Finished and working...

My new printer produced it's first output over the weekend, pretty impressed with the quality

A couple of hopefully minor questions?

The recommended temperature for PLA is 190 degrees but I cannot get anything to stick to the rubber base
unless it is printed at 235 degrees?

I upgraded the firmware to 1.11ch and replaced the SD disk image as indicated and have noticed that the button
'Auto Bed Compensation' is labelled as 'Auto Delta Configuration' (or something similar)

To get a successful print I have to click this or the print head appears to be too far off the print surface - and Sunday
evening this resulted in crashing the print head against one of the uprights and breaking two of the support rods

I had previously clicked this several times with no issues...

When this goes to the four points on the print surface it reports recalibrating/configuring 5 odd millimetres which
sounded a lot to me?

I have replacements on the way but can see why people replace them with carbon fibre rods, as I had already broken
the spare previously (clumsiness on my part) or are these considered sacrificial? (better to break these than damage
something more serious I guess!)

Having said that I am wondering what caused this? Looked like the print head was not centred before I tried this, but
I had also previously clicked 'Home all axis' and up to this point had been getting on reasonably well!

I can understand homing the axes when you first switch on but is this necessary before every print? How often should one
be 'levelling' the print bed (if this is what is happening?)

Some of the prints I tried did not stick to the print surface very well so I am wondering if when I 'slice' them I should be tweaking
the gCode before trying to print it - but I had managed to print 'robots.g' three times without any problems, using two different
filament spools as well

Thanks

Chris

It sounds like you have never edited the M665 and M666 commands in config.g to match the values found by auto calibration.

---

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 run the bed calibration every time I turn the printer on. Due to the homing switches you will always see a small amount of variation. Also I would recommend you increase the probing of the bed to the 6 point check it made a lot of difference on my fisher.

The acrylic rods are junk, I really recommend you don't waste your time with them, even if you don't crash into anything they will fatigue and crack

On my fisher I always seem to have to print on the higher side of the recommended temperatures but not as high as that for PLA. With excelvan PLA I usually do the first layer at 220c and then drop to 210 or 215 depending if i want a strong part or a pretty part. For small objects I would drop the temp and slow the print speed to give the pla more time to cool. This requires some experimentation and can be pla and model specific.

I gave up on the build tack and switched to glass (initially used double sided carpet tape to stick it to the original plate) and have been very happy with it. You may find that you need to print slightly closer to get good adhesion which can be tweaked in config.g in the G31 X0 Y0 Z-0.1 P200 command. making the Z-0.1 smaller (ie Z-0.2) will make you print closer.

---

DC42 Kossel 330mm x 2meters
My Thingiverse Creations

Ah - excellent, thought there might be some adjustment required - I called Traplet (where I got the printer kit from) and their technical support guy was very helpful, he also mentioned running the 'bed4' macro from the macros list although thought this might be the same as clicking the 'Auto bed compensation' button (the name changes to 'Auto Delta...' when the printer is connected??!)

I have noticed a message on the web ui about corrections being made after the auto bed compensation, and am wondering if this means the 'config.g' file is being altered by this process, I'll have to have a look - failing that I guess it's the bit below

M665 R81.0 L160.0 B75 H167 ; set delta radius, diagonal rod length, printable radius and homed height
M666 X-2.6 Y1.6 Z1.0 ; put your endstop adjustments here

But I have never altered these directly - home all does not show any messages/corrections? Looks like the M665 figures are fixed - the radius and rod length does not change surely - although R81.0 looks a bit large (bed is 150mm across so should this be 75mm?) might explain how my support arms got broken!

Good news is my replacement acrylic arms were delivered this morning so I am hopefully back in action, and I have some carbon fibre rods and ball ends on the way...

Pity the documentation about the 'config.g' is so sketchy (missing!) - there was nothing in the commissioning/printing write up from Traplet or eMaker about amending this file or what the macros do (not too familiar with Python but I am reading up on it)

---

Chris Parsons
-----------------------------------------------------------------------
e. parsons.cj@gmail.com
m. 07966 754528
h. 01392 278549
-----------------------------------------------------------------------
Exeter and District Model Engineering Society
Newton Abbot and District Model Engineering Society

Pilot of G-EFOX and Mixed Gas Rebreather Diver

You have to run M665 and M666 with no parameters, after running auto calibration a couple of times, to get a readout of the new values of those parameters. Then use the web interface to edit those values into config.g.

---

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].

Hi dc42, i would like to make the changes in config.g as described but when I run the Auto Delta Calibration I just get a small blue strip across the bottom of the screen showing with what seems a variance measure 'Calibrated 4 factors using 4 points, deviation before 0.022 after 0.000'
Where do I find or how do I get the x,y,and z parameters to enter into the config.g.

