LitexCNC: firmware and driver for 5A-75B and 5A-75E

So are you saying that the FPGA board has undefined states at its I/O pins
if they are floated so require the interfacing hardware to have pullups
or pulldowns to define the state after a watchdog bite?

That might explain the issue if the interfacing hardware lacks these.
 

Ofc. I doubt anyone would go about interfacing machine-tools straight into level shifters. Seen worst mind you, so all guesses as to what's happening in some other cases.

Yes. It indeed requires an additional board (1. colorlight fpga board, and 2.proper interfacing board).

RC net, schmitt triggers/buffers, pullups/pulldowns, octocouplers... you know. Usual motion control interfacing board stuff. Nothing fancy or spaceship. $10 worth but regardless, "mandatory" for obvious reasons.

Actually works beyond expectations, quite impressed. Recommend.

ps. with your hostmot2 I believe

PCW wrote:

So are you saying that the FPGA board has undefined states at its I/O pinsif they are floated so require the interfacing hardware to have pullupsor pulldowns to define the state after a watchdog bite?That might explain the issue if the interfacing hardware lacks these. 

 

I sent the header board schematic file
Of course, it needs to be modified in some cases
Also a short video of the problem I'm having
 
github.com/Peter-van-Tol/LiteX-CNC/discu...ssioncomment-9351198
github.com/Peter-van-Tol/LiteX-CNC/discu...8#discussion-6627097
 

I did not consider the pull down resistors for the input and output pins. I actually thought this resistor was intended on the fpga board.Where else should the pin intended for the watchdog be connected?

Mecanix wrote:

No it doesn't, neither ghosted. Works great (200+hrs so far, many parts). The I/O states are initialized and maintained hardware-wise (pullups/pulldowns). That's the machine-tool builder's responsability when designing the interfacing board/pcb.

Hello
Can you please share your board schematic and ini and hal configuration files with us?