DuetWifi Reboot randomly
I got duetWifi Chinese.
And I found that at the slightest heating, SAM reboots.
If the board is cooled, it works stably until it warms up.
At the command M122 issued:
=== Diagnostics ===
RepRapFirmware for Duet 2 WiFi/Ethernet version 2.02(RTOS) running on Duet WiFi 1.02 or later + DueX5
Board ID: 08DGM-9T6BU-FG3S8-6J1DJ-3SD6P-1SKMF
Used output buffers: 3 of 20 (9 max)
=== RTOS ===
Static ram: 25524
Dynamic ram: 99140 of which 0 recycled
Exception stack ram used: 272
Never used ram: 6136
Tasks: NETWORK(ready,124) HEAT(blocked,256) MAIN(running,3268) IDLE(ready,200)
=== Platform ===
Last reset 00:03:27 ago, cause: watchdog
Last software reset time unknown, reason: Hard fault, spinning module Platform, available RAM 5856 bytes (slot 2)
Software reset code 0x4030 HFSR 0x40000000 CFSR 0x00050000 ICSR 0x0441f803 BFAR 0xe000ed38 SP 0x00000034 Task 0x5754454e
Stack: 00000000 004488a1 00448905 004486fd 004486fd 004486fd 004486fd 00442021 004486fd 004486fd 00442419 004486fd 00444d79 00444d89 00444d99 00444da9 00444db9 004486fd 0040ce09 0043df75 004486fd 004486fd 00406f71
Error status: 0
Free file entries: 9
SD card 0 detected, interface speed: 20.0MBytes/sec
SD card longest block write time: 5.9ms, max retries 0
MCU temperature: min 24.5, current 25.8, max 26.0
Supply voltage: min 24.0, current 24.0, max 24.0, under voltage events: 0, over voltage events: 0, power good: yes
Driver 0: standstill, SG min/max not available
Driver 1: standstill, SG min/max not available
Driver 2: standstill, SG min/max not available
Driver 3: standstill, SG min/max not available
Driver 4: standstill, SG min/max not available
Driver 5: standstill, SG min/max not available
Driver 6: standstill, SG min/max not available
Driver 7: standstill, SG min/max not available
Driver 8: standstill, SG min/max not available
Driver 9: standstill, SG min/max not available
Date/time: 2019-01-20 18:12:50
Cache data hit count 755925200
Slowest loop: 15.79ms; fastest: 0.06ms
I2C nak errors 0, send timeouts 0, receive timeouts 0, finishTimeouts 0
=== Move ===
Hiccups: 0, StepErrors: 0, LaErrors: 0, FreeDm: 240, MinFreeDm: 240, MaxWait: 0ms, Underruns: 0, 0
Scheduled moves: 0, completed moves: 0
Bed compensation in use: none
Bed probe heights: 0.000 0.000 0.000 0.000 0.000
=== Heat ===
Bed heaters = 0 -1 -1 -1, chamberHeaters = -1 -1
Heater 0 is on, I-accum = 0.0
=== GCodes ===
Segments left: 0
Stack records: 1 allocated, 0 in use
Movement lock held by null
http is idle in state(s) 0
telnet is idle in state(s) 0
file is idle in state(s) 0
serial is idle 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 ===
Slowest loop: 21.75ms; fastest: 0.00ms
Responder states: HTTP(0) HTTP(0) HTTP(0) HTTP(0) FTP(0) Telnet(0) Telnet(0)
HTTP sessions: 1 of 8
- WiFi -
Network state is running
WiFi module is connected to access point
Failed messages: pending 0, notready 0, noresp 0
WiFi firmware version 1.22
WiFi MAC address cc:50:e3:14:fc:fb
WiFi Vcc 3.30, reset reason Turned on by main processor
WiFi flash size 4194304, free heap 27344
WiFi IP address 192.168.1.6
WiFi signal strength -64dBm, reconnections 0, sleep mode modem
Socket states: 0 0 0 0 0 0 0 0
There is also no connection with the display (panelDue can transmit commands, but does not see the answer) and via USB - the connection is established, but commands can be transmitted from the computer, but there are no answers.
Through HTTP and FTP works (cooled) normally.
Telnet - similarly, commands can be sent, no answers.
Before software update on 2.02 (it was 1.21), there was a connection to USB and it worked on the table stably
Is this a hardware problem or are there software solutions?
Thanks in advance for your help!
- WiFi -
And it is possible to roolback firmware to 1.21 for tests?
A watchdog reboot means the software has stopped executing normally. Watchdog faults are extremely rare on genuine Duets. The fact that it only happens when the board gets warm strongly suggests that it is a hardware issue.
Yes it's possible to regress firmware versions.
I found a hardware problem and now it`s work well.
This is a mystical story.
I have been developing embedded software for over 20 years. But I come across such a situation for the first time.
Originally DuetWifi worked fine on the table via USB. And from an external power source 5V.
I began to connect electronics, periodically checking the correctness of the work. At a certain point, I discovered that the board began to reboot regularly. Less often offline, more often when connected via WiFi, and 100% immediately after sending commands via WebControl.
The PanelDue display also stopped working.
Regardless of the type of power source.
It was possible to establish that the connection via USB and from the display still works, but in one direction - commands can be sent to Duet, but no answers are received.
I installed additional capacitors along the +5 and + 3.3V lines. It did not help.
Helped temperature effects. The cooled board worked stably until it reached room temperature.
Conversely, the board heated to +40 did not start at all, the Diag LED was on constantly.
The reset log on M122 showed a reset on watchdog due to hardfault in random points of memory.
It was like the unstable job of setting quartz due to its marriage or poor soldering.
Heating the board locally, I managed with its good warm-up (t. CPU = 45) to achieve stable operation until it is turned off and cooled down. Then it crashes again.
Establish the true (I hope :)) the reason came out almost by accident.
Before the appearance of failures, I just glued the radiators to the drivers of the stepper motors and to the CPU. During debugging, the radiator on the CPU interfered with me and I deleted it.
After that, everything worked steadily in any conditions.
Just for a anyone else reading this thread, the OP's first sentence was "I got DuetWiFi Chinese".
Beware cheap clones!
O, try to add heatsink on CPU to legal DuetWifi. CPU not from Chine.
@deckingman I'm not going to totally agree with you on this one! It's not generally true. Chinese stuff quality depends on price. But electronic boards are manufactured automatically, and I know first hand how low can be the cost of an expensive piece of equipment (like 300USD manufacturing costs for 4500USD retail price). In the end it is down to the quality check at the end of the production line, combined with the quality of the components used for building those boards (this is a different story!).
If you look carefully, you can find very decent Chinese boards for a fraction of the price of the original, if you are willing to wait for it to be cheaply delivered and ready to handle the customs papers yourself!
Just read again what the fault was - "Before the appearance of failures, I just glued the radiators to the drivers of the stepper motors and to the CPU. During debugging, the radiator on the CPU interfered with me and I deleted it. After that, everything worked steadily in any conditions." The guy simply affected the CPU functionality, probably with some parasitic capacitance from the glue used to install the radiators.
P.S. I have the original DuetEth because I hate waiting too much for delivery, generally, so the price was similar to the Chinese clones with express shipping! Customs papers doesn't scare me, though, and I buy quite a lot of things from AliExpress.
@catalin_ro Well we can agree to differ. I was just going by David's comment quote -
"A watchdog reboot means the software has stopped executing normally. Watchdog faults are extremely rare on genuine Duets. The fact that it only happens when the board gets warm strongly suggests that it is a hardware issue."
That and the fact that the OP found it necessary to stick heat sinks everywhere.
There is of course a case for buying non-genuine hardware and I know that even "genuine" boards and hardware are likely to use parts that originate from the far east. Meanwell power supplies are an example of quite highly rated products that are in fact all made in China.
What does tend to erk me somewhat though, is people who buy non-genuine hardware which is unsupported by the manufacturer or supplier, then ask the makers of the genuine product for that support. I just think it's a bloody cheek. Especially as one of the reasons why a clone might be cheaper is that the supplier or manufacturer does not have to pay the support costs.
If I buy something cheap and it fails, I just write it off and put it down to experience. I don't complain to the makers of the genuine product. But that's just me. I guess us of an older generation have different moral values. End of rant............
@deckingman Well, the OP never mentioned he had to glue the heat sinks. He simply said he glued them!
Clearly, a well placed heat sink would not hurt the Duet when the stepper drivers are pushed to the limit, but as specified in datasheets, the heat sink for those should be on the other side of the PCB. From my experience, a good fan blowing across the board is more than enough, and I use all drivers at 2.4A. Never had a problem with this solution!
Regarding the genuine vs. copies, I'm on the same wave length with you. When I go for cheap Chinese boards, usually those are various power supply modules, manufactured as indicated in the datasheets of the chips used on them. Support from designer has a different philosophy there.
But I can give you a very good example in the DIY/cheap CNC world... if you look at this 800W air-cooled spindle and at this one, you will see that they are totally identical. Even more, one of the Mechatron guys admitted long ago on a forum that this particular model is just a rebranded Chinese one, with no special manufacturing requirements. The "German quality" one costs about 500EUR and the Chinese one you can have for about 100EUR, even with express delivery. I simply call that a ripoff!
All true and I'm not arguing. There are certainly some high priced "rip offs" about.
The other side of the coin is that it's sometimes hard to tell if there is any difference between genuine and clone. The Duet board is an example - hard to tell if the clones use the same copper loading for the tracks which affects the heat sink properties. Other examples are easier to spot. I came across someone who bought a clone Diamond hot end that didn't work. It turned out to be made of brass plated steel rather than brass so wasn't much good for melting filament Still not that easy to spot unless you scratch through the surface though. Then there are clone E3D V6s. Until you measure the diameter of the heat break, it isn't immediately obvious that it can be thicker on a clone nor can you tell at glance what the internal diameter and surface finish are like.
I don't know but I'm guessing that genuine Duet boards are assembled using machines. So presumably are the clones. Therefore, the labour cost in any particular part of the world doesn't really come into it and the assembly cost must be similar. So if boards are being sold at significantly reduced prices, then where do the savings come from if not by using sub optimal and therefore cheaper components and cutting out any testing stage (as well as not offering any after sales support)? I've met all the Duet guys and can confirm that none of them live on a Caribbean Island or own their own private jet so they aen't making rip off profits. Maybe they could drop the price if they were prepared to live in tents and rely on food banks but that's hardly fair.
The other danger though is that the reputation of the genuine product can get tarnished. People will generally say something like "I bought an XYZ and it was crap". But they don't mention that it was a cheap clone and likely to be sub-standard. I have indeed seen comments like that about E3D V6 all metal hot ends when in fact the part in question was a clone and not a genuine item. In fact, the unfortunate guy who bought the steel Diamond hot end said exactly that - "Diamond hot ends are crap". Sooner or later, if cheap clone Duet boards start to flood the market, and those boards start to fail, then people will start posting on various forums and social media sites that Duet boards are crap. Uninitiated readers will take that on board, and that will be sad day.
@deckingman I'm the one that admitted almost a year ago that the "legit" Duet is dirty cheap for what it offers (I compare to a similarly powerful controller and the 5 higher-end/lower current stepper drivers, not the cheap Arduino with Pololu boards), something like 20% of the price with the closest competitor. I do myself a lot of software and I know the rather personal price you end up paying when really care about the end result. And I appreciate that!
I might have been the first WorkBee owner that went the Duet way. I also intend to document a Chinese 6040 conversion to Duet for a friend. It is the least as I can return to the community for the decently priced solution. I know that documenting the 6040 conversion is a double edged sword... those Chinese CNCs are known to be quite decent for many jobs if you spend time to re-build the whole machine (unscrew everything, clean and re-grease parts, put everything back together, while insuring proper alignment) and then replace the electronics. The only things that might be worth keeping are the PSU (if it is a 24V one) and the VFD (if it is a decent one). In the end you might start seeing 6040s on AliExpress delivered with Chinese Duets, but that's life!