Two Trees Sapphire pro 2 - exploring this corexy printer

Thing is, when wiring up a 3d printer board.
In this case the robin nano on my 'new' sapphire pro. (Second hand but has never actually printed anything).

I tend to think (erroneously as it turns out) that the positive and negative terminals along the power strip for mains in, extruder and hotbed out - are going to be the same for all of them.

And it turns out that on a robin nano 1.2 board - this is not the case.
The outputs are in fact reversed compared to the input.
So I wired the bloody hotbed up the wrong way round.

Soooooooooo, I'm wondering if I've killed something on the robin board or something on the hotbed itself !

Everything else works fine. printhead moves arounf and hotend and extruder works fine. Eventualy - after changing x&y motors round and then remembering to change endstops as well - all three axis home perfectly.
Which given it's a corexy and the belt setup is totally bizarre - I am amazed that it works at all.

But the bloody printbed will not heat up !

So, do i need a new heating pad or a new board ?

I might see if there's still current coming from the power connector tonight with the multimeter.

Not sure what's cheaper, a robin nano 1.2 or a 220mm heatpad. I do know that the board is probably a lot easier and quicker to change over :-)

But I suspect it's the heat pad that's buggered.

Anyone ever done similiar ?

ps. I have re-wired correctly and still no hot printbed. BUt everything else works fine.

well the hotend needs stripping down and probably having ptfe strategigically inserted, but you expect that.
But apart from the hotbed - it's all good.

Thing is I don't know if th horbed worked when i got the machine as the original owner said he could never get it to actually print. It would just sit there - almost as though it was waitign for the bed to heat up (my thoughts not his).

I'm hoping it's dodgy wiring on the hotbed and my mate can bring his soldering iron and have a play.

check the bed with an ohm meter..

So the heating element in these heated beds do not include any type of a diode. They get warmed by current flowing across a constrained path. And that resistance will build the same amount of heat no matter which way the current flows over it. Why there are red and black wires going to the heat bed idk, but 2 reds would be just fine. Now external mosfets very much care about that polarity coming in. If you are wiring the heat bed output from a mainboard into a external mosfet then you need to ensure that polarity is right. If you are wired directly from the board to the bed and connections are correct and tight and the bed doesnt heat up then you should look in marlin under configuration.h and make sure that the heated bed is enabled. And once you get it all up and running don't forget that PID autotune for them heaters.

Okay - it's worse than I thought ......
printer perfect - but the mobile organic bifurcated interface is f***ed ;-)

Yeah, turns out I didn't screw the terminals down tight enough.
After re-seating and using the correct size screwdriver - everything now working perfectly.
I've even added the top extruder 'pyramid'.
This cuts the bowden tube length down by about half.
And eliminates any really extreme angles in the tube.
Hopefully this'll make it half decent for flexible filament

So everything is now doing everything it should, levelled the bed and am now looking for a simplify3d profile for it and coming up totally blank.

Thing is - does the slicer have to know it's a corexy ?
I guess I'll just have to make my own profile and see what happens :-)

Oh and I need to add another 20mm to the hotend cables, it's kinda tight at the front right corner.
I don't usually do proper cable management and I think i got a bit carried away lol

But yeah, so far - as with all my other printers, only issues have turned out to be my fault :-)

Oh yeh I buggered the print surface while making the levelling nuts so that they don't rotate in their sockets when levelling.
But have just swiped the aluminium with some magigoo - and that will most likely do a much better job anyway.

There are 2 sheets of pei the correct size in my workshop somewjere.
But I put them in safe places and cannot find them, not for love not money.
So magigoo and a slightly reduced print area it is for now.

It Lives !
And oh man is it quiet !

The only noise is the tiny fan in the base and the occasional retraction.

The linear slides on the x&y are silent as are the tmc driven motors.

