Problems after updating firmware
-
@samlogan87 said in Problems after updating firmware:
Software reset code 0x4090
I'm not overly-familiar with error codes, but according to the error codes page I think that means:
memManage = 0x40
and...
moduleStorage = 9
(I think, please someone correct me if I'm wrong!)So I'm guessing SD card, or reading SD card, or SD card socket? Particularly if it's having problems reading the config.g from the card. You say you have recently got another SD card (16GB)? Maybe try formatting a card with the official SD card formatter https://www.sdcard.org/downloads/formatter/index.html
SD Cards should be formatted either FAT16 with 64kb clusters (4GB or below) or FAT32 with 32kb clusters for SD cards up to 32GB, with 512 byte sectors (should be default). SD cards larger than 32GB are not supported, and neither is exFAT format.Ian
-
@droftarts thanks for that.
I had no idea how to read the error codes. Something strange seems to be happening here as it has done it to two sd cards now. I am thinking there is something else happening perhaps. The amount of disconnects I get between dwc and the board has increased greatly since upgrading and I wonder if maybe when I was making changes to my config file if whether one of the disconnects happened and the file didn’t get uploaded correctly. I will give an official formatted a go. I usually just use the windows one.
Hopefully David can find sometime to shed some light on what could be happening as it is getting a tad frustrating now. Luckily when it started doing this last night I opened the config file and copied it so I have a backup before it seemed to have completely crapped itself. Last time I was not so lucky.
Kind Regards
Sam
-
@samlogan87, I am sorry, that software reset code is a known issue in firmware 2.03 when a DueX is present in the system and there is noise on the endstop inputs of the DueX during startup. You are only the second user to report it. There is a temporary 2.04RC2 version at https://www.dropbox.com/sh/mk3jlsoi6gyxd8q/AABbEvX2qb19McbRe6UvRmI9a?dl=0 that includes a fix for it. Or you can revert to firmware 2.02.
The noise on the endstop inputs may also be responsible for the increased rate of disconnections. Firmware 2.03 and 2.04 use a separate high-priority task to update the DueX endstop states recorded by the main processor when they change, in order to reduce endstop latency; but if they change very frequently, that task may use too much CPU time.
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42
I understand you are probably flat out at the moment and can’t answer everyone’s questions so thank you for replying.In a way that is great news as it is just not me going crazy. I will try the upgrade as I am hoping to get a magnetic filament monitor when they are released that I will need to be running 2.04 with won’t I? Is there anything else that I could do to mitigate these issues?
Kind Regards,
Sam
-
Also the only endstop I currently have connect to my duex is a filament run out switch if that is any help.
Kind Regards,
Sam
-
Hi @dc42
A bit of an update. I removed the sd card, formatted it (16gb formatted as fat32 with 64kb unit size), put the sd card files on it and then copied all of the appropriate printer specific files onto it that I backed up. I then updated the firmware to the release in your dropbox.
Unfortunately it has not worked and the same issues still exist. The bed leveling still pauses and when I did try printing something it lagged between straight moves.
Do you have anything else I could try?
Kind Regards,
Sam -
Could you try disconnecting the filament runout switch connected to the Duex, and disable it in config.g? This might at least remove this as a source of the problem.
Ian
-
Well i tried it and it seems to be working well. I am just had a look and when I put the water cooled hotend in I re ran some of the wiring up the the top of the printer and one of the stepper motors for the corexy mechanism runs down the same channel in the extrusion as the filament sensor.
The question is do I need to run shielded stepper motor cables now?
Kind Regards,
Sam -
If you can put some distance between the stepper motor wires from the endstop wires, so they're not running next to each other, it will help.
From what I understand, twisting the pairs of stepper motor cables should help, as will twisting the pairs of endstop cables, as interference is then cancelled out. I tend to braid/plait the stepper motor cables and twist endstop/thermistor cables, but that's more for wire neatness than for noise cancelling, so I'm really not sure if it works for the stepper motors.
It can be a bit of a pain running shielded cables. At RepRapPro we used foil shielded stepper motor wiring, grounded to the metal case of the Duet enclosure, which was grounded to the PSU. This seemed to work well, but for you it means changing all your motor wiring. Might be easier to run twisted pair shielded endstop wires instead first, and see if that's enough, though you'll need to ground the shield to something. I think shielded ethernet cable is the preferred choice, but I haven't had to do it myself, yet. At least you know where the problem is, if it reoccurs!
I found this guide to best wiring practices from LinuxCNC a while ago, thought it was a pretty comprehensive: http://linuxcnc.org/docs/html/integrator/wiring.html
Ian
-
Hi @droftarts
I will see what I can do. I am wondering whether the noise also was going through to my piezo probe as well that is connected into the z probe connector as it would just start randomly probing in mid air. I may just do the end stops as they are easy to change to shielded.
Kind Regards,
Sam -
Ethernet is "preferred" because it's easily available and relatively cheap. But it's also the lowest grade of shield (foil). Make sure to use stranded and not single core in any case (in the Ethernet case that would be patch cable instead of installation cable)
-
@bearer thanks for that. I have some cat6 cable laying around. But as an update, I have managed 3 5 hour prints on the printer without issue so it is definitely that. However the fix @dc42 didn’t seem to work for me. I am waiting on the magnetic filament sensor to be released so I can buy it and then I will do it on shielded cable.
Kind Regards,
SamCustom Core-XY
-
Hi @dc42
So I have had my filament out sensor disabled and disconnected for a few prints now and it has worked fine. Now all of a sudden I am getting probing issues again. The piezo probe seems to probe fine for homing and bed leveling then all of a sudden when it want to re home z after leveling it starts probing in mid air. I press the emergency stop and it will home and level correctly but when it does the last z home, it stops working. Same if I turn power on and off. Do you think this is related to the changes in the new firmware? The z probe is connected to the z probe connector and there are no end stops connected to the duex anymore. Just two stepper motors and one PWM fan control.
Kind Regards,
Sam -
@samlogan87 just to clarify cat6 doesn't say anyhting about shielding, but its usually accompanied by utp or ftp; so cat6 ftp is shielded and cat6 utp is not shielded.
-
@bearer sounds like I am going to have to rip the printer apart and shield everything now which is a pain in the backside. Will proper shielded cat6 cable be able to handle 1.5 amps if I double the pairs up for the steppers? The steppers are probably easier to do as the end stops for my y axis and precision piezo are in a sheath going from my carriage along with my pt100. It is quite a pain to do. Lucky I have a flex3drive and I don’t have a stepper motor on my carriage. I guess what I am however asking is am I going to waste my time doing all this if it is not something else.
Kind Regards,
Sam -
Sadly cat6 doesn't actually say anything about current handling, the rating is mostly concerned with the frequency response of the cable.
I think its commonly available as both 22 and 24 awg. 24 awg is rated for about 2A as a single wire, you might get away with a total of 3A peak on double pairs even in a 4 pair cable, as both coils won't be at full power at the same time(?) I went with flexible rubber 2 pair 22 awg with braided shield when I rewired my printers, a little thicker than 22 awg cat6, but a lot softer and likely to last longer in the cable drag chains.
-
I may just get some shielded cable off Ali express and put that in.
Another though, would ferrite beads work on the endstops to shield noise?
Kind Regards,
Sam
