6259 AI-Values are different to the values in the driver monitor panel

[LutzGeis]

Hello,

i have a problem measuring analog signals with a 6259.

For example:

- i want to measure 2 DC signals with 4.999V and 4.975V

- DMM shows the correct vaues

- the driver monitor panel shows the correct values

- the LabWindows/CVI software and one handle for each signal shows the correct values

- the LabWindows/CVI software and one handle for both signals shows 4.999V and 5,149V

Lutz

[Wolfgang_Z.]

Hello Lutz!

Please check the input terminal configuration (Differential, RSE, NRSE) of your CVI  program.
Do you have the same settings like in MAX?

Do you have a floating or a grounded signal?
In this measurement tutorial http://zone.ni.com/devzone/conceptd.nsf/webmain/177A8B29FEDC0F5886256FA90083C0F8 in Table 1 there is shown a summary of the recommended configurations for floating and grounded signals.

I hope it helps.

Regards,
Wolfgang

[LutzGeis]

Sorry that it takes some time to answer !!

The terminal configuration ist : Diff and there are 2 grounded signals.

The values are different if i use one handle to measure both signals. With one handle for each signal

the AI shows the correct values

I think the problem is the settling time for input of the A/D-converter. With a high output impedance

connected to the AI it takes a longer time to load the A/D-converter input to the new voltage. Setting down the

sampling rate does no effect. I have to slow down the rate that switches the multiplexer from channel to channel

in one handle so the A/D-converter has more time to change his input value to the new one. With an output resistence

of the unit to test at 3kOhm i have no problem but with 21kOhm it is a problem.

regards Lutz

[Wolfgang_Z.]

Hi Lutz!

The recommendation of the ouput impedance for the M-Series boards is 1KOhm.
I think you need an impedance converter to measure a signal source with an impedance of 21kOhm.
An other possibility is to use an S-Series board without multiplexer.

Regards,
Wolfgang

[LutzGeis]

Hello Wolfgang,

thanks for your answer.

After some more reading this forum i found an application note (AN045)

about the data aund the settling time.

I hafe to measure 2 different types of sensors. Each sensor basicly consitsts

af a simple bridge. The resistors of the bridge are effected of a magnetic field.

Each sensor has 2 bridges ( sin + cos )

So you can measure length with a magnetic yardstick the number of revolutions

withe a mag. field rotating over the sensor. We do this measurements on waferbase

with each single sensor. For the length sensors it is no problem to measure the sin and

cos signals separate in one handle each signal (Labwindows/CVI 7.1) but for the angular sensors

it is important to know the time dependence of the sin + cos signal so i have to measure them

in one handle.

An other possibilitie is to measure the sin + cos signals together withe a reference signal that makes it

possible to assign the signals to the angle of the mangnetic field. This is o.k. because the settling time error

concerns not on the 1st channel in a handle. So i can place the important signal at the 1st position in the handle

and the reference signal (hallssensor) in the 2nd. But the synchronisation of the signals costs time an time costs

money.

Regards Lutz Geis

[LutzGeis]

Is it possible to decrease the multiplexing rate  ??

Regards Lutz

[Wolfgang_Z.]

It is not possible to decrease the multiplexing rate. The switching time between the channels averages ~5us.
If you need simultaneous sampling you should use a S-Series board.

Regards,
Wolfgang

[Tom_W_[DE]]

Hi All-

Actually, it IS possible to decrease the rate of the AI Convert Clock (which controls the rate of multiplexing).  The best solution is to decrease the source impedance, but changing the timing parameter may help if that is not an option.  This setting is a DAQmx Timing Property and can be controlled by the DAQmxSetAIConvRate(TaskHandle taskHandle, float64 data); function.  Keep in mind that the convert rate you choose must be at least as fast as [Number of channels in task] * [sample clock rate].

Hopefully this helps-

[LutzGeis]

Ahhhhhhh ......................

Is this a "hidden function" ? It is not described in the DAQmx C Help.

Lutz

[Tom_W_[DE]]

Hi Lutz-

The AI Convert properties are listed under the comprehensive List of Timing Properties.  You can find this list by looking in the Contents tab of the NI-DAQmx C Reference Help.

Hopefully this helps-