The odd thing is that while I've now got it talking to simplify3d and happily printing over the usb cable.
It won't print the same g-code from the sd card.
It reads the card and appears to load the file a the hotend and bed heat up correctly.
But doesn't go so far as to home and start printing.

Wondering if I can plug an sd card reader into the screen.
Although it does read the card.
odd.
And a problem for tomorrow.
Yonight I' going to ramp the spee dip and see just what this sucker will do !

Also try and figure out why the fan control isn't working.

There are only two sockets for the c]print cooling fans. And the control panel has fan speed options that don't do anything.

Maybe need to swap over and double wire the fans into one connector.
we'll see. Also try a different sdcard and a micro to standard adaptor.

And sort the bloody hotend temp swing.
It's goes from 10c under to 20c ver and oscillates between.
Never seen anything like it.
The hotbed just hits temp and stays there - like normal.

Have to see what gcode I can throw at the robin nano to try and settle things down a bit.

But I am up to 200mm/s print speed.
hell I'm printing the first layer at 100.
I've aset everything else at 100% and just given the first layer 50%

Magigoo is good stuff lol

You need to PID autotune that heater. But first go over the heater and temp sensor connections and make sure everything is tight and heater and temp sensor are correctly attached to the hot end. Then connect via usb and enter M303 S235 C8 if you're printing with the PETG and it will return to you the correct kp, ki, and kd values for you to go enter into your firmware for the extruders heater. You can also do the same for the bed heater. For that you would enter M303 E-1 S60 C8 and wait for it to return your kp, ki, and kd values for you to go enter in marlin under the configuration.h tab. And then them heaters will work well. Also I like to apply thermal paste to the heater cartridge before i slide it into the heat brick. If i had one that was assembled dry I might be inclined to take it apart and apply some artic silver 5 or other cpu thermal paste. Because I use all E3D stuff all the temp sensors i use are also little cartridges and I also apply thermal paste to them before putting them in.

Quote:

---

Originally Posted by curious aardvark

It Lives !

The odd thing is that while I've now got it talking to simplify3d and happily printing over the usb cable.
It won't print the same g-code from the sd card.
It reads the card and appears to load the file a the hotend and bed heat up correctly.
But doesn't go so far as to home and start printing.

---

That was the problem I had and without the knowledge to know what was wrong my best choice was to sell it to someone who would get it working.
I'm happy it's done you some good.

yeah - it is an odd one.
The actual hardware and mechanics of the machine are excellent. I mean the top plate is 6mm aluminium and the base is made from 2.25mm steel.
It's ultra stable !

But even if card reader had worked - the temperature swing would probably have buggered you up.
Still looking into commands to fix that, pretty sure it's a dodgy firmware setting and those you can usually fix with the right gcode tagged on to the startup code.

That said -= even wth a near 40 degree swing - I set it at 210 and it oscillated between 192 and 230 - it STILL knocked out a couple of trolley keys at 200mm/s. they look a little rough - but they are dimensionally accurate and without the temp issue would be as good as a 70mm/s print from my other machines.

With the extruder mounted on a pyramid frame at the top - I've got the bowden tube dwon to half the length and with no really sharp angles. So I've got high hopes for flexible filaments.
Attachment 15041

Think I found the design for the pyramid on thingiverse. Originally it ws held together with little plastic plugs - so I had to redesign it to use bolts and nuts.
The tricky bit was fastening the feet securely to the top frame. I think he just had his resting - not good. I ended up using cable ties and it's now seriously rigid and going nowhere.
I've actually cjopped a bit more off the bowden tube sonce I took that pic.


It's weird that they obviously put so much effort into getting the frame and general mechanics right - the linear rails are amazing and coupled with the tmc stepper drivers - It's so smooth and quiet that if it wasn't for the filament making noise during retractios - you would not even know it was running.
I find it weird that you'd go the the effort to make something so well engineered and then bugger things up with dodgy firmware.

And just checked - it's running a custom version of marlin 1.

