Issues after 7 days of testing

Hi,

I wanted to check if anyone else has similiar problems.

1) If a print job is paused by the web interface, sometimes this wierd thing happens. The extruder will retract more than 500 mm of filament and keep going until I hit the emergency stop. In some cases it work correctly and retracts some mm and stops. Firmware issue ?

2) Home All button, how to get this to just home to the endstops and stay there ? The current Home All is dangerous as it sends the effector crashing down to 4 mm Z at high speed. Had a near crash when I paused a print job then cancelled and needed to retract the extruder and back to x0 y0

3) Extruder motor starts to get warm and the screws that tighten it via the printed part gets loose.

4) Duet CPU and PCB gets pretty warm after some hours operation. I used the bed sensor to hook up a temperature probe and it can get 40 deg at the bottom, suggest adding a 19V fan below

5) Screws that lock the the smooth rods ends tend to slip, these push at the 3mm acrylic sheets causing them to bow.

6) Quality of prints are quite bad at high print speeds. Due to the bed springs being rather week, the bed can move during printing if excessive material blobs on prints causing layer offsets.

7) Wiring on the back panel needs to be shifted else the carriage will hit on it.

Any comments appreciated.

Cheers,

Will

Thanks fo sharing.

1. It sounds like you have your slicer configured to generate absolute extrusion coordinates. Use relative extruder coordinates instead. If that's too difficult, add a G91 command at the start of sys/pause.g and sys/resume.g.

2. Edit the sys/homedelta.g file on the SD card and remove the G1 movement command at the end.

3. The Fisher extrudsr is direct drive so the motor current has to be set quite high. Using a geared extruder would allow you to use a lower motor current.

4. I think any printer that mounts the electronics horizontally and in a confined space should probably use an electronics cooling fan. Although if you use a geared extruder and reduce the extruder motor current, then the Duet will generate less heat.

5, 7 I can't really comment on as I don't have a Fisher.

6. I think someone is planning to install one of my mini IR sensor boards to use instead of the bed switches. That would allow the use of stronger springs, or no springs at all.

HTH David

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
DC42
6. I think someone is planning to install one of my mini IR sensor boards to use instead of the bed switches. That would allow the use of stronger springs, or no springs at all.

David
What is the price and availability of your mini IR sensor boards. I am presuming these are compatable with Build Tak
Regards
John

Quote
johneato

Quote
DC42
6. I think someone is planning to install one of my mini IR sensor boards to use instead of the bed switches. That would allow the use of stronger springs, or no springs at all.

David
What is the price and availability of your mini IR sensor boards. I am presuming these are compatable with Build Tak
Regards
John

Prices are at [forums.reprap.org]. I test them on BuildTak as well as on glass. I don't know whether there is room to mount one on a Fisher print head though. The board is 24mm wide x 18mm high x about 10mm deep.

Btw E3D has chosen this board as the Z probe sensor upgrade for their BigBox 3D printer, if they meet their £200K stretch goal.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

The problem is that the Fisher nozzle is about 17-18mm below the effector so not enough clearance for the IR board underneath.

I am considering adapting the nozzles part to carry a board beside the effector but it would be quite a distance from the nozzle and effector tilt will become a factor. Now if you could shave a mm or two from the board height or the nozzle could be mounted a mm or two lower we could get somewhere ...

I am not completely happy with tweaking the Z-dive for each measurement point as is currently proposed for fisher since it relies on the baseplate going back in the exact position after removal.

Tweakage ensues ...

Quote
dc42
Thanks fo sharing.

1. It sounds like you have your slicer configured to generate absolute extrusion coordinates. Use relative extruder coordinates instead. If that's too difficult, add a G91 command at the start of sys/pause.g and sys/resume.g.

2. Edit the sys/homedelta.g file on the SD card and remove the G1 movement command at the end.

3. The Fisher extrudsr is direct drive so the motor current has to be set quite high. Using a geared extruder would allow you to use a lower motor current.

