RevPi Connect 4 + DIO - Strange IO beaviour and crazy LED colors (A1-A5)

SiggiN
Posts: 8
Joined: 02 Nov 2023, 23:49

Re: RevPi Connect 4 + DIO - Strange IO beaviour and crazy LED colors (A1-A5)

Post by SiggiN »

Hey Webber,

back home I did some more testing an noticed that I made a mistake in mapping the IOs. Now everything works fine and I was able to translate it to FUP. The LEDs are still having their party but I will figure that out too.
Thanks alot again!

Siggi
User avatar
Amar
KUNBUS
Posts: 160
Joined: 15 Jul 2019, 12:58

Re: RevPi Connect 4 + DIO - Strange IO beaviour and crazy LED colors (A1-A5)

Post by Amar »

Hello Sigmund,
Thank you for bringing the issue to notice.The strange behaviour for controlling RevPiLED for Connect 4 is due to missing parameter RevPiOutput in I/O mapping of Connect4 which was adapted in PiCtory.
This resulted in writing the RevPiLED at incorrect offset in process image.This will be fixed in upcoming release for Revolution Pi Library for CODESYS next month.
Regards,
Amar
KUNBUS
hwopereis
Posts: 7
Joined: 23 Jan 2024, 15:12

Re: RevPi Connect 4 + DIO - Strange IO beaviour and crazy LED colors (A1-A5)

Post by hwopereis »

I'm experiencing the same problem, with the DIO. Whenever I use direct mapping, it seems that the input values are mixed up or shifted, depending on how many inputs I mapped. Also some pins will switch everything to TRUE instead of just one pin. The mapped pins will read <bad>.
E.g. like this:
Screenshot 2024-03-27 231819.png
I solved this by using the addresses in my GVL as follows:
Screenshot 2024-03-27 231402.png
Screenshot 2024-03-27 231402.png (8.16 KiB) Viewed 13394 times
Both for the inputs and the outputs.

The output with the same physical configuration then looks like this:
Screenshot 2024-03-27 231923.png
The versions I'm using are:
Codesys V3.5 SP20 64 bit.
Bullseye Lite 32 bit.
Codesys control for Raspberry Pi 4.11.0.0 (Problem also existed with Codesys control 4.10.0.0.)
Revolution Pi Library for CODESYS V1.4.1.0 (ID: 7530)

---- EDIT ----

Nevermind, this didn't work after all. In the end had to address the full word values for both the inputs and the outputs.
Post Reply