CONSTANT AJAX disconnect errors

[KeeganB]

The last two failures both happened within minutes of starting a print. I didn't get the diags for the first failure, but interestingly enough it came back online by itself and worked fine until a few seconds after starting the second print. Here is the M122 after the second print and subsequent AJAX failure. I suspect this will come back online again too, sometime in the middle of the print.

M122
SENDING:M122
=== Diagnostics ===
Used output buffers: 5 of 32 (13 max)
=== Platform ===
RepRapFirmware for Duet WiFi version 1.19.2 running on Duet WiFi 1.0
Board ID: 08DDM-9FAM2-LW4SD-6JTF0-3S86N-TLWZX
Static ram used: 21176
Dynamic ram used: 96280
Recycled dynamic ram: 1328
Stack ram used: 4048 current, 9104 maximum
Never used ram: 3184
Last reset 04:50:44 ago, cause: power up
Last software reset reason: User, spinning module GCodes, available RAM 4348 bytes (slot 0)
Software reset code 0x0003, HFSR 0x00000000, CFSR 0x00000000, ICSR 0x00400000, BFAR 0xe000ed38, SP 0xffffffff
Error status: 8
[ERROR] Error status: 8

Free file entries: 9
SD card 0 detected, interface speed: 20.0MBytes/sec
SD card longest block write time: 0.0ms
MCU temperature: min 35.7, current 35.9, max 36.2
Supply voltage: min 11.7, current 11.8, max 12.1, under voltage events: 0, over voltage events: 0
Driver 0: stalled
Driver 1: stalled
Driver 2: stalled
Driver 3: stalled
Driver 4: standstill
Date/time: 2017-09-16 20:04:42
Slowest main loop (seconds): 0.003906; fastest: 0.000000
=== Move ===
MaxReps: 4, StepErrors: 0, FreeDm: 221, MinFreeDm 189, MaxWait: 2ms, Underruns: 0, 0
Scheduled moves: 2311, completed moves: 2306
Bed compensation in use: none
Bed probe heights: -0.019 -0.083 -0.078 0.096 -0.127
=== Heat ===
Bed heater = 0, chamber heater = -1
Heater 0 is on, I-accum = 0.3
Heater 1 is on, I-accum = 0.4
=== GCodes ===
Segments left: 1
Stack records: 2 allocated, 0 in use
Movement lock held by file
http is idle in state(s) 0
telnet is idle in state(s) 0
file is doing "G1 X25.168 Y2.626 E93.1502" in state(s) 0
serial is ready with "M122" in state(s) 0
aux is idle in state(s) 0
daemon is idle in state(s) 0
queue is idle in state(s) 0
autopause is idle in state(s) 0
Code queue is empty.
Network state is running
WiFi module is connected to access point
WiFi firmware version 1.19.2
WiFi MAC address 5c:cf:7f:ef:51:6f
WiFi Vcc 3.11, reset reason Turned on by main processor
WiFi flash size 4194304, free heap 39392
WiFi IP address 10.0.1.38
WiFi signal strength -46dBm
Reconnections 1
HTTP sessions: 2 of 8
Socket states: 0 0 0 0 0 0 0 0
Responder states: HTTP(0) HTTP(0) HTTP(0) HTTP(0) FTP(0) Telnet(0)

[dc42]

Thanks. I can't see anything wrong in the M122 report. it's interesting that you have a a reconnection count of 1 indicating that the WiFi module became disconnected from the router and successfully connected again, even though you have a very high signal strength.

[KeeganB]

David,

I think I have established a bit of a pattern in my situation. It seems that right at the beginning of every new print, it disconnects. Then sometime after that (I estimate between 15-30mins) it just reconnects itself and is fine from then until I start the next print.

