@Notepad
Good to hear! Just to reiterate:
Its usual to have all the software at the same release level to avoid possible (untested) conflicts and
Most plugins are at least least major.minor version specific. In this case it does not matter since its just invoking whatever version of motion you have installed.
Try this. It is "grubby" in that it has this:
"dwcVersion": "3", "sbcDsfVersion": "3",
so it will basically accept any combination of 3.x
@Notepad
I have no idea what the zip file issue is. What happens if you extract my zip file then re-zip it ? It may be the zip mechanism itself.
Regarding the latest error, I believe its telling you that there is a mismatch between the version of RepRapFirmware and Duet Software Framework. Specifically, the Duet Software Framework is not at the correct level.
I'd suggest reinstalling and making sure all the versions match.
The motion plugin simply invokes the motion software without any of the UI dependencies of other types of plugins. In this circumstance, it does not have to be rebuilt i.e. you should be able to unzip, edit plugin.json and rezip.
A couple of things to watch (just in case):
"sbcDsfVersion": "3.5", e.g. to "sbcDsfVersion": ">3.5",Whenever I do this sort of thing, I just use windows (11) compress to zip.
Try this zip (remove .txt from filename)
My system is in being rebuilt at the moment so I cannot test.
@oliof said in Create a virtual temp. sensor based on average of multiple?:
another option would be to wire the sensors in parallel and adjust the reference resistance accordingly (halving it if my memory of Ohm slaw is correct).
If you want to go that route - then maybe 4 thermistors. Two in series and then wire those two in parallel. That way the nominal resistance would be the same as just one.
I do have to ask though: what is your objective (in real day-to-day terms)?
With reasonable stable conditions (i.e. no breeze across the bed) you'd expect the bed to reach a steady-state and more-or-less have a consistent distribution of temperature.
Is the complexity you are considering providing an advantage compared with (say): Manually take the temperature at different parts of the bed and compare to the desired temperature - create a simple lookup table. Use that knowledge (moving forward) to infer the average temperature from the reported temp given by a single sensor.
I took the liberty of adding some additional checks / logging to build-plugin.js. Warns if the plugin src/index.js file is missing. Hopefully helpful:
Figured it out. I omitted src/index.js for the plugin (possibly during a copy / paste) . The build thought it had suceeded but it failed silently.
I'm cannot figure out what's actually going on ...
This call .run("build", {}, ["build"]) (near the start of the build-plugin.js script) does not create files for the plugin in the dist/css/ or dist/js/ directory. This, obviously causes a downstream failure, specifically no zip file being created.
The script goes through the "Building for production" dance BUT does not report any errors.
Anyway to modify the script so that it reports errors ?
When I try to build a plugin with 3.5-dev, I get an error ' ERROR No Webpack chunks have be added to the ZIP archive.`. Any idea why? I can build in 3.4.6 OK.
This is on Win11.
There is no zip file is created.
When I first build the dev environment, npm run serve runs without errors.
After trying to build the plugin: If I try npm run serve again, I get the following error. It seems that the act of trying to build a plugin changes something ...
C:\Users\stuar\Github\DWCv3.5-dev>npm run serve
> duetwebcontrol@3.5.2-rc.1 serve
> vue-cli-service serve
INFO Starting development server...
