D-Bot 3D Printer

[tekkydave]

19 minutes ago, Gina said:

Just thought of where the error may be coming from, Dave.  You have a thick aluminium plate with glass plate on top.  I guess if the ali plate is not perfectly flat it could bend the glass to its profile.  Why not have the heater pad stuck directly to the glass?

Which error? Since reinstating the 4-point bed adjustment everything is working perfectly.

I don't like the idea of sticking the heater directly to glass. I have heard a few horror stories of shattered glass that way.

[Gina]

5 minutes ago, tekkydave said:

No reason why not. I just like everything being automatic - that's just me really

But it isn't.  You're manually adjusting the corner bed levellers.  I too like everything automatic if possible but I'm not averse to adjusting once if the adjustment stays put.  It's having to continually adjust things that I object to.

[Gina]

8 minutes ago, tekkydave said:

No reason why not. I just like everything being automatic - that's just me really

But it isn't.  You're manually adjusting the corner bed levellers.  I too like everything automatic if possible but I'm not averse to adjusting once if the adjustment stays put.  It's having to continually adjust things that I object to.

[tekkydave]

Just now, Gina said:

But it isn't.  You're manually adjusting the corner bed levellers.  I too like everything automatic if possible but I'm not averse to adjusting once if the adjustment stays put.  It's having to continually adjust things that I object to.

Once the 4 corners are set they hardly ever need touching. After that the levelling of the combined z axis frame & bed assembly is done by the automated leadscrew calibration.

The G29 Mesh Compensation is just a check that all is level and to iron out any minor variations. It probably could be taken out but I like to know what's going on

 

[Gina]

8 minutes ago, tekkydave said:

Which error? Since reinstating the 4-point bed adjustment everything is working perfectly.

The non-flatness of the glass having levelled the bed at three points.

9 minutes ago, tekkydave said:

I don't like the idea of sticking the heater directly to glass. I have heard a few horror stories of shattered glass that way.

I've had the bed heater directly against the glass on my Titan printer for quite some time and on the Mini too.  Both cases have borosilicate glass though which is supposed to stand temperature variations better than float glass, though I'm by no means sure this is true.

[tekkydave]

1 minute ago, Gina said:

The non-flatness of the glass having levelled the bed at three points.

That was when I tried the 3-point levelling. It was due to the heated bed itself not being stable and dipping in the unsupported corners. Now I have 4-point levelling back it is pretty level.

[tekkydave]

Now it is levelled there is no manual intervention unless G29 shows anything up. If I have been fiddling with anything and the z motors have got out of sync it will automatically set them correct when I turn on.

[Dr_Ju_ju]

1 hour ago, tekkydave said:

Do you have any pics? Sounds interesting.

Just a couple of quick ones, as the room is a tip & I'm currently doing some plumbing work....

[tekkydave]

I did try that method as mentioned before but I still found the bed would rock about a line between the two leadscrews. Even with 4 sets of wheels. Do you get any of that?

[Dr_Ju_ju]

nope, before I connected the motors, I spent some time (2~3 hours) fine tuning the inner frame so that it runs silky smooth vertically, with no twists, rocking etc. And only then did I connected the motors & fettled some more to ensure they introduced no errors, jumps, etc.

[tekkydave]

This is what I have now:

The 4020 frame is levelled (automatically) by the 3 leadscrews, but the bed is also set level wrt that by the 4 corner adjusters.

[tekkydave]

5 minutes ago, Dr_Ju_ju said:

nope, before I connected the motors, I spent some time (2~3 hours) fine tuning the inner frame so that it runs silky smooth vertically, with no twists, rocking etc. And only then did I connected the motors & fettled some more to ensure they introduced no errors, jumps, etc.

You've obviously had better luck than me with it. I got very frustrated with it which is why I went down the 3 leadscrew route. 

I notice your 3-wheeled z-axis wheels are the opposite way round to mine. I have the 2 wheels on the outside and single wheel on the inside of the frame. Not sure if it makes any difference.

Edited March 24, 2019 by tekkydave

[Gina]

The Z carriages on my Concorde use 4 wheels - I didn't think 3 would be sufficient.  In fact all the carriages use 4 wheels.

[Dr_Ju_ju]

I tried 4 wheels per corner, but found they were difficult to get 'even' contact, & for me having the 3rd wheel on the outside, which is also the adjustable\tensioner wheel, made it to set up.

[tekkydave]

Ok, these updates are getting more infrequent. It's not that I'm not doing stuff, just that it's probably not newsworthy.

The last few months have been pretty busy as we have had our kitchen done. I'm hoping to spend more time on projects now but we'll see what happens there

The only major change I have made to the printer, apart from keeping the Duet software up to date is a complete switch from 3mm to 1.75mm filament.

The main issue with 3mm was getting it to load into the Titan Aero easily. The filament path in the Titan is quite 'tight' and the filament often needs to be bent against its natural curl. No problem with PETG as it is naturally more pliable than PLA. But some PLA filament is definitely brittle and I often ended up taking the Titan front cover off to clear out the tiny bits after it had shattered (the filament not the Titan  ).

