Two Trees Sapphire pro 2 - exploring this corexy printer

[curious aardvark]

okay pid autotune total failure.
once it hits 20c past the set temp (tried several) it just says: temp too high, PiD failed.
So I think my next move is to try and flash it back to the factory original firmware.

That way if I still have the same issues - looks like I need a new board.
Which for around £30 Isn't a big issue.
Just a 2-3 week wait.

[Bikeracer2020]

The filament run out seems pretty essential to me when prints sometimes run into the early hours and then run out of filament.
Anyway I can cope with it not working, the E2 shows approx the amount of filament left on the reel and seems to err on the safe side. I like the Ideamaker slicer as well, it shows the amount of filament required and the cost of it.

I looked at the Leapfrog in the link, looks a super machine, but it's physically too big and much to heavy for my situation here.

If I do send this E2 back I'll be looking for an alternative with a similar build volume.
I think I looked at the BCN after I sent the Creator 3 back and they were possibly out of stock.

It will certainly be with reluctance if I send the E2 back because it really is a cracking printer, just needs a better support back up.

[AutoWiz]

okay pid autotune total failure.
once it hits 20c past the set temp (tried several) it just says: temp too high, PiD failed.

I am pretty sure this is your problem. In the configuration.h file with the heater settings there is a timer that the heater is expected to hold it's temp before the print starts and if your temps are not stable by more than 10c this might be your problem(there is also setting for how exact the temp has to be during this time). You need to calibrate your heaters. You can try tuning to a lower temp first. You can try changing the temp residency time to a lower number for that heater in Marlin. I think the default setting is 10(seconds) and I keep all mine set at 5 because I like my printer to print when I tell it to print. But you can lower it to 2 or even 1 so there is no delay and as soon as it crosses that temp it will start the pid autotune routine. And just put it back to 5 or 10 before you start printing just so the hot end has a good chance to heat soak at operating temp before it starts extruding filament.

[Bikeracer2020]

The filament run out seems pretty essential to me when prints sometimes run into the early hours and then run out of filament.

New sensor working okay and the problem with the lack of fan control is now sorted.
Caused by some numpty conecting the fan to the wrong socket on the assembly line. Good job the reason was discovered before at least two printers were returned.

[curious aardvark]

Okay - she lives !
Properly - or as much as will be for now :-)

I changed the firmware to a custom version a guy made.
Looks nice, has one button load and remove for the filament - which I like.

Fixed the temperature issues - done a couple pid's and updated and saved to eeprom viz g-code and the simplify3d control panel.

Added a micro sd - full size adaptor and prints are now working from the sdcard.

So it's properly functional.
Dimensional accuracy is excellent.
Did a 25mm cube at 150mm/s - great :-)

So I'm currently doing a more comples print. A print in place folding phone stand.
It's 2mm thick and has 6 hinges. Great bit of design.
not as tricky as an iris box, but still a good test print.

My only decision mow is whether to change the 0.4 nozzle for a 0.5.

And grease the z rods - it's developed a 'squeak' lol.

[curious aardvark]

todays task.
Convert it to direct drive.

The dual gear bmg extruder and motor are pretty hefty. But the rails ad top are really heavy duty so I can't see any issues.
Although I will have to work out how to get the belts a bit tighter probably.

But I really want thismone to be good with flexible filament - and even with the shortened bowden - it just isn't :-)

I will have to remove the pyramid though, which is a shame as I like the way it looks - not just another printer cube :-) lol

redesigned bvracket currently printing which should work.
we shall see.

[Bikeracer2020]

And grease the z rods - it's developed a 'squeak' lol.

Seem to remember one rod was not as smooth as the other regardless of which bearing I put on it.

[curious aardvark]

only squeaks when moving down at speed.
otherwise I struggle to even tell when it's actually printing.
In a way it's too quiet lol
I'm used to being able to tell what a printer is doing from the noises they make :-)

So latest news, I've converted it to direct drive - almost. There is a straight line 105mm ptfe tube between the extruder and the top of the nozzle - and that's as short as you can get.
It doesn't appear to have even realised that the extruder and motor are now mounted above the hotend. Just zips around as usual.
The frame is so sturdy and the linear rails so smooth - that the extra weight just isn't an issue.
Sapphire really should mhave made this a direct drive to start with - but then I guess they'd hit the same filament issue I have.

All the conversion took was modifying a simple bracket I found on thingiverse for the original sapphire pro.
The trickiest part was getting the tube the right length.

So my next project is a filament holder.

And i thought I'd go all autowiz and build something so totally over the top it just has to work :-)
Plus I miss the pyramid, she looks kind of naked without a 'hat'.

I've been giving this some thought - always a bad sign.

What I've come up with - sort of lol - is a large cubic framework mounted on top of the machine - the uprights will be made from wood fitted into printed corner brackets, with a crosspiece of 15mm copper tubing (which i have from my brief sojourn into electro-copper plating printed objects), from the centre of which will hang the spool holder on a bearing, so that while it's above the extruder it will also swivel with the movement of the printhead. I figure that way I can reduce the pull on the filament to the bare minimum.

At the moment, with the stock spool holder - small spools of pla and rigid filament work great. But the larger 1kg spools don't even turn well and the filament catches and flexibles just don't work at all. Too much pull and stretch.