So yeah - all-in-all, well chuffed with it :-)

How you getting on with the flashforge 3 - got a workling one yet ?

Quote:

---

Originally Posted by curious aardvark

How you getting on with the flashforge 3 - got a workling one yet ?

---

It was returned because using one extruder for supports and the other one for the actual print, it was not capable of putting the supports under the print.
The support would be offset and then the next layer of the print would be offset some more and so on until it looked like really severe layer shift with the supports being printed inside the print.

Even though I returned it with a small test print that would demonstrate the problem within about two minutes, Technology Outlet would only print a normal test cube or the two colour cube on the USB stick, which of course came out fine as I had already told them it would.

Anyway I had a refund and that was it.
Not sorry at all now because the Creator 3 has very very noisy fans and I do mean noisy.

I got a Raise3D-E2 to replace it from a different supplier. Just put my fourth roll of PLA in and apart from being on my third filament sensor the only bad prints have been down to operator error.
Although there is a glitch in the firmware meaning the only way to stop the Hepa fan is to unplug it.

Just waiting for a reply from Raise3D about this glitch that they themselves introduced into later firmware and depending on their answer I could be returning yet another 3D printer which with the one you had will bring my total to four manufacturer fails !!!!!!

I think we might start calling you 'lucky' :-)

Filament run out sensors are pretty pointless.
But for that kind of money you would expect something that simple to work. As basic working practice renders them useless.
ie: how big is my print - how much filament is on the roll.

Well if you want to keep trying out mid range idex printers here's another one for you to try: https://www.bcn3d.com/bcn3d-sigmax-r19/

Shame monoprice or qidi don't do them yet.

I mean if you want to push the boat out there's always my all time favourite fdm 3d printer (so far): https://www.lpfrg.com/products/leapfrog-bolt-pro/
Leapfrog bolt. Gorgeous looking machine with so,e seriously over engineered mechanicals.
They were also a nice bunch of blokes to chat with when tct still did non-industrial machines.

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.

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.

Quote:

---

Originally Posted by 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.

---

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.

Quote:

---

Originally Posted by 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.

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.

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.

Quote:

---

Originally Posted by curious aardvark

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.

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 :-)

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 !

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...

my latest sheet of pei arrived this morning.
So before i put it 'somewhere safe' and it vanished into the black hole I've got somewhere in my workshop - I cut it to size and fitted it !

The main reason I've done this is to cover the screw holes at the corners of the build plate.
If it proves as good to print on as the adhesives - even better - but it doe now mean I can print on the entire build plate.
And if it doesn't work that well - well there's always the adhesives - of which I have an abundance :-)

First print on the pei looks good.
Love the smooth base.
Attachment 15217
Adhesion was excellent.
My fault the top surface is a litle 'slumpy' - I had the infill a bit too sparse.
But the first layer went down at 30mm/s and the rest was a solid 100mm/s.
Dimensions are absolutely bang on !
These are shopping trolley 'keys' - I set the 'thickness' at 2.8mm and all three are bang on 2.79.
I'll take that every damn day !

And now I've been able to compare printbite and pei - they are quite different.
Printbite is a LOT harder.
In the three years the printbite has been on the he3d k200, it's been used and abused way beyond reasonable expectations. Getting the mymat nylon off with a metal chisel and a hammer would have destroyed almost all other print surfaces - the Printbite just winked at me and said: 'is that ALL you got ? '

So next task an actual 235x235 test print.

And apart from an extruder that cna do flexibles at speed - I honestly can't think of anything else that needs doing.

I even like the built in cooling ducts. One fan per side and a wide duct setup.
Given how good the 150mm/s iris box is - cooling is definitely working well.

now this is how you lay down a first layer !
there are 592 seperate 'links'. It's not agreat picture but every single one of those 'crescents' is a seperate part. The gaps between them are very small - but there are gaps.
Attachment 15218


So final verdict on the PEI - excellent !
Adhesion for pla is damn near perfect. And it comes off easy once cool.
Going to be interesting to see how long it lasts.
I think the sheet I bought which was 250x250 with 3m adhesive sheet, cost about £7
And hassle free printing from here on in :-)

Be interesting to see if pet-g sticks to it - it is supposed to, but who knows.
If it doesn't - no bigge I've got three different types of print adhexive that all claim pet-g works with them.
So we'll see.

