Author Topic: Oscilloscope to Function Generator Trigger Delay  (Read 364 times)

0 Members and 1 Guest are viewing this topic.

Offline arduic

  • Newbie
  • Posts: 3
  • Country: us
Oscilloscope to Function Generator Trigger Delay
« on: January 27, 2020, 01:25:57 am »
Simple question, is it common for there to be a delay between an external trigger out from an oscilloscope to the trigger in of a function generator?

Specifically, I have a Rigol 1054Z oscilloscope and a Rigol DG1022 function generator. The oscope is measuring a signal for a rising edge trigger, the trigger out signal is connected to my function generator via a BNC. My function generator is set to be a pulse for 350ns configured in burst mode with an external trigger. This output is connected to a T junction with one side connected to a Max chip and the other to a BNC connected back to my oscope. I'm measuring ~750ns of delay from when my trigger fires to when the function generator outputs. Is this normal? Is there some setting that could improve it? I tried looking in the datasheets but I do not see anything mentioning the external triggers and some delay except maybe some 10ns or less fields.

915416-0
Yellow=Control, Purple=Signal from generator, Blue=Output of Max chip.

Any hint as to this being normal or not would greatly help thank you.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 10493
  • Country: us
  • DavidH
Re: Oscilloscope to Function Generator Trigger Delay
« Reply #1 on: January 27, 2020, 10:05:38 am »
Modern DSOs use digital triggering so there are two delays.  One is caused by the trigger qualifier as with analog oscilloscopes and the other is a random asynchronous delay between the trigger and processing clocks.  At higher speeds, the trigger out is more useful for only indicating the waveform acquisition rate.

Analog oscilloscopes and DSOs with analog triggering have a short fixed delay which is typically 10s of nanoseconds when level triggering is used.  On most this is actually a sweep gate output but it is practically the same thing.

On the Rigol 1054Z you might try fiddling with the trigger AC coupling.  In the past when it was broken, it affected how the trigger output was processed.
 

Offline arduic

  • Newbie
  • Posts: 3
  • Country: us
Re: Oscilloscope to Function Generator Trigger Delay
« Reply #2 on: January 27, 2020, 12:46:35 pm »
Thanks for the reply just want some clarification here. Besides messing with the AC coupling and kind of just praying (which I'll do tonight) there is not much I can do to reduce this delay correct?

The only other thing that confuses me is how my scope can fire the trigger at t=0 so perfectly but then not externally release the trigger for so long. My only guess here is that the scope is buffering data within one waveform acquisition, processing it to see if a trigger has occurred, and then sliding the output back to t=0 for where it finds the trigger and the delay I see here is how far behind the buffer the trigger processor is. That would seem to line up with your mention of trigger qualifiers (I would hope removing extra qualifiers would reduce the delay but I guess not since I currently have none running).
 

Offline TurboTom

  • Frequent Contributor
  • **
  • Posts: 704
  • Country: de
Re: Oscilloscope to Function Generator Trigger Delay
« Reply #3 on: January 27, 2020, 01:46:09 pm »
Some time ago, I did a test of the trigger event -> trigger out  delay and jitter of some Rigol scopes. You find this contribution here: https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg1068070/#msg1068070
 

Offline arduic

  • Newbie
  • Posts: 3
  • Country: us
Re: Oscilloscope to Function Generator Trigger Delay
« Reply #4 on: January 27, 2020, 02:11:54 pm »
Thanks for the write up Tom, that was pretty helpful. Since you are measuring only half the delay I was I am curious. Do you think this delay stems from my source signal ramping up while yours was straight from a BNC? Or could it possibly be my setup in some other way? I don't think the function generator would be adding this much delay but maybe I'm crazy. Any suggestions on a way to test and find the source? I guess recreating your setup with a slow migration would work but I only have the 1 scope and 1 generator so I'm rather limited.
 

Offline TurboTom

  • Frequent Contributor
  • **
  • Posts: 704
  • Country: de
Re: Oscilloscope to Function Generator Trigger Delay
« Reply #5 on: January 27, 2020, 08:57:14 pm »
I never tested trigger in -> signal out delays of AWGs but I can well imagine that this may take several hundreds of nanoseconds. I haven't got access to a DG1022 but to several other AWGs (DG800, DG5000, SDG6000X) so I may do a comparison when time permits.
 

Offline graybeard

  • Regular Contributor
  • *
  • Posts: 157
  • Country: us
  • Senior III-V RF/mixed signal/device engineer
    • Chris Grossman
Re: Oscilloscope to Function Generator Trigger Delay
« Reply #6 on: January 28, 2020, 03:47:55 am »
You will get 1.54 ns/foot of delay in RG58 coax, but that does not account for the large delay you are measuring unless you are using an exxceptioally long test lead.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 10493
  • Country: us
  • DavidH
Re: Oscilloscope to Function Generator Trigger Delay
« Reply #7 on: January 28, 2020, 03:54:38 am »
The only other thing that confuses me is how my scope can fire the trigger at t=0 so perfectly but then not externally release the trigger for so long. My only guess here is that the scope is buffering data within one waveform acquisition, processing it to see if a trigger has occurred, and then sliding the output back to t=0 for where it finds the trigger and the delay I see here is how far behind the buffer the trigger processor is.

If I understand you correctly, that might be deliberate.  On an analog oscilloscope the gate output is asserted during the entire sweep making it handy as an indicator that the sweep is still active among other things and that another trigger will be ignored.  Rigol may be doing the same thing making it more of a gate output than a trigger output.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf