LinuxCNC-RIO - RealtimeIO for LinuxCNC based on FPGA (ICE40 / ECP5)

I’d guess that with the Mesa card, the voltage is slightly higher than that of your level shifter, which produces a more stable pulse.
It must be right on the limit, as some pulses get through and some don’t; even 0.2 volts could make the difference.
All in all, I’d recommend not using the drivers – not with Mesa, anyway – as the threshold is too close to the limit.
Or you’ll need to switch to 12/24V so that the driver is triggered reliably.

If you have a variable power supply, you could try increasing the voltage to 5.5 V and see if that helps; the level shifter should be able to cope with that.

DIR/PULSE issue resolved, thank you!! (2026-08-19)
Using a Mesa7i92 + 74hc245 (5V) + dm542 (unmodified dir input pc817), the zero-point return is correct and without error. This is why I raised the issue of the Riocore + Tang Nano9k + 74hc245 (5V) + dm542 (unmodified dir input pc817) failing to return to zero.

I believe you; I understand your problem too.
You could try measuring the voltages at the DIR pin (which can be held at a constant level) and comparing them.
Better still, use an oscilloscope to see how steep the edges are.
Otherwise, I’m out of ideas.

meister wrote:

I believe you; I understand your problem too.
You could try measuring the voltages at the DIR pin (which can be held at a constant level) and comparing them.
Better still, use an oscilloscope to see how steep the edges are.
Otherwise, I’m out of ideas.

 
Meister, thank you for your enthusiastic response! 

 

tuxcnc wrote:

viewsat wrote:

I'm showing four sets of drivers in the photos, but only this inexpensive DM542 has a problem
A long time ago, 3xTB6560 drivers were popular and I had them too.
There, PC817 supported not only dir inputs, but also step inputs. There were problems with this, but the chips were in sockets and it was easy to replace them with selected ones and it worked somehow.
So there are only two options:
1. Your driver is faulty.
2. This type of driver is poorly designed.
Either way, the cause is not to be found in riocore.

1.This DM542 was taken from my CNC engraving machine, which I'm currently using, for testing.
2.Although this DM542 is very cheap, I have used about 1,000 units, and its failure rate is very low, less than 1%. Most of the problems can be found as soon as they are powered on for testing.

I have a Riocore + ICEBreakerV1.0e + 26LS31 (pulse, dir) control board. I tested connecting it to an unmodified DM542 (dir / pc817), but it still cannot return to the zero position correctly.There shouldn't be any problem connecting this board to the motor driver using the 26C31 differential output. Adjusting the dirhold value had no effect.  I suspect it might be caused by a deviation in the dir/step during the forward and reverse rotation of the riocore.

 ICEBreakerV1.0e board
 

26C31 Differential Output Driver
 

FPGA
 
 

Hey, that’s a bit mean – I want a board like that too

Yes, that really is strange, but I can’t explain it at the moment and I’d be glad if I knew the solution.

All I can do here is take some more measurements; I’ll connect it to a logic analyser again and record the pulses for a bit longer to check them.
But on that board too, please measure the voltage arriving at the driver.

That board looks really good

Looks ok for me

Why doesn't anyone tell me that there's been a Rio Board in China for three years now?
www.ruten.com.tw/item/22418437051883/

cool

and a video: www.bilibili.com/video/BV15nG96kEQt/

If this is a drive level issue, note that a 26C31 does not swing to 5V
and its output level is about 2 diode drops below 5V (3.6V) with a 5 mA load.