The bracelet printed perfectly, and peeled off very easily. I'll take pics in better light tomorrow - but 4 layer plastic chainmail is a thing of wonder :-)

There was one other thing I did during the build of the printer. Or rather did not so.

It comes with a big metal baseplate, I did not fit that.
As long as you tidy the cables up with some cable ties so that nothing drags on the ground - there is no point fitting the base plate and several reasons not to.

1) much much better heat dissapation - all the plate does is to keep all the heat in there. Without it everything stays a lot cooler and there is plenty of air circulation.

2) if you ever need access to the electronics and wires - all you need to do is tip the machine on it's side. No big awkward metal plate to remove.

It's not needed, so I'd recommend you don't use it.

Some of those little sticky anchor points would be useful.

One thing I haven't mentioned is how long it takes to get to temperature.
The new firmware has a 'preheat' button - one for pla and one for abs - which I'll probably eventually use for pet-g.
Because it heat the bed and printhead up at the same time. To go from ambient room temp to 65c bed and 200c hotend take exactly 2 minutes !

That's bloody quick.

So today I decided I'd have aplay with pet-g.

I have aroll of creality white pet-g that I believe I bought by accident.
Pretty sure it was really cheap and I was out of white pla and just clicked without checking.

So bed temp 85c.
Made a little model for testing purposes. It's a 7 minut print with one piece inside a larger version. Got some gentle underhangs and overhangs and the gap tests retractions.

First print was 240c and 50mm/s.
I've kept my retraction and fan settings from pla.
First print - looks great, zero stringing stuck to the bed like a limpet and came off easily when cool.
I am really liking this sheet of PEI.
Actual strength of the print is not brilliant, about the same as pla- but you can peel layers as well.

Second print up to 250c and two models a decent way apart on the build plate.
still 50mm/s.

Well ! colour me impressed.
Zero stringing, really strong prints - much mangling with pliers and no layer seperation, just squashed plastic.
Really clean sharp transition areas from one model to the other.

Either this pet-g is not normal or it's really not as tricky to print as I've been led to believe :-0

Let's see what it will do at 100mm/s :-)

the 100mm/s pet-g looks pretty good.
My cooling settings have been left - so 0% for first layer and 100%b for all subsequent layers.

So I'm running the same print on the delta with white pla and same speed.

Curious to see if fast pet-g is noticeably stronger than pla.
No actual scientific tests here - just a pair of crushing pliers.

***

well that was interesting - and children, before you destruction test pla - PUT ON SAFETY GOGGLES !
Umm, yeah the pla thin walled shape did not deform it just exploded and a bit did bounce off one of my eyeballs.
The solid pla piece was just totally non-deformable. The pliers won't touch it.
The 100mm/s petg - was not only a better looking print than the pla. With no stringing and sharper lines and zero blobbing. But while it will crush, it does not shatter. The solid basically ends up looking like a piece of mangled chewing gum. So layer adhesion - excellent, actual strength - nowhere near the pla.

Really interesting, as I've never done this kind of comparison test before.
I'll stick some pics up in a bit.

But this pet-g is really interesting stuff - not at all what i was expecting.
I can see it being really good for things like boxes with tight fitting lids.

I'll try a 75mm/s iris box next i think.

So what I actuallyn did was to make alittle scraper.
At which point I discovered that pet-g adheres to the PEI a bit too well.
Could not get it off in one piece.
The other prints were small and knocked off fairly easy. Obviously the larger the amount of material in contact with the bed the stronger the adhesion.

