Duet WiFi firmware bundle 1.15rc3 available for testing
-
I have put a new Duet WiFi firmware bundle at https://www.dropbox.com/sh/dmqh94ylfo9jaxa/AAB9Klpo2Dtbdy_ZFYhPFXJea?dl=0. If you want to try it, please proceed as follows:
-
Install the new DuetWiFiFirmware (1.15-rc3) and check it is working. Note: if you are defining PID parameters in config.g then you will need to double the I parameter and halve the D parameter (or you can use the new auto tune function instead).
-
If you are happy with that, install the new DuetWebServer.bin (1.02). This version requires DuetWiFiFirmware v1.15-rc3, otherwise you won't get the connect message.
-
You may also install the new DuetWebControl.bin. A big thank-you to Chrishamm for this, the improvements in it are all his work.
If you try DuetWiFiFirmware 1.15-rc3, please let me know how well the experimental heater auto tune support works for you.
Here's the change list since 1.15 beta 3:
DuetWiFiFirmware version 1.15-rc3
Implemented experimental PID automatic heating model calibration and PID tuning. See https://duet3d.com/wiki/Tuning_the_heater_temperature_control for details and instructions.
Implemented much better heater safety monitoring, based on a model (first-order-plus-time-delay) of each heater/sensor system, using default models if none have been configured
Implemented TMC2660 driver enable signal to support production Duet WiFi boards
Implemented M38 (thanks Chrishamm)
Added support for the external SD card connector on the Duet WiFi and implemented M21/M22 to support this
Added support for new Duet Web Control functions, in particular the sys file editor (thanks chrishamm)
Increased the PWM resolution
On starting WiFi, display additional info from the wifi module if available, in particular the received signal strength indication
Included the status of the SD card-detect signal in the M122 response
Made the WiFi firmware version available to DuetWebControl
Bug fix: PID parameters entered using M301 had the I parameter doubled and the D parameter halved, and vice versa for PID parameters reported by M301
Bug fix: the object height was sometimes incorrectly extracted from gcode files
Bug fix: slicer comments that included certain characters (e.g. '') would give ajax errors when the file information was returned
Restructured the source code tree
Refactored the TMC2660 driver code, in preparation for future work on standstill current reduction and dynamic microsteppingDuetWiFiServer version 1.02
Report the received signal strength indication in the status message
Allow any file type to be served compressed, not just .js and .css filesDuetWebControl 1.12-RC1
I'll leave it for Chris to tell you about all the new goodies he has implemented!
-
-
Great work! Looking forward to trying this out.
-
I realise this firmware version involved a lot of work (AFAICS all of the PID code has been rewritten which is no easy task), so once more great job, David!
For those who are interested, here the changelist for DWC 1.12-RC1:
Merge of dc42's latest DWC changes (thanks again!)
Refactored the entire JavaScript logic
Modified and extended build script to generate SPIFFS and ZIP packages (not published yet)
Redesigned settings page, introduced new values for Duet WiFi and improved data binding
Changed several settings from milliseconds to seconds
Allow creation of custom macro files
Implemented /sys file editor
Allow download of /gcodes and /sys files
Improved AJAX handling, added support for two new HTTP requests and parsing of new rr_connect values
Auto bed compensation button now visible on small devices and added quick connect button to error notification for small devices
Firmware temperature limit is applied to temperature chart
M117 messages can be automatically discarded
Implemented file info caching to speed up G-code file list (although rr_fileinfo is already way faster thanks to dc42's improvements)
Improved firmware update dialogs
Added Hungarian translations (thanks nhadrian) and extended German ones for new version
Minor bug fixesI haven't compressed the main HTML files yet, but I doubt this factor will be significant if DuetWiFiServer 1.02 is installed. When I upgraded to a newer DuetWiFiServer, I could already significantly improved performance and reliability.
-
Thank you both for your work on the firmware! I'll give it a try in the next day or so.
-
David
Is that a typo in the .bin names above should the DuetWebServer.bin be DuetWiFiServer.bin?
Doug
-
David
Have installed all the stuff from your Dropbox but the DWC still reports as DC1.11
Firmware Name: RepRapFirmware for Duet WiFi
Firmware Version: 1.15-rc3 (2016-08-20)
Web Interface Version: HTML: 1.11b-dc42, JS: 1.11b-dc42Have you by any chance posted the wrong File?
Doug
-
Did you run M997 S2 after you had uploaded DuetWebControl.bin? DWC v1.11 for Duet WiFi doesn't ask if a new web interface shall be installed.
-
Hi,
Thanks David and Chrishamm for your great work. I've installed the files and it's working as expected.it's awesome that we can now edit the config files / create macros directly in the interface
I did some tests with pid tuning but I need to play with it a bit more.
I haven't been able to complete the tuning.. i'll try to decrease the PWM -
David
Yes I did in fact I did it twice
Doug
I have just done it again to double check this Time I made the S a upper case instead of lower case not sure if that will make any difference tho
Yep still the same?
-
Perhaps you already have DuetWebControl.bin on your SD card and your attempt to upload the new one to SD card failed silently?
-
David
Refreshing Chrome did the trick obviously didn't pick it up on the Re-Connect maybe worth a note?
All good now tho
Doug
-
After updating firmware, I can't get the extruder heater to heat up. The PID tuning works, but the web control, paneldue, and M104 all fail to raise the temp. The bed works fine. Any ideas?
-
More info: the extruder heater works if I go back to legacy PID via M301. So there's something in the new mode that's keeping it from heating up.
-
Aaaaand now I just saw the temp slowly creeping up (hit 80C) when it was supposed to be off. Looks from the curve like every few seconds the heaters would turn on for a moment. Don't quite remember the sequence I used to get there, but I'm pretty sure I did M0 and them M307 H1 to check the model parameters (which said "in use:no") and then left it alone for about 5 minutes. An emergency stop shut the heater back off.
Getting some really weird behavior here.
-
PanelDue has stopped talking to the Duet or responding to button touches… huh.
-
Hi Ryan,
Thanks for your reports, which I have been looking into. Progress so far:
1. Fixed incorrect error message strings.
2. Heater coming on when it shouldn't and temperature creeping up: I think it is responding disproportionally to temperature reading noise or occasional bad temperature readings. I haven't seen this myself, but I have identified a possible mechanism for it, and implemented a fix which should stop it.
3. I find that I can set temperatures manually, but a macro that I normally use to set temperatures doesn't work when I invoke it from PanelDue. I am about to look into that.
I hope to have updated firmware later this evening.
-
I've just put a new firmware build (1.15rc4) at https://www.dropbox.com/sh/dmqh94ylfo9jaxa/AAB9Klpo2Dtbdy_ZFYhPFXJea?dl=0. Only DuetWiFiFirmware.bin has changed. This fixes the three issues listed above.
-
Hi David or Chris
First thing i've notice with RC4 is that an edit of config.g uploads on save but doesn't prompt for a reboot.
Same Chrome browser that always prompted btw.
Cheers
Phil -
Yep that's correct, i noticed that as well after I had published RC1. It will be fixed in the final version.
-
Thanks Chris.
@David, I've had two goes at auto tuning using P0.5 and P0.4 with S240 and both times it has cancelled due to exceeding 240…. is there an actual target temperature the tuning is aiming at?
I really don't want to run pwm down to low normally and would expect the tuning to work with the desired pwm (P).For reference my M301 was
;M301 H1 P22.0 I0.2 D180 T0.2 S1 W180 B30
This gave <1c overshoot with the 60W heater and was relatively stable with about 0.8c variation once at temp.Edit PS: with smile: it's frustrating to wait for the hotend to cool down to room temp
