meauring the period of a differential TTL signal

[shyammenon]

Hi all, My system has a pci-6052e series card connected to an scxi-1000 chassis and i am using an scxi-1302 terminal block to access the counter pins from the card. I have a differential square wave (ttl compatible according to the manufacturer) signal that i want to measure the period (and frequency) of. The signal is a low-mid range frequency signal (upto 300 hz). When i use the measure period.vi to measure the period of the signal (which i connect to the gate pin of the counter), i keep getting incorrect measurements (i am checking using an oscilloscope). But the same vi works perfectly fine if i use a pulse train generated by another counter(even at very low frequencies). So i am wondering if it is because this is a differential ttl signal and if thats so, how do i fix this problem. Where do i connect the reference end of the signal to..? Would be glad to get any help i can on this.
Thanks in advance, Shyam.

[ShotSimon]

Sounds to me like a signal level problem.

Can you get an o-scope screen shot of the working signal and the non-working signal?  Which of the two signals are you driving your logic with?  If for example you are using the low end, you may want to try using the high.  Sometime differential signals use a different reference than ground say 2.5V and one signals swings to 5V while the other swings down to 0.  The level of the signal you are using to drive your logic may not be adequate.

Regards,

-SS

[shyammenon]

Hi, I am attaching a picture of the scope outputs from the two signals. Is there a grounding pin on the scxi-1302 where i can connect the reference (ground) of the signal..guess i should be checking the manual!!.
Thanks, Shyam.

[AnalogKid2DigitalMan]

I can't get to the manuals now (slow web) but could it be a slow rise time issue on your sensor signal? You may want to look in the manual for any min/max risetimespecs on the counter inputs. Signal looks TTL compatible, you could use a schmitt trigger circuit to get faster edges.

[ShotSimon]

Can you include the excel file so we can see the rise time between data points and transistions.  Sorry but your plot is formatted so it's hard to make out details other than that the amplitudes look the same.  -SS

[shyammenon]

Hello..Thanks for the quick responses..am attaching the excel sheet here.
Shyam.

[shyammenon]

Hi all..Still no luck.. I tried to implement the schmitt trigger..acquired the waveform on an analog input vi..acquired waveform components and then generated a new waveform based on if the component value was high or low..thought that cleaning up the waveform would help. but when i output the new waveform using scxi-1325,1124 modules and routing the voltage output to counter gate and ground to digital ground on 1302..the measure period vi still gives me wicked values..unlike the oscilloscope which gives good measurements for both signals. I am starting to think if there is a  simple way to get teh frequency value from my scope to labview..so i dont have to use the counters at all!
Shyam.

[ShotSimon]

Does your scope have GPIB?  -SS

[shyammenon]

Yes..it is a textronix 3034b scope with gpib output. I saw that there is a set of instrument vi's available for this scope from labview..but only for versions of labview 7 and above..while i still have labview 6.0 and doesnt seem like i can upgrade right now!
Shyam.

[ShotSimon]

You could start a new thread and see if people in the community will back convert it to 6.0 for you.  I have done this with other vi's.  Takes about a day assuming the vi are backward compatible.  If you just need to make a frequency measurement it may be as easy as sending a few GPIB commands to the scope using the programming manual. 

On a side note here is your excel in a easier to read format, rise and fall are around 0.1ms.

-SS

Message Edited by ShotSimon on 07-18-2008 03:01 PM