Getting occasional "mcu shutdown - timer too close" shutdowns.
This time it occurred while running "beacon_ratos_calibration bed_temp=85"
When things are running normally, the short term htop load average is like .1 or .3, but when I get the timer too close shutdowns,
the load average is very high.
38 Replies
For What it's worth, I had been struggling with similar errors on my RR VCORE3.1 (https://discord.com/channels/582187371529764864/859890291591217162/1259952416619495484) and some entries in the logs pointed to USB-hub resets. It was suspected to be a USB powersupply issue (I ran a unpowered USB hub connected to Manta M8P and connected USB camera, EBB42 and Beacon of of that). I now have the Beacon directly connected to the Manta M8P and EBB42+webcam on the unpowered HUB and my errors (for now) have disappeared. (I still need to buy a power USB HUB)
Take care about powered hubs. They power the Pi too. That is something you want to prevent, because otherwise you run into issues on power-up. Ask me how I know... 😉
Doesn't the beacon have a warning about connecting it directly to the manta usb?
?.... Please enlighten me? I am not aware of that.
from their website
https://beacon3d.com/product/beacon/
If you search in this discord it was discussed, and I think the very first version was ok, but 1.1+ had the issue where the usb isn't properly grounded?
Thanx, @blacksmithforlife 🇺🇸 ! Not sure if this was already on Beacons website when I purchased the unit or I missed it... Al swap the USB connection so to connect Beacon to the Hub... Hope that that doesnt bring my errors back though...
Does anybody else get these timer too close errors?
I got some once in a while... I just toggle the power and it´s gone. 3 times if I remember correctly over the last 2 months.
I haven't had any errors/shutdowns in the last two days, but before that, I got several a day! I was doing more calibrations, PID tuning, shaper graphs, etc., and now I'm just printing. I don't understand how the Raspberry Pi load numbers can go from .1 and .2, then shoot to almost 6, but maybe the tuning is more demanding than I realize. It would suck to be several hours into a print when it happens, so I'd like to find out what's causing it.
It will not happen during printing.
I think I fixed it by reordering the USB cables on my setup. Beacon and toolboard are now using the USB3.0 connectors, Board is connected to USB2 same as my USB hub. There I just connected my cam and screen. Since then I had no such issues. Take it as a BSOD on M$. It happens, but you would not be able to localize this issue yourself, even experts don´t.
timer too close is always related to system load and doesn't actually have anything to do with USB. The connection is fine, but the system isn't fast enough to send the messages in the timeframe the other end is expecting.
Data congestion could cause it i guess, but you would have to wire your stuff in such a way that all devices would be running over the same low-speed usb connection. USB 3.0 connector on the pi to an MTT USB 3.0 hub with all the devices connected, takes care of that issue (but i'm fairly sure this isn't it - it's not something you do by accident)
Good point... I was thinking about a BIG7 7-port MTT USB hub.
On the other side: For now I don´t get this issue since I reordered my USB on the Pi.
Main issue is: To get it stored in the electronic box on the back side of the VC4... May be there is enough place in the box to get it mounted under the Pi... I don´t know tbh. Time will tell.
Read this thread at Superslicer Discord. Not sure if it is related: https://discord.com/channels/771316156203270154/858997549491027978/1277588712070316104
@miklschmidt are you aware of this issue in SS?
...read the thread and mostlikely not related as it deals with too many request for fan speed changes in the slicer (Orca in this case). Apologies for allerting you...
That's cool. I appreciate people pondering the issue. I have started to notice that most of my errors/shutdowns are during the startup phase when the bed is homing and mesh leveling, and once it gets past that, it seems to finish the print. Thanks.