Can the OPI5 be Configured to Run LCNC?

In case he is bashful, royka is definitely one of those people. I've been following his directions to excellent success.

That being said. I haven't worked on installing the wayland drivers, yet. Don't know if I'll bother. Mesa has not stock of the 7c80 or 7c81, so I ended up with a 7i96S. Given the performance I documented, the fact that I'll never be trying to push the system that hard, and I'll be offloading some of the work to the Mesa card, what's the point of faster video?

Anyway, today I'm trying toi cram 10lbs of shinola in to a 5lb box. My system will have two enclosures. One for power, and one for control. Today, the power box is going together.

buzzFab, perfect then it'll probably arrive about the same time as mine.
I've looked at the source of the NanoPC-T6 but it's not clear to me which dtb it needs. It seems to use an UEFI bootloader and Armbian doesn't support it yet either.

If you like you could try to write an official FriendlyElec SD image with Ubuntu Jammy and install the kernel for OPi5+ on it since it has the same ethernet chip. I know it's a lot of information at the beginning but in the end it's not so hard. The kernel we are using is the realtime kernel of Rockchip and isn't updated since august. To support newer devices I made these patches (it basically just adds the dtb sources and some updates) with these patches everyone could build their own rt kernel with the drivers they need.

That look like some pretty nice drivers/motors, There will be plenty of IO pins available indeed with the driver made by Guglielmi.
The OPi5+ does have an eMMC slot, to be safe I would use a m.2 drive in case it has influence on the latency. A heat sink is required indeed, otherwise the chance it big it will throttle when it gets hot which will induce latency. Haven't seen one yet but with a good fan the small ones you could stick on the chips will do. For a pwm fan you could search for "rpi4 pwm fan". 

I installed heatsinks and a 512Gb .m2 card, then transferred the OS and ran the test again overnight.

The results were exactly like the one before, under 4 on the servo thread, except for the ouliers at about 25. I have to leave it running overnight for these to show up, and nothing is printed to the console. Is there a way to get the histogram to print a timestamp when the max or min latency has been exceeded?

I think the evidence is clear at this point.  I've been running this latency histogram for days, trying to determine where the glitch up to 25us was coming from.   I've got the all but perfect results above.  The only thing I haven't done is run any video. 

Is it fair to blame the glitch I got before on This Old Tony?  Maybe I should have been watching tubalcain videos? 

Update:  Didn't even get through the first youtube commercial before ouliers started showing up.

 

Nice, for someone like me who wont ever be watching videos on my milling machine this is good to see. Hopefully the OPi5+ gets similar results, I woke up this morning with an email saying that my OPi5+ shipped and is already at the "departure hub" so it wont be much longer. I have a M.2 SSD but will still need some heatsinks. It would be nice to use something like this new solid state active cooler from Frore Systems. 

AirJet mini   

It looks like it could just fit but wont know until I have the OPi5+ on hand. I might contact them and see about pricing and availability. 

Eric

We have found that playing youtubes in FF give us the worst latency results on any processor. It might not be what you see when running the machine, but it sure gives you a worst case to keep in mind.

I think you nailed it, Bari.
Look at the timestamps on the screenshots. I ran for two days, doing everything imaginable to trip it up, and you get the first screenshot. A few minutes of playing a video in FF. . . and the latency has already expanded by a factor of five. If that don't tell the story, I'm reading the wrong book.

On a different note, I'm getting my control box wired up. I've ended up using a 7i96S, due to availability, and Ill be posting my connection sheet shortly.

The one big issue I'm having is how to get +5V to the OPi5. There only seems to be an option for USB-C input. How do I get power to that from a 10A, DIN mounted supply?

But still very good numbers, nice to see a test of multiple days. 

You can connect the power supply directly to the pins. I had it connected that way and worked perfectly until it got loose while I was busy and connected them both to the 5v pins (40A)  the plastic melted but nothing broke.

Ah! You're talking about the GPIO pins. Right? Pins 2 and 4 are +5 and pins 6 and 25 are GND. (Choosing 25, because the position is distinctive)