Hence my thinking about how to reduce filemant pull to a minimum.

It's still very much 'vapourware' but as I design in my head and then go straight to my wood shop and openscad - it's looking pretty decent.
Still working on the swivel mechanism - I have some bearings from making my rotary axis for my laser cutter.
Just trying to work out how to incorporate them into a hanging spool holder.

The rest is really knocking out some basic brackets and working out just how tall to make it :-)

Too short and you start getting tight feed angles at the corners of the print. Too tall and I won't be able to reach the spool - lol
Also looking at it - my dehydrator sits above the printer and might get in the way.

My other thought - was a much simpler single extension at one corner that just raisess the spool holder - I guess that could also swivel, but probably not as smoothly.

I like the frame idea, plus it's minimal printing and i have a lot of wood and enough copper tubing :-)

[curious aardvark]

right well - machine finished.

In the end I just went for a raised filament spool holder in the front left corner. The spool is held at 45 degree angle to the bed. So pull is even no matter where the print head is.

So last few days I've been putting her through her paces.
WOW !

Original size iris box - one of the hardest things to get clean and smoothly working. Printed them all at 0.2mm layer height.
Started at 75mm/s - best iris box I've ever printed.
Then went up to 100mm/s - just as good, I can't tell them apart.
Then did two at once at 100mm/s.
Both perfect - zero indication that they weren't made one at a time.
I then did one at 150mm/s.
It took a little more twisting to get working - but once freed up, works as smooth as the others.
I can just about tell it from the others - plus i engraved 150 on the base just to make sure.
https://www.thingiverse.com/thing:1817180

So I'm now attempting a print I have only ever managed on my mini-delta with a 0.25mm nozzle.

print in place rounded fidget cube.
https://www.thingiverse.com/thing:1752508

same thing as iris boxes. I'll start at 75mm/s and work my way up.

Unlike a benchy - these items are hard to print and great things to give away or use.

iris boxes great for screws when dismantling things - as when you knock them off the table they don't scatter the screws :-)

And everybody loves the fidget cvubes :-)

I still have the issue of flexible filament somehow managing to wrap itself nehind the extruder drive gear when you attempt to push it fast.
I cannot even see how it manages it - and I have had that thing stripped down to it's individual components.
there is no discernible gap. IT shouldn't be able to do what it does.

So until - if - I decide to change the extruder - I'll just stick to the replicator pro for flexibles.
Not a biggy - but a little annoying.
I do have a spare mk8 extruder that I might make a bracket for.
It would be interesting to compar ethe two. Plus with a single really short piece of ptfe you can make a mk8 100% tangle proof.

Once I've finished mucking around with fidget cubes I'll have a go with some pet-g as well.

Also interesting to see which of the three printing adhesives I currently have, pet-g like best :-)
Plus I have a sheet of PEI on it's way from china.

Yes I do have two other sheets in my workshop - somewhere !
But look as I might, I simply cannot find either of them.
When i put something in a safe place - it's completely safe !

[curious aardvark]

many years ago, when I got my first 3d printer - flashforge creator - the hardest thing to print was plastic mesh.
I pretty much only used abs for the first 6 months and I've not touched it in the last 5 1/2 years. It's not just worth 3d printing with abs.
But the one aspect of abs that's really useful when printing something like plastic mesh - IS the shrinkage factor. As long as the first layer stays stuck to the raft - then what the shrinkage actually does is open up the gaps between printed components.
So while I succesfully printed 2 mesh bracelets in abs - I never managed to do it with pla.

Skip forward 5 years and throw in simplify3d and fine tuning internal gap adjustment, as well as the sheer precision of the sapphire pro - and I thought I'd give it another go :-)

So I was sitting here after knocking out perfect iris boxes and fidget cubes (yeah they rock) at 100mm/s and wondering what the toughest thing to print was that I could throw at Saffy next.

It was a coinflip between my own pin-cushion picture thing and some plastic mesh in pla.

So i thought I'd start with the mesh :-)

In case you hate your 3d printer or just fancy a challenge - I've attached the bracelet file I'm currently attempting to print :-)
Why is this such a tricky print ?
Retractions. Each layer of chain has hundreds of seperate bits and the system has to retract and reload between each and every one.
And if any single bit is pulled up and out of line, it kind of snowballs into several other 'bits'.
I've never seen a 100% complete mesh on one of these prints.

Would I even attempt to print this on a bowden setup ?
Well yeah, but at a stupidly low print speed - maybe 20-30mm/s.

The first layer is PERFECT !
And there are hundreds of tiny little seperate bits in each layer.
Even on the ones that worked in the past - I don't recall a perfect first layer. There was always at least one of the tiny bits that came loose or just got knocked out of line.
The way this mesh is designed you can lose quite a few bits and it will still work

Speed wise I'm runing this at 75mm/s - or as I like to call it now : ultral quality, super SLOOOOOW speed.
Yeah - that's my slow speed :-)

(so i just looked at the rpinter panel - someone forgot to reduce the speed - so it's actually running at 'normal' speed 100mm/s - hot damn !)

I'm also using 0.2 layer height. Honestly can't remember what the recommended layer height for this is - I do have the scad file somewhere and I believe you can generate mesh for any given layer height.
So we'll see how it manages at 0.2

More later...