Using Trigger 1 as output

[av_disp]

Hello,

Being sort of disappointed in AD2 logic analyzer capabilities (see  the  post https://forum.digilentinc.com/topic/2004-ad2-continuous-digital-acquisition-speed-is-slow/  and similar for details)

I obtained a different logic analyzer with a lot of memory on board + support of RLE. But I do really like the flexibility of AD2 when it comes to setting different conditions/triggers soI decide to use it a master device/LA that should control the second LA  (aka slave LA with limited triggering capabilities but with a lot of RAM).

The idea is simple - AD2 uses Trigger1 as output and it is connected to Trigger input of the second LA. AD2 has very limited sampling range / almost non-existing buffer but as soon as it senses the right condition it triggers the second LA with very wide range (up to 2 G (RLE)) and that allows me to capture and analyze  quote long periods of time.    But the use of Trigger 1 as an output pin is confusing and I simply cannot explain its behavior. Eventually I started using the second LA just to observe the correlation between the scope triggering mechanism and the Trigger 1 on AD2 and have several questions in this regard. By that I mean the following: 

AD2 scope is connected to an AVR generating short pulses (about 8-9 ms period with 4-4.5 ms width). Trigger 1 is set to "Scope" - i understand it as the scope's triggering condition is mirrored on Trigger 1 pin. The scope runs in "singe" trigger mode and it works but  works "kind of" - see the  attached screenshot (I attached 2 high resolution screenshots providing all necessary info just please don't be lazy to examine it).

 Questions:

 a) Why did Trigger 1  generate 2 impulses (shown as Channel 1/TRIGGER1 on DSLogic)?  I would expect 1 and I cannot explain why 2 and not 3 for example.

 b  ) What defines the width of the triggered impulse (right now it is 182 ms)? I assume this is the number of samples * by the sampling rate. Am I correct?

More questions will come later, the post is already long enough.

Thanks.

 

[attila]

Hi @av_disp

You have an interesting trigger solution

I'm not sure if you have seen the Digital Discovery.
This does not support RLE (yet) but it is equipped with 2Gb DDR, which can store up to 256 MSamples of 8 signals at 800MHz.

1. Selecting the Scope trigger as output it drives the trigger signal high while it is triggered and acquisitioning, for post-trigger length of time.

2. The second pulse is a bit of bug when using single acquisitions.
The second pulse is due to the automatic rearming and acquisition of the device. This is also caused by trigger but the data is never read by the WF application.
Use Run to have one pulse per read acquisition.

[av_disp]

Hello @attila

>> Use Run to have one pulse per read acquisition.

I would gladly do that if I knew that it works as expected. But this mode also makes me wonder what is going on.

I am attaching 2 screenshots  (one is the zoomed version of another) for continuous acquisitions (i.e. RUN mode). I don't understand why the time between  trig. events  is not constant or at least "kind of constant", I mean the time between cursors 4 / 1  (105 ms) and 2 / 3 (170 ms).  Where does this delay come from?   In my view it can only happen if AD2 uses some sort of software implementation of triggering the signal on Trigger 1 output -  otherwise I simply cannot imagine why any reasonably fast hardware implementation triggering mechanism  would not be able to generate a steady flow of triggered events on Trigger 1 output at such low speed - we are talking here about hundreds of milliseconds.

Is there any workaround for this problem?

 

 

 

 

[HansV]

Hi, @ Atilla too

9 hours ago, av_disp said:

I would gladly do that if I knew that it works as expected. But this mode also makes me wonder what is going on.

I still have the erratic triggering problem as discribed under here, perhaps we experience the same problem?

Sorry if i make things only more complex but i hope that we find a solution for all.

Greetings,

Hans.

 

[attila]

Hi @av_disp

I will try to reformulate my earlier reply...
The Scope trigger output represents the instrument state: Ready->Prefill->Armed->Triggered->Done
The output will be high while this is Triggered state which lasts for post-trigger time length (from trigger arrow to the right edge of the plot). 
After the acquisition data is read by the software the instrument is restarted, going to Prefill then Armed state. On trigger condition it will enter again in Triggered state and output high.
The time uncertainty is induced by this software restart.

I think you are looking for Scope trigger detector signal. This would generate a pulse for each trigger event regardless of the instrument state.
Unfortunately this option is not available to choose from the application.

[av_disp]

Hello @attila,

Thanks for provided explanation, I see what is going on now. Looks like nothing is easy.

I found in the manual the following information:

----

Trigger Detectors and Trigger Distribution Networks are implemented in the FPGA. This allows real time triggering and cross-triggering of different instruments within the Analog Discovery device. Using external Trigger inputs/outputs, cross-triggering between multiple Analog Discovery devices is possible.

----

May I ask what FPGA pin is assigned to the trigger detector?  As a simple workaround I could simply use that pin at my own risk.

>2. The second pulse is a bit of bug when using single acquisitions.
>The second pulse is due to the automatic rearming and acquisition of the device. 

Well, if nothing else works, I can use an ATtiny85 and convert that  multi-pulse triggering event for single acquisition into a short, single impulse with 10-20 lines of C or assembler code. This solution will require just one extra element (AVR or other MCU).

Thanks.

 

[attila]

Hi @av_disp

You can find in private message a beta software version adding Scope and Logic Detector option.
These output a short 10ns trigger pulse which does not depend on the software rearming.

[av_disp]

6 hours ago, attila said:

Hi @av_disp

You can find in private message a beta software version adding Scope and Logic Detector option.
 

That's a very impressive customer support. I could not even dream about a possibility to add a new feature which solves my problem just in a couple days.

Thank you very much!

 

 

[av_disp]

Hello @attila,

The new triggering mode "Scope Detection" option works great. I am observing short (you said 10 I see 7.5 ns but I think that's the result of inaccuracy of my LA which has only 400 MHz max sampling rate) impulses on each trig. event, it runs in nonstop mode  ignoring the GUI settings (single vs continuous acquisitions) and I assume this is how it supposed to work.

At max. sampling rate of AD2 (100MHz) there is a small delay (300 ns) between the actual trig. event and generation of the impulse on External Trigger pin, which is more than suffice for general use measurements. The aforementioned delay  grows with the decrease of the sampling rate of AD2 which also fully understandable.

Thanks.

[attila]

Hi @av_disp

Thank you for the confirmation.

The delay is due to the ADC conversion time and device FPGA logic. 
The Scope detector by default uses averaged samples, which increases the delay.
To have minimum delay set low (us/div) time base to have 100MHz sampling rate, or change the trigger or leave this at auto and change the trigger channel sampling mode to Decimate.

[Xieo]

@attila

That is very very helpful. ... I will try this today ...