I’ve been getting random MCU shutdowns for the past few days and it’s driving me nuts. The most common has been the timer too close error. It happens at random hours into prints. At first I thought it was a cooling issue, per the alarm message, so I put a cooling fan on the MKS Pi - happened again after a few prints. Thought maybe it was because I was running from the PSU (MKS Pi is native 24v), so I reconnected it to the 24v, 2A supply I’ve got - happened again. Thought maybe it was a pwm cycle time issue, removed the pwm_cycle_time commands (put them in to eliminate LED flicker), ran 2 more prints (5 hours total) and now I’m getting another MCU shutdown Move queue overflow. I’m running on the UART pins from the MKS Pi with their EMMC memory module (no SD card) and have been doing so for the last few months without any issue. I’m beginning to think it might be something with the HDMI touchscreen that is attached as it seems like prints tend to not fail as often without the screen being attached, but I’m not really sure at this point. That’s the only thing I can contrive because I’ve been running headless until just recently. I’ve even tried running 2 different versions of their Armbian distro to see if that might help. I’m just at a loss here.
DOH! Didn’t even think of that. Here’s the latest. I’ll get the others in a few. I have to power the unit back up with both the EMMC and SD card I was using to test. klippy (4).zip (1.2 MB)
here’s some more. the majority of the time, i’m getting timer too close, other times, it’s move queue overflow (machine A, let’s call it). On machine B, I got the timer in the past error. I’ve been keeping up with the latest updates, but I’ve also been running screens on the controllers as of late, which is something I hadn’t been doing until recently. Not sure if there’s a correlation or just happenstance, but after 3 shutdowns last week on machine A, I disconnected the screen and it ran again. I’ve since reflashed the EMMC a few times to try and rectify the issue and last night/today flashed an SD to see if it might have been the EMMC doing it, but it still happened. Machine B has been running great until today. The setup is almost identical on both machines, besides having dual Z motors and a direct drive setup on machine A and a slightly different STM32 chip
MKS Pi with 8gb EMMC (Rockchip RK3328)
UART connection with short cables
Elegoo Neptune 3 and 2 printers with ZNP Robin Nano boards (STM32F401 and STM32F407 respectively)
3007 cooling fans and heatsinks on both units
Hosyond 7" HDMI touchscreens
I had chalked it up as a dying board until I got an error on machine B.
Currently running a Bigtreetech CB1/Pi4b with a BTT HDMI5 screen attached with the same print that failed earlier and it’s just about to finish. klippy.log.zip (6.8 MB)
Looking through the logs now, it seems you’ve enabled serial support and set the baud rate to 115200. It seems like you are running out of bandwidth at this lower data rate.
Happens on two machines, one with the 115200 baud, one with the default 250000. I did the 115200 because i had heard that there were issues with uart disconnects with higher bandwidth, so I wanted to test it. Mainly happens with the rk3328 chip on the mks pi with a display attached. Not sure if there’s a correlation. I’ve taken both mks pi out of rotation and switched to btt cb1 with no further issues.
