Minimum hardware for precision timed digital outputs

[eldiegoefe]

Hello. I need to control some instruments connected with a multifunction DAQ through its digital lines, during 8 channels analog acquisition.

With a counter I set up a stimulus pulse at a rate selected by the user between 0.1 to 5 Hz. With different delays after that pulse I need to drive 4 digital outputs. I need the delays (and the pulse widths generated) to be accurately set with a resolution of about 1 to 10 microseconds if possible. I understand that I cannot do it polling the counter through software. Do I need "hardware-timed DIOs" or "correlated digital IO" boards? Or is it a way to set this thing through NiDAQmx VIs?

Right now we are working with a USB-6215. Which is the minimum hardware that I need (considering also the simultaneous 8 channels acquisition, each one at a 25 KHz rate)??

[tmatos]

Hi,

Considering you are using an USB hardware and Windows, you don't have determinism to control your I/Os with an accurate time, if you take a look at NI USB-6215 datasheet, you can see that it's clock is controlled by software. For your task you'll need a hardware-timed I/O as you told. To help you with a hardware set that will meet your needs, we need to understand a little more about your application and what you are trying to do, things like if this 8 inputs are logically related with the 4 outputs so you'll need a deterministic application for this and etc.

You can provide us with a little more information about your application here so we can help you or you can call the closest NI office to you so one of our field engineers can visit you and set a specific hardware to meet your needs.

We wait to help you.

[John_P1]

You need something that can clock out at least 4 hardware-timed signals (and maybe a hardware-timed clock for a stimulus pulse):

The 6215 does have:

• 2 counters which can generate single pulses with a configurable duration and delay (12.5 ns resolution) after your stimulus pulse.
• 2 clocked AO lines that can be updated at 250 kHz (4 us resolution) and triggered by the stimulus pulse.  The AO lines on the 6215 are not retriggerable but you might be able to re-arm them in software since your stimulus pulse is at such a low rate.

The catch is that using the above to generate your 4 signals doesn't leave you with a hardware-timed stimulus pulse (you mentioned you are already using analog input, so you can't export the AI sample clock for this purpose).  If your stimulus pulse timing doesn't need to be too accurate, then perhaps you could get by with timing it in software (toggle a digital output line every N ms based on a SW-timed loop).

So, you'd be much better off with something else, but maybe the 6215 can get close enough to what you need.  

If you are able to purchase new hardware for this application, I'd look into a USB-6341 or a PCIe-6320.  They each have 4 counters for generating the 4 pulses you need (at 10 ns resolution) and meet your analog sampling requirements.  The exact 0.1 to 5 Hz clock can still be generated using a digital output (or just any digital I/O task with an exported sample clock--both these devices support hardware timed digital I/O).  Or, you could generate the 4 pulses using clocked digital output (1 us resolution) and continue using a counter to generate the stimulus pulse clock signal.

Best Regards,