ERROR AssertionError [ERR_ASSERTION]: Missing entry point (index.js, index.ts, dwc-src/index.js, dwc-src/index.ts, src/index.js, src/index.ts) in plugin DuetLapse3
AssertionError [ERR_ASSERTION]: Missing entry point (index.js, index.ts, dwc-src/index.js, dwc-src/index.ts, src/index.js, src/index.ts) in plugin DuetLapse3
at C:\Users\stuar\Github\DWCv3.5-dev\webpack\lib\auto-imports-plugin.js:54:17
at Hook.eval [as call] (eval at create (C:\Users\stuar\Github\DWCv3.5-dev\node_modules\webpack\node_modules\tapable\lib\HookCodeFactory.js:19:10), <anonymous>:9:1)
at Hook.CALL_DELEGATE [as _call] (C:\Users\stuar\Github\DWCv3.5-dev\node_modules\webpack\node_modules\tapable\lib\Hook.js:14:14)
at createCompiler (C:\Users\stuar\Github\DWCv3.5-dev\node_modules\webpack\lib\webpack.js:86:28)
at create (C:\Users\stuar\Github\DWCv3.5-dev\node_modules\webpack\lib\webpack.js:145:16)
at webpack (C:\Users\stuar\Github\DWCv3.5-dev\node_modules\webpack\lib\webpack.js:173:32)
at f (C:\Users\stuar\Github\DWCv3.5-dev\node_modules\webpack\lib\index.js:73:16)
at serve (C:\Users\stuar\Github\DWCv3.5-dev\node_modules\@vue\cli-service\lib\commands\serve.js:185:22)
at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
Full output from plugin build attempt below
C:\Users\stuar\Github\DWCv3.5-dev>cmd.exe /c npm run build-plugin "../DuetLapse3/pluginv3.5-dev"
> duetwebcontrol@3.5.2-rc.1 build-plugin
> node scripts/build-plugin.js ../DuetLapse3/pluginv3.5-dev
All browser targets in the browserslist configuration have supported ES module.
Therefore we don't build two separate bundles for differential loading.
/ Building for production...[BABEL] Note: The code generator has deoptimised the styling of C:\Users\stuar\Github\DWCv3.5-dev\node_modules\vuetify\dist\vuetify.js as it exceeds the max of 500KB.
- Building for production...
DONE Compiled successfully in 236835ms 1:56:47 PM
File Size Gzipped
dist\js\monaco-editor.54de9d18.js 4748.92 KiB 1058.10 KiB
dist\js\app.fc942cc4.js 3764.19 KiB 816.44 KiB
dist\js\GCodeViewer.2c2d2e4a.js 2699.53 KiB 590.23 KiB
dist\js\HeightMap.0a7868ce.js 1464.97 KiB 316.40 KiB
dist\js\OnScreenKeyboard.08821710.js 106.99 KiB 33.03 KiB
dist\js\InputShaping.3edf6046.js 81.09 KiB 17.29 KiB
dist\js\ObjectModelBrowser.b7c78d60.js 25.50 KiB 6.78 KiB
dist\workbox-6567b62a.js 14.11 KiB 4.93 KiB
dist\js\monaco-json-worker.1cd76b1b.js 2.80 KiB 1.01 KiB
dist\js\monaco-worker.6d781179.js 2.79 KiB 1.00 KiB
dist\service-worker.js 2.16 KiB 1.01 KiB
dist\css\app.1e72278e.css 768.96 KiB 107.11 KiB
dist\css\monaco-editor.bda37737.css 112.91 KiB 18.12 KiB
dist\css\GCodeViewer.5c28e225.css 14.26 KiB 2.96 KiB
dist\css\ObjectModelBrowser.562feac3.css 6.60 KiB 1.24 KiB
dist\css\OnScreenKeyboard.425af7fa.css 3.53 KiB 1.13 KiB
dist\css\HeightMap.6241cfd5.css 0.49 KiB 0.24 KiB
dist\css\InputShaping.bd3578ab.css 0.44 KiB 0.24 KiB
Images and other types of assets omitted.
Build at: 2024-06-10T19:56:48.845Z - Hash: 030dae31db75d394 - Time: 236835ms
DONE Build complete. The dist directory is ready to be deployed.
INFO Check out deployment instructions at https://cli.vuejs.org/guide/deployment.html
ERROR No Webpack chunks have be added to the ZIP archive. Make sure it is imported by src/plugins/index.ts
@droftarts said in manual/Tuning/Tool_changing:
@stuartofmt from the 3.5 release notes https://github.com/Duet3D/RepRapFirmware/wiki/Changelog-RRF-3.x :
- When RRF needs to deploy a Z probe ...
Ian
I'll take another look as soon as I get the chance. This was not the behavior in one of the 3.5 rc's - hence the workaround. I've not checked for a while - will report back as soon as I can.
