Topics created by Eumldeuml

@dc42 Sorry, apparently I misread your statement somehow. Thanks for clarifying!

The WiFi version is v1.23 so that seems to be the current one.
I'll try using the new DWC before deciding if I want to go back. The new design has some improvements that I really like 🙂
Also, the .json files are now visible in the systems folder. My guess is that it has to do something with the option of storing the settings on local storage (although I don't know if "local" means on my computer or on the Duet).

Regarding the settings for the scope, I'm not sure which ones you mean.
The step pulses are what you'd expect but the DIR signal is not. On the picture you can see both signals, STEP on top, DIR below. The sampling rate probably isn't high enpugh to catch all the pulses (pulse width is set to 2 µs) but otherwise you wouldn't be able to see the whole spike on the DIR channel.
20200805_115209.jpg

Is there a way of testing that without modyfing the board itself? A resistor in series with one of the connecor wires perhaps?

@Eumldeuml I think I see what you mean. It looks rather like the top layer is more "squashed" on the right than it is on the left. Almost as if the part isn't level so you have a bigger gap between the top of the part and the nozzle on the left than on the right. Try printing it again but rotated 180 degrees on the build plate. If it still looks like gaps on the left, then it's a mechanical issue with your printer, rather than anything related to extrusion or slicer.

Thank you all for your answers! 🙂
I did consider using a macro for that but for me it seems a little bit excessive that one needs to create a macro in order to display the correct object height... I don't like having too many files in the /sys folder because it clutters up the view which gets very annoying on a small tablet that I use to control the printer (and since it's a DIY printer I do need to access the config and toolchange macros quite often).
But anyways, if adding one letter to the end-gcode fixes the problem then I'm happy 😄 Will try that tomorrow!

@fma You are right, that's why I asked the initial question (we closed the loop, so to say :D)
A printer that tilts the bed to tram it is already a 5-axis machine (or more like a 7-axis machine, if you count all the axes of the printer). What's missing is a software, that allows for non-planar printing and is fire-and-forget like a regular slicer. Still a long way to go I guess...

@JamesM I did calibrate the E-Steps quite a while ago and found no deviation from Bindtech's recommended 420 steps/mm. The problem I see there is the tension of the drive gears and the hardness of the filament since there's no way of adjusting it repeatably. More tension means more bite into the filament which reduces the effective diameter. As I print mostly PLA (which has a comparably high hardness), this isn't of much concern I think.
Also the flow rate was calibrated so that the printer produces accurate line thicknesses and overall dimensions so I am very confused how that could result in those artifacts I mentioned...

@EasyTarget Thank you for your detailed answer! 🙂
I also think that those artifacts are in some way related to a bigger nozzle (I changed from 0.4 to 0.6 mm quite a while ago). Since I never cared really much about aesthetics so I considered the faster print time was worth more than the drop in optical quality. But now that my printer is getting better and better I start to dislike the uneven walls (which btw show up way more with 0.2 mm layer height than with 0.3 mm).

For slicing I use the Prusa Slicer. I don't have a 'vanilla' profile for my printer since it's completely designed by myself 😅
The filament I use most is PLA NX2 from extrudr which I print at 215/60°C. I measured the diameter at several points and it is quite consistent (1.75 +-0.01 mm).
I also checked all the set screws on the pulleys and the belt tension, no problems there.

My print bed does indeed move in Z (intentionally of course :P) but it also does wobble the slightest bit at certain print speeds. There might be some hysteresis in the wobbling so that could cause those artifacts on the walls which get amplified by lower layer heights (and presumably the flow rate). This might be the most likely cause but I don't know how to get rid of it since I already beefed up my Z-axis quite a bit (if you're interested check this thread.
Could the problem be a result of interpolation or microsteps on the Z-axis? I'm thinking that the Z-axis might 'fall' to the next full step when commanded somewhere in between. Maybe even a similar effect on the other axes which could lead to lines being printed not exactly on top of each other?

I'll try that! Thanks so much for your support! 🙂

@edgars-batna said in Grounding a frame made of aluminium extrusions?:

I used the plain normal metallic T nuts and corner brackets from the extrusion supplier instead of plastic when assembling and used quite a bit of force. This provided me with connections everywhere basically by accident. The T nuts are rough on the sides and bite into the material when tightened. Another way that I found is to use longer than appropriate screws with T nuts or T-slot nuts so that the screw bites into the back of the slot. Just don't use plastic as it will fatigue and break the connection.

I found the same with the hammer head nuts. For the connections between extrusion pieces I used a small piece of wire to join them. This maybe isn't the best practice since you can't guarantee that the connection will remain over time, but it gave me steady continuity from any two points on the frame I could find, so I called it good.

@eumldeuml started off with a cheap low profile heatsink, so that reduced the milling quite a bit, then "lapped" the top surface with wet/dry sandpaper on a surface plate. should be reasonably flat but looking forward to seeing how it works once mounted and heated.

Thanks for your replies! I will test that as soon as I get my printer up and running again! 🙂

@nitrofreak Regarding the purge/wipe after the nozzle change: In my current design with the tilting extruder I already used a small piece of stainless steel sheet metal which blocks the nozzle so no filament comes out. It worked quite good and it eliminates the need for purging since no filament can ooze out and create an empty space in the melt chamber.
In theory the filament degrades over time in the hot nozzle but I haven't found it to be a problem. Most of the time the infill will print first anyway so a potential degradation doesn't deteriorate the strength very much as the infill doesn't really contribute much to it anyways.
But regarding my initial question: Your extruder is bowden drive, right? I'm still not sure about the need for simultanoeus retraction of nozzle and filament when using a direct drive setup. But I guess I'll have to do a little bit of pioneering here 🤔

@deckingman As I already told you in another thread I am really impressed by your mixing hotend solution 🙂 It's just not the best way for my needs (apart from being waaay over my budget lol) since I don't print with more than one color/material very often. But when I do it's mostly for a combination of a rigid and a soft-ish filament (nylon for example). In the future I also want to experiment with different nozzle sizes in one setup.

OpenSCAD difference is probably your best bet. If OpenSCAD won't load it, run it through one of the online repair tools. Even if you've never used OpenSCAD before... it is free, and this use is extremely easy.

You said you can't post the STL on a forum. Or your not sure. I get that. If you are willing to email it to

danal (dot) estes (at) gmail (dot) com

I'll give it a shot, and send you the OpenSCAD program (as well as the resulting "fill" STL).

Well, we've drifted off just a little bit 😂

In order to add a little bit to the original topic: I managed to get the end stops and the trigger commands to work but I realized that my magnetic latching mechanism is too weak. So tomorrow I'll try to improve that. I also need to design a part cooling duct but the first print with the new Extruder showed that a lack thereof isn't all too bad

That makes sense and eases my mind! Thank you very much for your help! 🙂

Yes, the printer has been homed. It executes the macros since it retracts one filament and unretracts what would be the other filament, just the servo doesn't move.
There are indeed two commands for the servo, one with S-1 to disable it.
I've tried to put a pause in between those lines and now it works! Seems be that the servo was being disabled while it's on the verge of moving...
Kinda embarrassing that the solution is so stupid simple... But thanks a lot anyways! 🙂