Frequency measurement of a very short pulse

[tintin_99]

Hi I am using PXIe 6368 to measure Oscilliator Frequency. This signal is connected to PFI lines and I would like to use counters to measure frequency but the pulse width is too brief ( 12 ns ) . How can I determine if the hardware counter can even detect this short pulse for frequency measurement?

Thanks 

[Norbert_B]

How does the signal look like exactly?

There seems to be no detailed specifications on the signal form in order to be counted correctly (rising/falling time, minimum time to stay at a level (H, L), overshots, ...).

As the internal clock of the device is 100MHz, the decive can handle signals with 5ns H/L time, at least internally. But as generation is limited to 25MHz, it is possible that it is limited to 20ns....

In reality, i would expect the device to be somewhere in between these two numbers if the signal is "perfect" TTL/CMOS signal.

Norbert

[GerdW]

Hi,

There seems to be no detailed specifications on the signal form

The specs clearly state "TTL" - this includes signal form definitions. TTL describes more than just logic levels…

[Norbert_B]

Gerd,

i am not aware that TTL specifies detailed timing information. The main reason: TTL was introduced as "standard" in 1961. In the mean time, all our semicondictors are way faster than back in that time. So if TTL would specify timing in detail, there would be the requirement for different versions of TTL (just like USB 2.0 vs. USB 3.0).

I don't know about TTL 2.0 (or higher), so i doubt timing specs to include hard numbers.

NI specified TTL signals for older controller chips (compare here) but i am not aware of any updated document for STC3 chips which are used for X-series devices.

Norbert

[GerdW]

Hi Norbert,

NI explains all the needed details (including rise times) here.