You sure everything is ok on your USB cable to your TinyG? Is it too long? Have you tried swapping cables? This feels like a low-level serial port problem. I have never seen an error come back like that on my runs. It could be OS drivers too. I hear the FTDI drivers on Mac have issues sometimes.
The USB cable is only a few feet long and has a ferrite bead on it. Haven’t tried switching it out.
I’m using Win7 64 bit. It’s worth noting that when @Riley_Porter updated the firmware on my TinyG (it’s a version 6 board so had no bootloader) he struggled with it and offered to send me a later board instead, but since I had already made an enclosure for the v6 board I told him I’d much prefer to keep it (the enclosure won’t fit the later boards). I just mention it because it could be an issue with the board itself. ??
It’s not a deal breaker at this juncture because it doesn’t seem to interrupt any of the gcode files I’m sending and doesn’t even happen every time. Maybe if it continues to bother me over time I’ll suck it up and get a later version TinyG and make a new enclosure for it.
Thanks for your help and please don’t think I expect you to be able to fix whatever this weird problem is that I’m having.
Does other software communicate ok with your TinyG? Like a standard serial port console app like CoolTerm? Or TgFX? It would seem I could try to ignore that error and retry, but getting an actual error back from the OS like you’re getting really is rare. I could count up the errors and after 5 retries fail, but if there are errors it would seem data could get corrupted over the serial wire and your job could get messed up. It could be Serial Port JSON Server as well since it is pretty new software so I don’t want to rule that out. I just know I’ve never seen that.
I haven’t spent a lot of time trying CoolTerm or tgFX since I got the firmware updated, but the whole reason I got it updated was because I couldn’t get through full gcode files without it failing in one way or another. Because the failures only seemed to happen when I was running the spindle, I chased down EMI for quite awhile, by installing shielded, grounded cables for the motors, building a conductive enclosure, etc.
It may also be worth noting that my laptop is SSD and the failures seem to occur more frequently if I run json from the D: partition rather than the C: partition, but that may also be because whenever I run it from C: it’s because I want to run it verbose. ???
Could you try running SPJS from another computer and just remotely connecting to it from CP from your laptop? Maybe on old PC? Or Raspi? Or old linux box?