If you are using a common (i.e. same for each tool) BLTouch or similar: it might be worth checking to see if you also need to add deployprobeN.g and retractprobeN.g
I'm not entirely sure if its fixed / working as intended and may depend on the firmware version you are using.
I had to add them even though they only call deployprobe.g and retractprobe.g
It looks like I have some work to do. I've not dug deeply into it yet but it seems Bookworm does not like pip3 and would prefer / needs to use pipx.
I can check versions of Debian and make the appropriate differences transparent to the install program. I just need a few days to carve out some spare time and then it will mostly be testing.EDIT: Looks like using apt is the prefered way ...
If I'm missing something obvious here, that you have already encountered / resolved, please let me know.
I've created a chatroom to take any implementation discussion off-line. I'm progressing on the basis of providing each python plugin with its own virtual environment.
It looks like I have some work to do. I've not dug deeply into it yet but it seems Bookworm does not like pip3 and would prefer / needs to use pipx.
I can check versions of Debian and make the appropriate differences transparent to the install program. I just need a few days to carve out some spare time and then it will mostly be testing.
EDIT: Looks like using apt is the prefered way ...
If I'm missing something obvious here, that you have already encountered / resolved, please let me know.
Sorry - for some reason I have not been getting notifications about this thread until today.
I will take a look at the installation script to see if it can be improved but ...
Please note that it's all a bit "vague". By that I mean there is no sure-fire way to check versions across all libraries. It depends too much on how the author chose to identify (or not) versions.
Regarding the use of exec. I was being a bit lazy but at the same time its limited to calling version. I'll take a closer look.
After scanning the article ...
I think its a bit premature (at this stage of the research) given (emphasis is mine):
Once our method calculates flow rates in mm/s for these four speed categories, it outputs equivalent linear feed rates in mm/s for a selected track width and layer height. We then manually input these selections into the slicer in order to generate GCode and print the test artifacts. This obvious shortcoming of the method is a primary focus of our future work, as we discuss in section “Integrating with Motion Control and Slicing.”
This is not a dynamically generated set of parameters. More-or-less, once you have the parameters, its unlikely that you would need to change them often.
The same basic parameters can be gleaned with a small number of "calibration" prints.
So while the possibilities are interesting, its unclear if there would be a compelling cost benefit (for the average user). Perhaps this might appeal to the batch oriented commercial user (print a test object - adjust parameters - print a batch).
Now what would be elegant (cost aside) is a closed loop system with feed-forward during print time 
This macro may be overkill, but may help. It will calculate the probe points for you. I use it for quick calibrations (small number of probe points) t0 get the bed more-or-less level then a more comprehensive calibration.
To use you would likely need to have global zprobe_index = 0 at the start (I use a couple of different zprobes hence the variable)
; Creates a mesh calibration
; Assumes the bed limits (actual area reachable by the nozzle)
; have been set with M208
; Assumes the probe offset has been set with G31
;
;--- Prerequisites --------------
; global.zprobe_index MUST be defined. The probe number to use.
;
; --- CHANGE THESE SETTINGS TO YOUR REQUIREMENTS ---
;
; Heat Bed and Nozzle ?