I tried it at ambient temp and no-stick.

So now trying it at 60c printbed.

well that worked surprisingly well.

Now whether this particular pet-g has very different properties to everything I've ever read about it - or whether it simply loves being printed on this particular make PEI - I have no idea.

But it's basically printing as easily as pla and almost as fast.

So of course I went to see if I could buy any more creality pet-g.
Of course not !
What was I thinking.

Nobody on ebay or amazon sells it - not even sure creality still make it - I bought this roll last year, from sovol on ebay, actually.

PETG is nice to print with. It has opposite the adhesion issues of PLA. PLA comes off the bed nice and separates at the layers if dropped. PETG will break chunks of glass off of your bed and break right through the layers. I usually have my heated bed temp set a little lower around 70C and you can raise that z offset the 0.05 or 0.1 or so from where you print with the PLA. You just do not want as much squish.

PETG seems to like moisture so if there's lot's of stringing, drying the filament seems to be a good idea. I had quite dramatic improvement in quality after I dried couple of PETG rolls. The problem was actually worse with bigger nozzle, I had ugly gaps in seams where retractions had happened and of course stringing and it was mostly fixed with dry filament.

well the roll I'm using has been hanging around in my workshop since last year.
I did open the vacumn pouch.
But I generally don't have mositure issues anyway. once it's been opened most of my filament is just stacked up on shelves. While it does rain alot in england - we don't tend to have high humidity that often.

And yeah next time i play with it I'll put the z-gap right up.

I'm also starting to suspect that this particular PEI is menat for printers without heated beds.

I've started making some pasta extrusion plates for my hand cranked mincer/grinder.
Even at at 50c the pet-g only came off the plate after heating the plate back up and eventually getting a metal scraper under one part.

Ended up making the first one out of pla anyway.

But absoilutely no stringing, prints fast, I've had the cooling fans full on from layer 2 on.
So whether it's just this particular brand (whihc no longer appears to exist) I don't know.

I do have some transparent pet (no glycol). So I'll have a play with that.

I'm also thinking - maybe polycarbonate might stick to this stuff.
So that's probably what I'll play with to0night.
The hot end is supposed to be good for up to 260c - though I have noted the official config file has max set to 290c and the one I'm using has it set to 285.

So we'll see - if it goes up in smoke - I'll know better for next time lol

this is the script for the chainmail bracelet - I modified it slightly to remove the supports he had and make the clasp actually print properly and not bugger everything up :-)

Quote:

---

/* chainmail bracelet

Gian Pablo Villamil
March 2013

Based on Zomboe's chainmail (http://www.thingiverse.com/thing:8724)

Added clasp mechanism
Printed two layers of chainmail to prevent link rotation
Added (hacky) offset to allow horizontal bridges to clear

Extra bonus bitchain() module allows for multi color printing

v2 reduced clasp bump, added more support
v3 added internal support to female end, opened up clearance on female end, thickened tongue
v4 retightened clearance, removed internal support on female end

*/

// Variables

layer_thick = 0.27; //filament layer thickness
line_width = 0.4; //filament line width, = layer thickness * width/thickness

b_width = 2; //block width in lines
b_height = 6; //block height in layers
d_height = 2; //diagonals height in lines

r_halfwidth = .95+1.5*line_width*b_width+0.45; //ring inner half width
corner = tan(22.5)*line_width*b_width; //fills in corner between block and diagonal

claspwidth = 38;
claspheight = b_height * layer_thick * 2;
claspthick = 2; // in layers
clasptongue = 6; // in layers
claspsides = 6;
claspclearance = 0.2;

offsetmod = 0.8;

bitmap =
[
[ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 0, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 0, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 1, 1, 0, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 1, 1, 0, 1],
[ 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1],
[ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]];


module ring(modifier)
{
module blocks() //four blocks
{
for (i = [0:3]){
rotate([0,0,90*i])
translate([-0.6001,r_halfwidth,0])
cube([1.2002,line_width*b_width,layer_thick*b_height]); }
}