The 1.75mm PLA and PETG I have now loads very easily so no more swearing during colour/material changes.

One thing I have discovered is that I need to reduce the hot end temperature for 1.75mm filament. For example I have some 3mm and 1.75mm REAL PETG. The 3mm prints best at 250C but the 1.75mm at 240C. Maybe the 1.75mm can heat up quicker in a narrower melt zone or it could be any number of things:

• The 1.75mm is physically different to the 3mm filament despite being the same Supplier & Brand.
• The hotend thermistor was replaced at the same time so maybe reporting a different temperature.
• Something else I haven't considered.

I wonder if anyone else has noticed this effect or is it just a quirk of the filament.

I've also decided to learn a different CAD package. I have been happy using Sketchup for the last 5 years despite the bad press it gets. It has served me well personally and I will still need to use it from time to time for editing existing models.

I don't use a Windows PC which restricts me to the older 2016 version. I run it in a Virtualbox VM and the 2017 version refuses to play ball. Also they will not be releasing any more free versions apart from the 'Sketchup Free' which is web based. I have tried it and it is slooooooooooooooooooooooooooooooow....

I tried Fusion 360 but it also only runs on Windoze so was much slower than SU in the same VM.

So I have started learning FreeCad which runs on the Ubuntu Linux I use on my main PC. It works a bit differently to SU but after following an online tutorial and playing around for a couple of hours it is quick to pick up. It is also truly FREE being open source. It seems to have more modelling features than SU and doesn't need loads of add-ons installing just to export to STL or add a fillet/chamfer.

All my 3D printing workflow is now Linux based:

FreeCad --> Cura --> Duet Web Control (browser-based).

 

[Gina]

1.75mm filament melts more easily due to geometry - more surface area compared with volume.  Also, 1.75mm filament is more easily controlled as it goes through the extruder faster for the same volume of print.  The chemical composition of the different sizes of the same brand and colour should be identical.  The difference you are noticing between 3mm (or 2.85mm) and 1.75mm filament is well known.

[Gina]

The CAD software is interesting.  I'm using a Windoze only CAD software running in VirtualBox with Linux Mint (a fork of Ubuntu).  This is a very old version of Autocad Inventor and very powerful though I have found the odd problem.  No doubt this applies to any software. 

Duet was the best upgrade I have applied to my printers.

[tekkydave]

17 minutes ago, Gina said:

The CAD software is interesting.  I'm using a Windoze only CAD software running in VirtualBox with Linux Mint (a fork of Ubuntu).  This is a very old version of Autocad Inventor and very powerful though I have found the odd problem.  No doubt this applies to any software. 

Duet was the best upgrade I have applied to my printers.

I thought you were using Sketchup.

[tekkydave]

26 minutes ago, Gina said:

1.75mm filament melts more easily due to geometry - more surface area compared with volume.  Also, 1.75mm filament is more easily controlled as it goes through the extruder faster for the same volume of print.  The chemical composition of the different sizes of the same brand and colour should be identical.  The difference you are noticing between 3mm (or 2.85mm) and 1.75mm filament is well known.

I thought the geometry would be the most likely explanation. Some of the "3mm" filament I had was actually 3mm diameter. I've had 2.85, 2.9, 2.95 and every other value between 2.85 and 3

 

[Gina]

54 minutes ago, tekkydave said:

I thought you were using Sketchup.

I was but not now.

[Chriske]

2 hours ago, Gina said:

Duet was the best upgrade I have applied to my printers.

Don't think so Gina.  AI was imo..😉

[Gina]

Could be indeed.

[JamesF]

I'm trying to find time to learn my way around FreeCAD at the moment.  Unfortunately with all the other things going on, it's not happening very fast.

James

[tekkydave]

Decided to recalibrate my E steps/mm for the Titan Aero. PETG in particular seems to be still over-extruding after the switch from 3mm to 1.75mm filament. I currently have it set to 419 steps/mm which is what E3D recommend as a default for a 1.8deg motor and 16x microstepping. It worked fine with 3mm/2.85mm filament.

I removed the hot end parts from the Titan and extruded 100mm of filament - it measured at 104mm.

So:

New Steps/mm = (419 x 100)/104 = 403

Issued an M92 E403 to the Duet.

Tried the above procedue again and got 100.5mm extruded

(403 x 100)/100.5 = 401

Put 401 in the M92 command and now it extrudes 100mm every time.

I put this down to the diameter of the hob being different for the different sized filaments due to it contacting them in a different part of the profile.

Just shows you though - don't just go with the manufacturer's default value - test it yourself 🙂

 

Edited September 29, 2019 by tekkydave

[Gina]

I wouldn't have thought 4% was a significant error in extruding.  I fiddle with my extrusion rate during printing (in Duet) by as much as several tens of percent!  Maybe I'll get things right one day!!!  🤣

Edited September 29, 2019 by Gina