Question about Trigger signal with fly by point

Moonyeongbin

First time posting here.
Currently, I am performing a vision inspection application using a trigger signal.
The quality is checked by triggering a vision signal at a specified position with a robot speed of 1000 mm/s.
The program was configured using the TriggIO, TriggL, and TriggJ program commands, but each time a cycle is executed, an error of +/- 0.5mm or more occurs in the position.
This error is out of the customer's specification, and the trigger signal must be precisely controlled within 0.5mm.

If there is a function that can be tested for this, please recommend it.

matti

Hallo,

is your signal send by an ABB Local I/O (DSQC 1030)? Check the request package interval (RPI) of that device and try to decrease.

Moonyeongbin

Since the vision signal output type is NPN, I used DSQC1030 using a relay, but it didn't work, so I'm using an NPN board.

How can I reduce the RPI for my board?

matti

There is no RPI on such device I suppose.

Can you find a correlation between robot speed and deviation (by checking different speeds)?

You could use EquipLag Parameter in that case.

Moonyeongbin

The higher the speed, the greater the deviation.
It is a condition that the robot speed cannot be reduced below 1000mm/s due to cycle time.
I use TriggIO command now. Do you think it's better to use the TriggEquip command?

matti

We use a combination of triggios and TriggLIOs:

VAR triggios TriggerMessen_7{ 14 };

...

TriggLIOs *, VMessen, \TriggData1:=TriggerMessen_7, ZoneMessen, tool0 \Wobj:=BT1;

TriggerMessen contains several trigger-positions.

We have measurement speed just 200 mm/s. According to the distance of the trigger-events,  it was necessary to set and reset i/o every 16 ms. Using a high RPI, controller was missing 50% of the events.

In your case it is more about accuracy, not missing. So EquipLag may help.

You may also try to change the "connection type"  of your ethernet-device in system-parameters.

Moonyeongbin

First of all, thanks for the reply. Let's test it the way you suggested.
and then We also shouldn't be missing out on events. At the same time, the accuracy has to be high, so this part is not solved.
As a rough calculation, signal Set and Reset should operate within about 5ms, so we want to solve this part.

matti

Hallo,

5 ms sounds challenging. RPI might be important: https://scadahacker.com/library/Documents/ICS_Protocols/Schneider - Principles of EtherNetIP Communication.pdf