var set_temp = false
;
; Set bed and nozzle temp -- Note that this is not too critical
var bed_temp = 50.0 ; set at an reasonable temp (e.g. 50 should be good enough for PLA and PETG
var nozzle_temp = 150.0 ; set high enough to allow for some expansion but not enough to ooze
;
; Set margin to limit area of bed to be probed should be >= 1.0
var X_margin = 1.0
var Y_margin = 1.0
;
; Set the number of probe points in X and Y axis
var P_X_points = 0
var P_Y_points = 0
;
; ------------------------------------------
; --- SHOULD NOT CHANGE BELOW THIS POINT ---
;-------------------------------------------
;
;Ensure a tool is selected
;
if state.currentTool < 0 ; no tool selected
var msg = "Cannot run if no tool is selected"
M117 {var.msg}
abort {var.msg}
M291 R"How Many Probe points" P"Select a number" K{"9","16","64","100","Cancel"} S4
if input = 0
set var.P_X_points = 3
set var.P_Y_points = 3
elif input = 1
set var.P_X_points = 4
set var.P_Y_points = 4
elif input = 2
set var.P_X_points = 8
set var.P_Y_points = 8
elif input = 3
set var.P_X_points = 10
set var.P_Y_points = 10
else
abort "Selection cancelled"
M291 R"Hot or Cold Calibration" P"Turn on Heaters" K{"YES","NO","Cancel"} S4
if input = 0
set var.set_temp = true
elif input = 1
set var.set_temp = false
else
abort "Selection cancelled"
;
;initialize local variables
var X_grid_min = 0.0
var X_grid_max = 0.0
var Y_grid_min = 0.0
var Y_grid_max = 0.0
; Get the limits for accessible parts of the bed i.e. where nozzle can reach
var X_min = move.axes[0].min
var X_max = move.axes[0].max
var Y_min = move.axes[1].min
var Y_max = move.axes[1].max
; Set the probe offset
var X_offset = sensors.probes[{global.zprobe_index}].offsets[0]
var Y_offset = sensors.probes[{global.zprobe_index}].offsets[1]
;
; ------------ MACRO STARTS HERE-----------------
;
G90 ; use absolute coordinates
M83 ; extruder relative mode
;
if var.set_temp
echo "Wait for Calibration Temperatures to be reached"
M104 S{var.nozzle_temp}
M140 S{var.bed_temp} ; set bed temp a compromise between PLA and PETG
M116 ; wait for temperatures to be reached
;G1 E-4 F3600 ; retract filament quickly so no ooze
M290 R0 S0 ;remove any babysteps
M561 ; reset any bed compensation
;
; Calculate X max and min probe points
if var.X_offset >= 0.0 ; right hand probe point is limited
set var.X_grid_min = {var.X_min + var.X_margin + var.X_offset}
set var.X_grid_max = {var.X_max - var.X_margin}
else
set var.X_grid_min = {var.X_min + min(var.X_margin, -var.X_offset)}
set var.X_grid_max = {var.X_max - var.X_margin + var.X_offset} ; X offset is negative so we add
echo "X offset " ^ var.X_offset ^ " (" ^ var.X_grid_min ^ ":" ^ var.X_grid_max ^ ")"
;
; Calculate Y max and min probe points
if var.Y_offset >= 0.0
set var.Y_grid_min = {var.Y_min + var.Y_margin + var.Y_offset}
set var.Y_grid_max = {var.Y_max - var.Y_margin}
else
set var.Y_grid_min = {var.Y_min + min(var.Y_margin, -var.Y_offset)}
set var.Y_grid_max = {var.Y_max - var.Y_margin + var.Y_offset} ; Y offset is negative so we add
echo "Y offset " ^ var.Y_offset ^ " (" ^ var.Y_grid_min ^ ":" ^ var.Y_grid_max ^ ")"
;
; Set the grid probe points
M557 X{var.X_grid_min,var.X_grid_max} Y{var.Y_grid_min,var.Y_grid_max} P{var.P_X_points,var.P_Y_points}
M557 ;report the settings
;
echo "Performing Mesh Calibration"
G28 ; must be homed first
G30 K{global.zprobe_index} ; Perform a single probe to establish Z=0 (should already be at center because of home)
G29 K{global.zprobe_index} S0 ; perform mesh calibration save and load
G0 Z50 F3000 ; lower bed
echo "Calibration Complete"
if var.set_temp
G1 E4 F300 ; put filament back
; Turn off heaters
M568 S0 R0 A0 ; Reset temps and turn off hotend
M140 S0 R0 ; Reset bed temps
M140 S-273 ; Turn of bed heater
This should clarify:
Screenshot of Object Model from DWC
Relevant setting on config.g (global.zprobe_index = 1)
;Setup BLTouch as probe 1. This is to allow probe zero to be free for analogue sensor e.g. simple piezo
echo "Setting BLTouch"
; K = probe number P= probe type R=Settle time before probing H=Dive Height F=Probe Speed
; T=Speed between probes A=Max probes at each pont S=Repeat deviation
M558 K{global.zprobe_index} P9 C"io2.in" R1 H10 F120 T6000 A5 S-1 ;probe 5 times and average
M950 S{global.zprobe_index} C"io2.out" ; This is the mod probe - used for deploy and retract
G31 K{global.zprobe_index} P25 ; set trigger threshold
M280 P{global.zprobe_index} S160 ; reset any error conditions```My somewhat strange use-case stems from some experiments I was doing with a secondary calibration code for auto Z-offset. Because of some limitations in DSF - I could only use probe 0 for the calibration code and therefore relegated my BLT to probe 1.
