gain calibration constant."

Does anyone know where I can find troubleshooting information for this issue, or have any suggestions on how to fix it?

Thanks,

Ron Norton

Faculty Research Assistant

Gravitation Experiment Research Group

Department of Physics

University of Maryland

College Park, MD

[Abhinav T]

Ron,

This is an error in the 4472 hardware and you will have to send your board in for repair. Please create a service request from www.ni.com/ask and one of the Applications Engineer will be more than happy to help you out with your repair.

[Farside1]

I am posting this for an applications engineer, intended for this to post last in a linear fashion ~

I have attached screen grabs of the self test and calibration panels.
When running the "self-test" panels within MAX, all tests pass.
I performed the tests after hitting "reset" in MAX.

Current versions of software installed: Labview Full Dev 8.2., and Max 8.5.
All upgrades have been applied.

Ai1 is setup to measure a voltage differential of 1-5 VDC based upon a "differential" input
across a precision resistor. I have setup custom scale, 1V = -40`C 5V = +180`C.
I have tried setting sample rates of: 3 @3Hz, 10@60Hz, 100 @600Hz, etc...

As I run the self test, and it's the same when I launch Labview and work within a VI, I see
the wave output with what looks like a square wave. I have put a scope on the wire pair
and do see steady voltage levels. Sensor has a current loop supplying external excitation.
Sensor output passes thru a linearization circuit operating at 60Hz.

In the 3rd image I captured Ai0 which is similarly setup in "differential" 1-5VDC input to scaled output.
You can see the graph image captures the trailing signal just like on Ai1 analog channel.
I have verified from Omega that the sensor is wired correctly supplying current loop,
and also taken the wires and "twisted" the pairs for less EMI/RFI crosstalk/noise.
The sensor and it's power supply are brand new purchased directly from Omega.

I am planning to send this to a local vendor for calibration, SE Labs in Santa Clara, but if it's
a matter of the boards circuitry being damaged than it needs repair. I started here about
3 months ago and after asking around found the PCI-6250 DAQmx board was available.
Problem is I don't know the history, besides the 2 year calibration expiring Nov of 2006.

The 1 thing I have not tried is moving the PCI to another slot, but FYI I have just this week
moved the PCI card over to another workstation that exceeds minimum hdw requirements.
This card was installed in a Pentium III @848MHz workstation with 424MB of ram.
Facts are the graph output of both channels look exactly the same in either pc.

The previous user was an engineer, whom may have exceeded voltages/parameters
on the analog and/or digital inputs. Thank you again for your knowledgeable replies!
Sincerely,
Phil Johnson
Hardware Technician
http://www.digitalpersona.com/

[Farside1]

Script error? Attached files here ~

[Farside1]

More images captured ~

I measured a battery @1.625VDC with a DMM, connected to Ai2 as "differential" input,
measured input max 10V, min -10V and ran "task" to acquire analog signal.

Images show what sampling rates were tried ~

[Abhinav T]

Hi Farside1,

This defenitely looks like a repair and not a calibration. The board should not show a square wave for a constant DC voltage.

Once again you should contact an Applications Engineer at www.ni.com/ask to get support over the phone. This is most likely a repair of your board.

Do all other functions work? What if you setup analog output and then loop it back into analog input?

[philj@digitalpersona.com]

I had dis-connected the cable, trying both scenarios below ~

With SCB-68 dis-connected from cable (cable connected @ DAQ)
the signal looks different, sort of a combination of a
sawtooth/square wave on the graph.

With cable dis-connected (at DAQ card input) the signal is
same as with cable connected to DAQ and not to SCB-68. The
only difference is the amplitude jumped from ~1.3 to ~3.0.

As I keep running the task, on different analog channels,
the amplitude drops steadily, like a capacitor dis-charging.

I was thinking of creating a loop back to test input, even if the
output signal generated might be questionable I can hookup
scope to measure and verify what signal actually looks like.

I will perform this before Monday, but am already in touch
with app engineer to request repair quote.

Device: PCI-6250 Multifunction Card
Serial Number: Z21473636
Product Number: 779069-01

Thanks,
Phil