Running 6 counters at the same time

[Euaruksakul]

I want to use 6 counters in NI PCIe-6612 at the same time to count pulses from 6 channeltrons, and I found that the counters miss some counts when I use TTL out of a function generator as a test source. For example, it shows ~40 counts in one second when I apply 60 Hz TTL pulse. The fluctuation is also quite large (~30%). When I use only one counter at a time, it reads more accurately, but still ~10% less counts than what it should be.. Is it possible that the computer cannot take the load since it's quite old (about 8 years, Pentium 4 with 1GB RAM).

[Dennis_Knutson]

You need to attach your code. You are hiding everything by using the DAQ Assistant.

[TJWhit]

---

@Euaruksakul wrote:

I want to use 6 counters in NI PCIe-6612 at the same time to count pulses from 6 channeltrons, and I found that the counters miss some counts when I use TTL out of a function generator as a test source. For example, it shows ~40 counts in one second when I apply 60 Hz TTL pulse. The fluctuation is also quite large (~30%). When I use only one counter at a time, it reads more accurately, but still ~10% less counts than what it should be.. Is it possible that the computer cannot take the load since it's quite old (about 8 years, Pentium 4 with 1GB RAM).

---

How often are you reading the count? I have found that when I use 6612 or 6602 counters in event counting mode with random (or at least asynchronous) events, there is a dead-time after the sample clock pulse is generated to read the count. Therefore, the more often you read the count, the more likely you are to miss counts. I have not seen a problem with using multiple counters but I don't think I've ever used six counters in event counting mode. 

Regards,

[Euaruksakul]

It is about once every ~2-3 seconds. I understand the each of them counts simultaneously independent of each other?

[TJWhit]

The effect I'm talking about requires about a thousand reads to miss an event. Your issue seem to be much worse than this so it is likely not the pulse collision problem I was describing.