Simple Stepper move with synchronised analog data aquisition

[Toneles]

Hi there,

following application:

a stepper moves max. 12 turns forward and backward (2 to 3 secs each direction)

during these moves a analog value (0 to 10V) needs to be sampled with around 1500 to 2000 S/sec (each step a value)

the result should be a table (graph) using y for the analog value and using x for the stepper-position

My questions to you are:

_will a 7332 in combination with a 6220 (using RTSI) serve this application?

_will it be necessary to use a 7352 instead (periodic breakpoints...?) ?

_since the stepper has no enocder - can I still use RTSI to trigger the DAQ-card AI with each commanded position instead?

_how about only using a 6220 card (to accomplish the simple moves of the stepper and measure the analog value) ?

I haven't bought any hardware yet.

Maybe there is still a more enlightening solution to this application.

Ciao

[Kevin_Price]

How important is the stepper's velocity profile to your measurement?  A 6220 board by itself can easily configured to use a counter output as its AI sampling clock, helping to sync your analog measurement to your step command.  As for generating the step pulse sequence, that'll depend on your needs.  Here's what you can expect with the 6220 alone:

1. Can generate fixed # of steps at constant rate.   Both counters work together to generate finite pulse train.

2. Can generate "continuous" pulse train whose freq can be changed on-the-fly.  Will be difficult to accurately control total # steps generated, but can count them with other counter.  Freq changes through software calls, so can not count on smooth ramping.

3. Can use "correlated DIO" to generate a more precisely-defined trajectory using a hw-timed digital output.  Will take just a little more work and/or wiring to  get this in sync with analog input.

-Kevin P.

[Toneles]

Kevin, concerning your question:

The velocity-profile can be as simple as a linear ramped one (constant accelaration/deccelaration) driven by the drive's and physical restrictions.

The path for itself is predefined and does not need to be changed on the fly (just a quick halt before the end of the move depending on the readings of the AI may become an issue)

So, pretty much I'd go for the correlated DIO. Hence, some more open points:

_can I generate a pulse train (the whole profile of one move including acc/decc) in advance?

_how hard will it be to route this pulse train as sample source for the AI?

_DIOs on 6220 need a sample clock from another subsystem or external source - what does that mean (to this setting)?

_any instructions on the web?

For testings for now I solely have a 6036E card.

So far so long

Nevertheless can still somebody tell me about the concerns using the 733x board to trigger a DAQ-card via RTSI using the commanded position instead of encoder feedbacks (refer to opening message)

[Abhinav T]

Toneles.

- Yes you can Generate Pulse Train using the coutner output (There are several examples of these within LabVIEW, C/C++ and VB)
- You can internally route the output of a counter to be the sample clock of an Analog Input in the Timing VI (using DAQmx)
- DIO can not use its own internal clock, but it can use the AI Sample Clock (which is variable). Please refer to the M-series manual for more information on that (Page 6-4).

You should not have issues routing lines over RTSI between the Motion board and the DAQ board. Are you trying to trigger the DAQ card off the Motion card or vice versa? Is the analog input triggered off the Motion card or the DAQ card?

Abhinav T
Applications Engineer
National Instruments

[Toneles]

Abhinav,

I'm moving a motor forth and back and sample at the same time the analogue value of a force measuring device. At the end I need a force-distance diagramm.

Several hardware configurations can be possible:

1. Using simple a DAQmx-device which outputs steps and direction to a servo drive. The drive does the position control loop. The analogue input gets sampled by the counters internal output or by the encoder feedback attached to the motor.

2. Using a 7332/7342/7352 controller in combination with a M-series 6220 device. The advantage are ramps and stuff I can do with the motion controller. Via RTSI the motion controller sources the sample clock for the AI of the DAQmx device.

Since the servo drive excepts step and direction signals the 7332 should be good enough for the task. From NI I was suggested to use the 735x series. Is that really necessary? And, can I use only the meassured position (of an encoder feedback) as sample clock which goes via RTSI to the DAQmx-device? Can I also use the commanded position instead?

So long.

So long.

[T-Route]

Toneles,

They probably suggested the 735X series because it has built in Data Acquisition to the motion card.  The Data Acquisition is not real great and I do not think will be good enough for your application.  It has high resolution but it can only be software timed meaning you will be constantly polling for the values.  So I believe the 733X (only stepper control) combined with the 6220 card should be sufficient for your project.  You could easily route the two cards triggers together so they start at the same time.  It will then just be a matter of you determining the amount of time for each step and acquire the correlated measurements.  Good Luck.

Mark T
Applications Engineer
National Instruments