module diagonal() //thin diagonals
{
mirror([-1,1,0])
translate([0.6,r_halfwidth,0])
union() {
cube([corner,line_width*b_width,layer_thick*d_height]);
rotate([0,0,-45])
translate([-corner,0,0])
cube([corner,line_width*b_width,layer_thick*d_height]);
}
translate([0.6,r_halfwidth,0])
union() {
cube([corner,line_width*b_width,layer_thick*d_height]);
rotate([0,0,-45])
translate([-corner,0,0])
cube([corner,line_width*b_width,layer_thick*d_height]);
rotate([0,0,-45])
cube([sqrt(2)*(r_halfwidth-0.6),line_width*b_width,layer_thick*d_height]);
}
}
module 4_diagonals(mod) //all four diagonals
{
translate([0,0,0]) union(){
diagonal();
mirror([1,1,0]) diagonal();
}
translate([0,0,layer_thick*(b_height-d_height)*mod])
union(){
rotate([0,0,90]) diagonal();
rotate([0,0,-90]) diagonal(); }
}

union(){
blocks();
4_diagonals(modifier);
}
}

module chain(w,h,m) {

module line() {
for(i=[0:w-1]){
translate([i*(2*r_halfwidth-.55),0,0])
ring(m);
}
}

for(j=[0:h-1]){
translate([0,j*(2*r_halfwidth-.55),0])
line();
}
}

module bitchain(w,h,c,m) {
for(i=[0:w-1]) {
for(j=[0:h-1]) {
translate([i*(2*r_halfwidth-.45),j*(2*r_halfwidth-.45),0])
if(bitmap[i][j]==c) {
ring(m);}
}
}
}

module clasp_male() {
translate([8,0,0]){ //moves reduced clasp to chain
cube([2,claspwidth,layer_thick*(b_height*2+1)]);

translate([0,0,-0.8])union(){
translate([-8,claspsides+0.25,(claspthick+1)*layer_thick])
difference() {
cube([8,claspwidth-(claspsides*2)-0.5,clasptongue*layer_thick]); // tongue

translate([7,0,0]) cube([1,3,clasptongue*layer_thick]);
translate([2,2,0]) cube([5,1,clasptongue*layer_thick]);

translate([7,claspwidth-(claspsides*2)-3.5,0]) cube([1,3,clasptongue*layer_thick+1]);
translate([2,claspwidth-(claspsides*2)-3.5,0]) cube([5,1,clasptongue*layer_thick]);


}

//bumps

translate([-2.75,claspsides+0.5,(claspthick+1)*layer_thick]) cylinder(r=1.0,h=clasptongue*layer_thick,$fn=16);
translate([-2.75,claspsides-0.5+(claspwidth-claspsides*2),(claspthick+1)*layer_thick]) cylinder(r=1.0,h=clasptongue*layer_thick,$fn=16);
} // end union
// supports

//translate([-8,claspsides+0.25,0]) cube([0.4,claspwidth-(claspsides*2)-0.5,3*layer_thick-0.1]);
//translate([-7,claspsides+0.25,0]) cube([0.4,claspwidth-(claspsides*2)-0.5,3*layer_thick-0.1]);
//translate([-6,claspsides+0.25,0]) cube([0.4,claspwidth-(claspsides*2)-0.5,3*layer_thick-0.1]);

//translate([-5,claspsides+0.25,0]) cube([0.4,claspwidth-(claspsides*2)-0.5,3*layer_thick-0.1]);
//translate([-3,claspsides+0.25-2,0]) cube([0.4,claspwidth-(claspsides*2)+3,3*layer_thick-0.1]);

//translate([-1.5,claspsides+0.25,0]) cube([0.4,claspwidth-(claspsides*2)-0.5,3*layer_thick-0.1]);
} // end mod


