DAQ 6230 Trigger

[Clint_Eastwood1000]

I have a PXI6230 DAQ.,..250KS/sec or 1 sample every 4 micro sec.  If I externally trigger the rising edge analog input, is the 1st sample taken at T0 or 4 microsec later? My pulse COULD BE 100 mv/10 microsec sqwave.  my reqment is that if its 100 mv duration is <=8 microsec then pass else fail so I'm going to trigger on the rising edge then look at the next 3 ?? elements. 1 at 0 sec, 1 at 4 microsec and one at 8 microsec.  If the mag of each of these elements meets spec it passes.  Any comments ?...  I have to work w whats been given to me.  I would have picked a better daq if I was buying one.

[John_P1]

The first sample will by default be taken pretty close to the trigger.  There two configurable delays that will introduce a few timebase ticks of latency between the trigger and the first measurement though:

Start.Delay (DAQmx Trigger property), its minimum (and default) is 2 ticks of ai sample clock timebase.

DelayFromSampClk.Delay (DAQmx Timing property), its default is 3 ticks of the ai convert clock timebase, but it can be set to 2 to minimize delay.

The minimum of 4 timebase ticks of delay would (nominally) equate to 200 ns when using the 20 MHz timebase (the max timebase on this device for ai).  The trigger signal is asynchronous to the timebase though so there is an additional 50 ns of uncertainty (the end result should be a delay between 150 and 200 ns).

Having said that...

With the device you have, you could instead convert the external signal to TTL (you're already providing a TTL signal for the trigger, so it sounds like you're at least pretty close to this already) and use one of the counters to make the measurement.  You would be able to measure the width of the pulse with a 12.5 ns precision (the counters can utilize the on-board 80 MHz timebase).

Best Regards,

[Clint_Eastwood1000]

Thanks John. 

I don't want to introduce delay so I guess I wouldn't be using those vis.  Could you point me to any KB or tutorials that show how to use the external triggering w that daq?

When you say ..."you could instead convert the external signal to TTL" is this done w some kind of external circuitry??

[John_P1]

The delays exist whether you configure them or not, the default is 5 timebase ticks total but you can reduce this to 4 timebase ticks by explicitly setting DelayFromSampClk to 2 ticks instead of the default 3.  You can't set a delay less than 2 timebase ticks for either of the two delays.  If you're only interested in whether or not the pulse is less than 8 us though you could increase the total delay to 4 us and know that the second sample would occur exactly at 8 us.

The 6230 only supports digital triggering, so to even use triggering in the first place you need to generate a TTL compatible signal to give to the 6230.  Once you have that, configuring the trigger in software is fairly straightforward by adding the DAQmx Trigger VI to your task.  However, if you're going to generate a TTL compatible signal anyway, I would strongly consider just going all the way and using the counter approach to get a usable measurement of the pulse width.

If you don't use triggering at all and just sample continuously at 250 kHz, you could parse the result in software to detect pulses.  However, you would have a huge uncertainty to your measurement.  For example:

0 high samples => you might have missed a pulse up to 4 us long.

1 high sample => the pulse is somewhere between 0 and 8 us long.

2 consecutive high samples => the pulse is somewhere between 4 and 12 us long

3 consecutive high samples => the pulse is somewhere between 8 and 16 us long

etc.

I can't imagine this approach would be suitable for your application.

Best Regards,