4. I think any printer that mounts the electronics horizontally and in a confined space should probably use an electronics cooling fan. Although if you use a geared extruder and reduce the extruder motor current, then the Duet will generate less heat.

5, 7 I can't really comment on as I don't have a Fisher.

6. I think someone is planning to install one of my mini IR sensor boards to use instead of the bed switches. That would allow the use of stronger springs, or no springs at all.


HTH David

Yeah, I am planning on making a new effector plate setup and put the mini IR. Still working on the effector design so I can have the IR board close as possible. First Duet board setup and delta so I will see how it goes. Also replace the bottom plate with aluminum so I can put a heat bed. Removing the bed springs so that make it just use the mini IR once I get it.

I working on a design so it will fit on a new effector design. Congrats on the E3D deal. I might have to get more IR boards off of you later down the road. I like the Duet setup and much easier to deal with than individual stepper boards and then another controller board it get messy.

Quote
pjr
The problem is that the Fisher nozzle is about 17-18mm below the effector so not enough clearance for the IR board underneath.

That's a shame. With care I think it would be possible to file off 1mm from the top of the board, but that would still make the board almost 17mm tall. I guess next time I modify the PCB I'll try to get it down to 16mm high.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I don't have my calibration tools on hand what is the thickness of the effector assembly on the fisher.

Quote
johnmayo
7) Wiring on the back panel needs to be shifted else the carriage will hit on it.

Completely agree. Would prefer at least a couple of the cables moving to another pillar. I had to do a lot of fiddling to stop the carriage hiring the cables. Moving a few cables would eliminate this issue and stop it becoming an issue when the cables shift during use.

I would also prefer the cable entry hole in the base plate to be a bit bigger to make it easier to feed the connectors through.

At the end, I think the simplest solution is to route the cables OUTSIDE the printer. You shall drill the panels for re-entry, but this is not a problem if using sharp bits. If found unelegant, electrical rails could be used to hide the cables.

As others, I got many problems with the cable routing and I used nearly half of the supplied zip-ties routing and rerouting. It shall be noted that following an initial mistake on base plate position, my effector cable is on another pillar than the switch cables, so my situation is supposedly better than others. Also, I added complementary attach for the effector cable to make a 'S' around the panel opening and limit conflicts.

As I've proposed early, I think the printer design shall be modified to offset the panels by 3mm, which also will allow to enlarge the pillars as they could clear the carriage.

Edited 2 time(s). Last edit at 09/08/2015 05:46AM by PRZ.

Regarding point 1, can anyone explain why the E coordinate should be relative while all the others are absolute?
The answer ('if that is too difficult') is misguided passive aggression and helps no-one. On top of that it does not work: G91 only affect the XYZ coordinates, not E. (try it for yourself). M82 and M83 set the extruder to absolute or relative. Unfortunately, M82 seems to reset E to zero, so that is no help at all.
Some further digging in the gcode wiki found this: [reprap.org] and [reprap.org].
That seems to work (details moved to separate post)

Quote
Jelle
Regarding point 1, can anyone explain why the E coordinate should be relative while all the others are absolute?

Because it makes more sense, and avoids the rounding errors you get with absolute E coordinates. However, if for some reason you can't use relative E coordinates (e.g. because you are using Cura to slice models), add a M83 command at the start of the pause.g file. I updated all the pause.g files in my repo to include this command some time ago.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

As I said, that does not work. M83 does as expected, but M82 then resets the E value to zero. M120 + M83 followed by a M121 at the end of the file does work.

Quote
Jelle
As I said, that does not work. M83 does as expected, but M82 then resets the E value to zero. M120 + M83 followed by a M121 at the end of the file does work.

I said use M83. I did not say anything about using M82. You don't need M120/M121 because the push/pop happens anyway.

Edited 1 time(s). Last edit at 09/09/2015 05:53AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].