}

module clasp_female() {
cube([10,claspwidth,layer_thick*claspthick]); // bottom
translate([0,0,claspheight-(layer_thick*claspthick)]) cube([10,claspwidth,layer_thick*(claspthick+1)]); // top

translate([0,0,layer_thick*claspthick]) cube([2,claspwidth,claspheight-(layer_thick*claspthick)]); // back

translate([0,0,layer_thick*claspthick])
difference() {
cube([10,claspsides-claspclearance,claspheight-(layer_thick*claspthick)]);
translate([7,claspsides+0.5,0]) cylinder(r=2,h=claspheight-(layer_thick*claspthick)+1,$fn=16);
}

translate([0,claspwidth-claspsides+claspclearance,layer_thick*claspthick])
difference() {
cube([10,claspsides-claspclearance,claspheight-(layer_thick*claspthick)]);
translate([7,-0.5,0]) cylinder(r=2,h=claspheight-(layer_thick*claspthick)+1,$fn=16);


}
// internal supports

// translate([3,14,layer_thick*(claspthick+1)]) cube([7,0.8,1.8]);
// translate([3,25,layer_thick*(claspthick+1)]) cube([7,0.8,1.8]);
}

module bracelet() {

//usage: chain(w,h) gives w columns, h rows

translate([-34*r_halfwidth,-7*r_halfwidth,0])
chain(38,8,0.8);
translate([-34*r_halfwidth,-7*r_halfwidth,b_height*layer_thick])
chain(38,8,1.0);

translate([170/2+1,-(claspwidth/2+2),0])
clasp_female();

translate([-170/2-16,-(claspwidth/2+2),0])
clasp_male();

//translate([-170/2,-(claspwidth/2+2),0])
//cube([170,claspwidth,1]);

}

module colorchain() {


translate([-15*r_halfwidth,-15*r_halfwidth,0]) {
color ("blue") bitchain(16,16,0,0.8);
color ("white") bitchain(16,16,1,0.8);}



translate([-15*r_halfwidth,-15*r_halfwidth,b_height*layer_thick]) {
color ("blue") bitchain(16,16,0,1.0);
color ("white") bitchain(16,16,1,1.0);}

}

//
bracelet();

---

https://www.thingiverse.com/thing:4246413

Dead simple - 100% effective.
Simple belt tensioners. Or rather something that makes tensioning the belts really easy :-)
They are currently being printed.

The belt tensioners needed a little adjustment.

I put -0.1mm on the internal surfaces of the 'screw/belt holder' part. As it was almost impossible to get belt into it otherwise. I also made the outer collar part 3% larger, otherwise it was too tight a fit.

That done, they look good and should work well.

Today, however, while i wait for a second motherboard for a super computer I'm building for a client to play flight sim 2020 on - I am playing with polycarbonate.

Currently running at 260c and 70 for the bed.
Given how different my pet-g bed temps are (60c) to the universally accepted temp, I thought I'd start low this time :-)

So just printing a simple small flat scraper - as I have a pet-g one for comparision.
It's 2mm thick , tests nothing apart from bed adhesion and printing temp.
Also only running at 50mm/s - mainly because 260c is on the cool side for polycarbonate and there's no point pushing it for the first print :-)

Results sa far: pc sticks well to pei at 90c with a small amount of 'smush'.
Any hotend temp over 260c and it struggles.
Fortunately that does seem sufficient to print pc.

I have reduced fan speed to 0%
Curious to see how that will effect stringing when I do a multi part test.

But at the moment it does look like I can print polycarbonate :-)

Okay I'll qualify that. I can make scrapers and pretty decent spudgers. And they are really strong and tough. Noticeably stronger than the pet-g.
But complicated prints are going to struggle.
Realistically you need to be able to reliable print at 270c and above.
And that is not happening with the stock hotend.
Bless her she tried - but anything over 260 just wasn't stable.

But an interesting experiment and I can at least print some things from polycarbonate.