Detect edge locations on a digital waveform

[tintin_99]

Hi,

I have a  digital waveform and I would like to detect transitions from 0 to 1 or 1 to 0  with their location ( or the edges ). Is there any Edge function that works with digital waveforms ? How can I do it in the fastest way possible without scanning through all the bits to detect transitions times

Thanks

[tintin_99]

ANy help really apprecaitred

[tintin_99]

Could you please let me know how can I detect locations ( in time ) of transitions in a digital waveform signal

Thanks

[Taki1999]

Have you tried anything yet? I'm pretty sure Search for Digital Pattern on the Waveform Palette will do what you're asking for.

[tintin_99]

Thanks Taki, Search for Digital Pattern can detect the firts 0 to 1 or 1 to 0 edge but I would like to locate all the edges 

So I think I need to put it in a loop and then change the start point everytime I search for a new edge. I think it is a very efficient way of doing this. Do you have any suggestions?

[Taki1999]

Sounds pretty straightforward to me.

Loop through using a shift register for the start and collect all the transitions.

Is there anything in particular about your dataset or application that has you worried that this won't be efficient enough?

[tintin_99]

atually in the previous post I wanted to say it is not a very efficient way fo doing this because everytime I search I should go through all the samples. So why not just search one time and find all the tarnsitions ?

Yes time is very important . I should do as fast as possible so it would be great if in just one search I can detect all the transition times 

[Taki1999]

---

@tintin_99 wrote:

atually in the previous post I wanted to say it is not a very efficient way fo doing this because everytime I search I should go through all the samples. So why not just search one time and find all the tarnsitions ?

Yes time is very important . I should do as fast as possible so it would be great if in just one search I can detect all the transition times 

---

The red highlighted part is not true. If you wire the start time appropriately, the search for the next match will proceed from where it left off. This will end up doing exactly one sweep through the whole dataset and return all of your transitions.

[tintin_99]

Great I didn't know that. However I still need to search 2 times .one time to detect 01 pattern and one time to detect 10 pattern. Is it true?

[Taki1999]

You could also sweep through with an XOR and then use the result to index back to timestamps.

Do you need to be able to distinguish between rising and falling transitions?