Please help us test RRF 1.17 and DWC 1.14 release candidates

[dc42]

Hi to all Duet and Duet WiFi users,

I have just released firmware 1.17RC3 for Duet WiFi at https://github.com/dc42/RepRapFirmware/ ... -WiFi/Edge and for Duet 0.6/0.8.5 at https://github.com/dc42/RepRapFirmware/ ... 0.8.5/Edge. As usual with git, to download the files you find there you must follow the link to the file and then press the Raw button.

Chrishamm has made a Duet Web Control 1.14 beta release available and I have included it in the above locations. Please note, on the Duet 0.6/0.85 at least, DWC 1.14 requires RRF 1.17RC2. So please upgrade the main firmware first, then check the version number in DWC to make sure the upgrade was successful, and only then upgrade Duet Web Control.

This is your last chance to report new bugs before I do the 1.17 release. So if you have a working 3D printer running RepRapFirmware and can spare a little time, please try out the new RepRapFirmware and Duet Web Control and let us know of any new problems you find (or that there are no problems). Please visit https://github.com/dc42/RepRapFirmware/ ... /WHATS_NEW and scroll down to and read the upgrade notes before you install the new firmware.

For those of you using the new G29 grid bed compensation feature, Duet Web Control now gives you a visual representation of the height map. Thanks to Chrishamm for his hard work on this. Here is an example which shows quite clearly the Y axis twist on my Ormerod.

[img]https://dl.dropboxusercontent.com/u/193 ... ghtMap.png[/img]

[gestalt73]

Oooh, Fun!

But it looks like I have some frame issues that I didn't even know about. :-/

(EDIT: Thanks to this height map, I dug deeper and discovered that I didn't update the effector plate with 50mm spacing for the magball arms. I just finished installing the upgrades to kosselorange. The carriages were updated to 50mm, and the effector platform was still 48. I knew my frame wasn't that far off. Doh!)

http://i.imgur.com/WYn7xE8.png
[img]http://i.imgur.com/WYn7xE8l.png[/img]

[dc42]

It's odd that your bed doesn't look more circular in the height map. Try a finer grid.

[Micael]

This is my Rostock Max V2 with a onyx rev7, 3mm boro glass, snowflake mounted on an FSR system (FSR positioned between the towers) with heatbed set to 70º and a 125mm radius, done with RC1 not RC3.

Heightmap.csv

(1.48 KiB) Downloaded 235 times

Here are older versions (done with pre RC1), with cold bed and different settings from before:

heightmapCold.csv

(643 Bytes) Downloaded 241 times

And the bed at temperature of the previous test:

heightmapHeated.csv

(639 Bytes) Downloaded 233 times

[gestalt73]

There we go, now we're talking. I knew my frame geometry wasn't that off.

Very nice!

[img]http://i.imgur.com/xds3fC6l.png[/img]

[dc42]

There we go, now we're talking. I knew my frame geometry wasn't that off.

Very nice!

[img]http://i.imgur.com/xds3fC6l.png[/img]

I wish mine was as good as that!

[mhackney]

Not my V3 (it's a D300VS)

[DGBK]

Hi there. I just decided to try out the RRF 1.17rc3 after a long hiatus. After hitting the update button I lost my connection to my duet. I now think I may have to go about recommissioning the entire duet. Is it possible to reinstall everything from my SD card, or do I need to start over? I think it's been close to a year since I did the original commissioning and I completely forgot most of the process.

Update: I fixed the problem. It turns out that my download somehow got cut off. I only had about 1/10 of the proper firmware, so it completely messed up the system. I noticed after successfully flashing 1.17b that it was much larger than 1.17rc3. After downloading a new copy, everything is working.

[DGBK]

Hi there. I finally got everything up and running, but I have an odd bug with my height map. Any clue what could be causing this huge trough?

[dc42]

Were you probing off the edge of the bed on that side?

What type of probe are you using? If it doesn't use the nozzle itself as the probe, did you enter the probe X and Y offsets from the nozzle in the G31 command in config.g?

[DGBK]

I'm on a Rostock max V2 with an FSR setup. Every point seemed to trigger and none of them missed the bed. The probe offsets are set to 0.

I thought maybe I was missing a command prior to homing my system because I got another error about moving the head before homing.

I'll run the test again to find the exact wording.

[DGBK]

I was unable to reproduce the problem. It happened the first three times I ran the grid compensation, but this time it did not. It may have just been as simple as power cycling the printer.

[dc42]

My best guess is that the FSR at that point was triggering really late for some reason.

[mhackney]

David, this may be the "ringing" FSR issue I mentioned to you a few weeks ago. I have a new D300VS delta that I discovered this on. It happens when probing points opposite (or near opposite) an FSR sensor location. You can see it if you watch the JohnSL - the LED or LEDs instantaneously light up on the initial trigger - this is quick. Then a split second later usually 1 LED (and it is at the FSR opposite the probe point) will illuminate and stay illuminated for quite some time, a second or more. This happens when the bed "pivots" when the point is probed and then falls onto the FSR on the other side after the initial probe is complete. I don't know why that secondary trigger takes so long to clear as compared to the initial trigger.

Grid probing exasperates the problem since the probe points are closely spaced and get hit before that ringing FSR is cleared. You can add a delay with the R parameter and really slow things down but that's a Pain and/or slow down probing speed. You can apply a little built-in top force to each FSR to prevent it from lifting.

I've been aware of this issue during normal auto calibration probing - I've seen it on several of my FSR equipped deltas. That led me to rethink the FSR mounting system and I've come up with a better solution that places the FSR contact points outside of the perimeter of the build surface. Ideally you want to completely eliminate the hinge point but that is not practical. This is a reasonable compromise that has been working well and I've not seen the ringing effect at all.

Here's a top view so you can see the overhang.

[DGBK]

That's an interesting observation. I have my FSRs mounted half way between each tower and that point does appear to be exactly across from the FSR mounted between my x and y towers (if I'm reading that graph correctly).

[Polygonhell]

I assume if I add a G29 S1 to the end of my config.g file, it will load the bed height map file at start up?

FWIW I had issues upgrading the web interface, (The firmware uploaded fine on the first try), when uploading the web interface, it appeared to upload the files, but I continued to run 1.11, I ended up pulling the SD Card, and the web files had been uploaded to the sys directory.
I manually deleted the web folder, and manually copied the 1.14 files into place.

Might also be nice to have some indication in the Web interface that bed compensation is in use.

[DGBK]

I assume if I add a G29 S1 to the end of my config.g file, it will load the bed height map file at start up?

FWIW I had issues upgrading the web interface, (The firmware uploaded fine on the first try), when uploading the web interface, it appeared to upload the files, but I continued to run 1.11, I ended up pulling the SD Card, and the web files had been uploaded to the sys directory.
I manually deleted the web folder, and manually copied the 1.14 files into place.

Might also be nice to have some indication in the Web interface that bed compensation is in use.

I had to do the same thing to upload from version 1.11.