When not running an auto-z-offset macro (still a wip): The default state is state.sensors.probes[0] = null and state.sensors.probes[1] is the active probe (and hence has a .value). The object model does indeed have a element at index zero but it reports null.
Aesthetically - the second code attempt is a bit ugly (makes several assumptions) but does work as expected.
you were testing state.sensors.probes[0] and state.sensors.probes[0] which are the elements of the probes array, not arrays themselves
I had assumed (incorectly) that each element of the probes array was itself an array. I should have looked at the object model more closely
@gtj0
declaring as static moved a step closer but "no cigar" 
I tried a bit of debugging. Line 4 (below) gives this error (note there is no probe 0 ). I expected a null or undefined - I suspect deeper in dueui its simply not returning anything on error ??
{timestamp: Fri Mar 29 2024 14:28:00 GMT-0600 (Mountain Daylight Time), severity: 'E', message: "Cannot set properties of null (setting 'value'): `…rs.probes[DueUIConfig.getProbeNum(state)].value}`"}
Test code:
static getProbeNum(state) {
var myprobe = "1"; // set to one for testing. If statement below is failing
console.log("Checking Probe");
console.log (typeof(state.sensors.probes[0].value));
console.log (state.sensors.probes[1].value);
if (Array.isArray(state.sensors.probes[0])) {
console.log("Probe 0 found");
myprobe = 0;
} else if (Array.isArray(state.sensors.probes[1])) {
console.log("Probe 1 found");
myprobe = 1;
}
console.log("Probe Number is - ", myprobe);
return myprobe;
};
I ended up backing away from Array.isArray and just using this. I'm unsure if the use of the global myprobenum to exit the function early on successive calls buys me much ....
static getProbeNum(state) {
var myprobe = null;
// If already know the probe number just return it
if (myprobenum != null) {
return myprobenum;
};
// otherwise Find the active probe
console.log("Checking Probe");
if (state.sensors.probes[0] != null) {
console.log("Probe 0 found");
myprobe = 0;
} else if (state.sensors.probes[1] != null) {
console.log("Probe 1 found");
myprobe = 1;
};
myprobenum = myprobe; // assign to the global variable
return myprobe;
};
I started this conversation in another thread but it warrants being in its own thread so others may search more easily.
I'm trying to execute some javascript inside DueUI. There is an example in the wiki that I'm following but without success.
Note that my javascript is "code by example" - so I do not really know what I'm doing
I'd appreciate any pointers.
The top of my DueUIConfig class looks like this:
class DueUIConfig {
getProbeNum(state) {
var myprobe = "No Probe";
if (Array.isArray(state.sensors.probes[0])) {
myprobe = 0;
} else if (Array.isArray(state.sensors.probes[1])) {
myprobe = 1;
}
console.log("Probe Number is - ", myprobe);
console.log("Probe ", myprobe);
return myprobe;
};
Later on I do this:
"field": "${state.sensors.probes[DueUIConfig.getProbeNum(state)].value}"
The error I get claims that :
{timestamp: Fri Mar 29 2024 11:07:37 GMT-0600 (Mountain Daylight Time), severity: 'E', message: 'DueUIConfig.getProbeNum is not a function: `${stat…rs.probes[DueUIConfig.getProbeNum(state)].value}`'}