[dc42]

I did a little investigation this morning, and I found that for homing moves (i.e. G1 S1 moves), the maximum step rate possible is many times lower than for ordinary moves. So a high microstepping value can cause MaxReps to get very high during a homing move even if the combination of M350 microstepping and M203 maximum speeds is OK for normal moves. So if your print files have a G28 homing command at the start, that could be why you get a disconnection at the start of a print. The workaround for this particular issue is to use slower homing moves.

[KeeganB]

I will try slower homing moves. I don't have a separate G28 at the start of my prints, but there is on inside of beg.g that runs (as G32) at the start of all prints.

My micro stepping is 16x

I should add though, that it doesn't necessarily disconnect during the homing move, sometimes it can be a few minutes into the print.

[dc42]

I would not expect you to have this problem with only x16 microstepping.

[Jarery]

Similar to KeeganB, my board is set to 16x micro stepping (with interpolation).
I also do not have a G28 homing in any of my Gcode but as KeeganB states in my Bed command is a G28 at the start. After that the Gcode has no homing until finished.

My disconnects happen early in a print but unlike KeeganB, mine has never reconnected on its own that I'm aware of.
I have never seen the reconnect count above 0.
Edit: mine don't always happen only near start of a print, just that they are a high chance at start. Mine also disconnects just sitting idle sometime.

I'm about a week away from commissioning a second delta with another duet wifi, based on dc42's large Kossel build. So in a week or so I'll be able to tell if both boards exhibit the same behaviour in my wifi network or if its specific to the one board.

[lolorc]

Hi there.

I also have the constant disconnections.
the disconnections also happen when the printer is idle…

I also have a tons of issue uploading g-code files through wifi. (when usb is plugged)

[BlackTom]

I have had the same kind of issue with 1.19.2, 1.19 is correct.

I had issues when I was running my bed compensation routine and I had it too when I was printing,
I always finish by being disconnected of the web interface when I print and I cannot reconnect to it.

The print continue without issue until it's done so it doesn't affect it

The only way to get back access to the web interface is by powering off / on the board,

[lolorc]

@bloood3d:

The only way to get back access to the web interface is by powering off / on the board,

same here.

[bot]

What about when you do a "hard refresh" of the web page through the web browser? On chrome this is done by holding the CTRL key while clicking refresh. In the past, this has helped me get back into a web interface session.

[KeeganB]

bot,

That hasn't worked for me. I have also tried loading the page on difference devices.

[RGN01]

I'm commissioning my new Duet WiFi and have the same issue. I've tried it in two locations (both no more than 10 feet from the WiFi router) and with three clients (desktop PC, Android tablet and Android mobile phone all running Chrome). The desktop PC is best but it still drops the connection frequently - perhaps every 15 minutes - despite the printer not even being fully commissioned (the most it has done is home the towers and heatbed auto set up) and the PC is being used for browsing and email.

I hope that this is resolved as it is a bit concerning to someone new to the Duet (which I think is absolutely great!)

[dc42]

Have you tried increasing the number of retries in the DWC settings?

[RGN01]

Thanks, David.

I increased it to 5 earlier this evening and it seems to be OK so far on the desktop PC (about 90 minutes) - but the printer and PC are effectively idle.

It was set to one initially, is there a recommended larger setting, or is it trial and error?

I'll monitor this as I start adding and testing functionality to the printer - and report back.

[dc42]

1 is rather low, I think 3 would be a better default.

[KeeganB]

At this point I can confidently say mine fits the pattern I mentioned earlier. Every print it starts fine and after a few minutes it disconnects. After probably 30 mins or so, it reconnects without issue and stays connected until I start the next new print.

David, I am considering downgrading to 1.19. Is there any reason not to?

[Jarery]

Interesting yours reconnects. Mine never has, retries I've changed to 5, 10, to no effect.

[KeeganB]

Jarery,

It may just be that the webpage is auto reloaded by the browser when I come back to the computer and pull it up.

[RGN01]

Thanks, David. I've dropped it to 3 and am testing that. SO far, so good!