Regards Gary J

Quote
Gary J
Hi dc42, i would like to make the changes in config.g as described but when I run the Auto Delta Calibration I just get a small blue strip across the bottom of the screen showing with what seems a variance measure 'Calibrated 4 factors using 4 points, deviation before 0.022 after 0.000'
Where do I find or how do I get the x,y,and z parameters to enter into the config.g.

Regards Gary J

With 4 point calibration the XYZ parameters are not changed. You need 6 point calibration if you want to adjust them too.

---

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].

Hi
As said in good advice above after doing bed calibration run M665 and M666 with no parameters then feed these readings back into config.g on the settings tab, not forgetting to click the tick at the bottom of the web interface to save the changes and then do a software reset. Re-home and calibrate again.

I had no problems on clean BuildTak with the supplied white PLA ('filament print' branded) at 210 but then tried black YS and this would just not stick to the bed or the previous layer. I increased the temp bit by bit and started to get better at 218 with the Z parameter in G31 of the config.g reduced from -0.2 to -0.05 to print closer on the first layer, this helped a lot. Also better was printing first layer a bit faster (25% speed) the 10% that was programmed into the slic3r config I received is really slow and the first layer cools before the second is applied.
I then tried Velleman grey filament and this also needs to be up at 218/220 (but seems much better layer to layer than the YS) I think the heater - fan - thermistor combination just means the fisher reads higher than the real extrusion temp.

*** By far the best improvement I've found is using PVA glue diluted 50/50 with water (Evostick weatherproof wood adhesive) applied in a very thin layer with a tissue onto the BuildTak ***
This can be left for several prints just wiping in between with water (it is waterproof), then cleaned of easily with isopropyl alchohol and re applied for the next few prints.

A couple of other points -

Built as per instructions I fitted 16mm screws to fix the hot end together !! these hit the bed contacts when calibrating and had to be shortened.

The acrylic arms seem to be working ok, I didn't rely on "the screws self tap into the other side" during any of my build, I have a 3mm tap and tapped the holes in the arms (and the rest of the parts that required it) and used three bond 1401 that I use at work to lock them. fingers crossed all still intact. Lubricated with Ambersil AMS 4 grease (also from work, don't tell the boss !)

Hope some of this is useful,

Steve

Spoke (a bit) too soon...over the weekend I had a few problems with prints not sticking to the bed, and have ordered some blue painters tape to try - surgical alcohol didn't make any difference.

More worryingly is that on several occasions the 'Home All' caused the print head to crash into the print bed (usually off centre) requiring an emergency stop and power off reset - at which point it behaved again

It's a pity the end stop leds are not more visible, I am wondering if the 'Home All' is missing one of the stops?

I have tried running the 'bed4.g' macro as well as the 'Auto Delta configuration' and they are not the same - use difference G codes? I was advised by Traplet to use 'bed4.g' a few time to get the corrections down low enough but sometimes the corrections do not get smaller (and on a couple of occasions actually get bigger!) When the calibration is close to the towers it occasionally makes odd noises which might be normal but it looks to me like the nozzle is scratching off the edge of the buildtak surface.

I saw the comment about amending 'config.g' to record the end stop corrections, and the M665 and M666 amendments, and that it is better to have a 6 point probe, is there an idiots guide to all this somewhere? I saw the calculator referred to by DC42 but am not sure how this works? The 'bed4.g' and 'bed4meas.g' macros were supplied with the web interface (1.11ch) but there is no documentation for them? Where does one get the 6 point probe macros from?

My colleague commissioned his Fisher over the weekend and is getting similar problems, he managed to print one robot.g after 10 attempts!

I do run the 'bed4.g' macro a few times but every time I have my heart in my mouth waiting for a grinding noise and another broken support arm...

Chris

Quote
cparsons
More worryingly is that on several occasions the 'Home All' caused the print head to crash into the print bed (usually off centre) requiring an emergency stop and power off reset - at which point it behaved again .

It sounds to me that one of your homing switches is not triggering consistently at the same point every time you home. I don't think you are the first Fisher owner to have this problem.

---

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].

What would be a solution? Change the switches for something else?

Chris

Hi Chris, I had similar problems and as highlighted by dc42 it was the micro switches. I slotted the panel to allow the switches to move closer to the arm assembly which cured the problem. However, the side panels seem to buckle after a little and could cause further issue. I lucky it has not, but others have replaced the panel with greater thickness and others have replaced the switches with hall effect switches which are not mechanical. How these are installed I'm not sure so can't help you any further, but i'm sure there is more detail in the forum.
Gary

Thank you Steve - I checked my build last night and you are correct, the 16mm effector screws are far too long, I have changed these for 12mm and things are a lot better (no more horrible grinding noises!)

I have written to Traplet and suggested they correct the build instructions before anyone else ends up breaking something!

The calibration is now working correctly and the prints are a lot better...this is turning out to be a bit of a minefield

What I now want to do is try a 6 point calibration, the four point calibration does appears to be recording it's results in config.g (?) - and now when I run it several times the accuracy gets better and better - down to 0.02 at one point

Chris