On second thought scratch the above video. Just watch this: https://www.youtube.com/watch?v=DCdQkgSwNoU
Printable View
On second thought scratch the above video. Just watch this: https://www.youtube.com/watch?v=DCdQkgSwNoU
I have been busy this weekend wiring this machine. I am still not done but here is my progress so far. I do NOT build rat's nests and I profoundly respect a well laid out electrical system with a clean appearance.. https://i1050.photobucket.com/albums...oWiz/158_1.jpg https://i1050.photobucket.com/albums...utoWiz/159.jpg
Hey CA what do you think about "Kumonga"? It is the giant mutated spider from Godzilla.
Well I better decide on the name soon. This thing is almost done. The wiring is complete and I just need to mount the touchscreen and the step down converter and then I get to start playing with marlin 2.0. I am pleased with how all this has come out.. https://i1050.photobucket.com/albums...utoWiz/161.jpg
have a look at this: https://coda.io/d/M3D-Official-Troub...e_sudLH#_luU6H
Maybe for the Huntsman mkII :-)
If you want to go for fictional spiders: Shelob is the original mother of them all :-)
But given that huntsman's are real and as wandering spiders, prone to turn up at random in people's houses - much scarier !
A lot of the quadfusion info would seem to be relevant to the mumu setup as well:
Quote:
- Foreword
- The QuadFusion (QF) is a first of its kind - a printhead cramming 4 full direct drive extruders into the same space that most bowden printers use for one the hot end. This is possible thanks to the direct-drive, fixed crush-distance gears, and highly geared stepper motors in the QF. QuadFusion is innovative, there has never been a printhead like it, and this creates unique challenges that requires retraining for success. Thus, it is important to start with the right mentality to achieve success: The QF requires new knowledge - clear your mind, and do you best to avoid the assumptions and biases of other printers, and adopt the methods laid out in this guide.
- It's all about the force. The QF has 4 times the force (up to 50 lb), and up to double the pressure of a standard 3D printhead. This leads to many unique challenges, such as extra force on the nozzle, and competition between ports. You should read through this guide as many times as it takes to appreciate this statement to achieve maximum success.
- There are four times the number of spools, filament flow paths, extruders, etc. So if things are going as well as a traditional one-extruder 3D printer, one should reasonably expect 4 times the rate of failure. Luckily with good knowledge and understanding of the print head, failure rates caused by things like jamming should approach that of a single-extruder, noting that this result would mean each individual filament extruder as a subsystem is more effective and and successful.
- Material Selection
- The QF was optimized primarily around PETG, and M3D's ABS-R brand. For learning purposes, please use PETG, and not PLA. The low temperature stability of PLA compound the difficulties of learning the QF technology, as it will Grind Through (Concept)easier and needs more force.
- Spool Size and Placement
- It's all about the force
- There are many forces acting on the printhead and filament at any time. The printhead pulls filament and unravels the spools. Too many spools of heavy size requires too much pull force from the print head and can cause the printhead to move out of it's current Z plane, causing wobbly looking prints or Grind Through (Concept). Following this guide will prevent these issues.
- Spool Placement / Filament Path
- The number one key to success and cause of failure is spool placement. The spool should be placed so that the filament has a direct line of sight to any position the printhead will be during a print. The filament ideally will have a straight path as well. It is okay to bed filament around a low-friction object like the edge of the print frame, provided is does not cause more than 20-30 degree bend angle from an ideal straight path. The bend angle over the entire filament path should never exceed 90 degrees, unless a customer PTFE/POM feed tube is provided. Ideally, the filament should be allowed to unwind freely from the spool without constraints.
- The ideal spool position is above the printhead. For smaller spools, placing them directly behind the printer works well too. Spool can also be placed to the left/right sides of the printer for printing small objects, however, there is a significant chance of X axis stalling during very large prints so it is generally not recommended. Never feed filament over the top of the printer frame as this bend angle is too aggressive.
https://coda.io/contentProxy/zE73kMb...8a39ffb5f5e1b0https://coda.io/contentProxy/zE73kMb...91876d1e4bc6ebhttps://coda.io/contentProxy/zE73kMb...c208b61f4c5298
Example of successful spool locations and filament feed-paths
- Spool Size
- The ideal spool size is 250 ft spools as these can be placed behind, to the sides, or above the printer and generally lead to success. Remember, there are 4 filaments, so this is already 1 kg combined. 500 and 1000 foot spools can be used, provided they are in a very low friction system, like one that places spools on bearings and places spools above the printhead. Do not use any spool larger than 1 kg weight / 1000 foot length.
- Spool and Filament Quality
Hey thanks. I've already got the cards stacked against me here a bit as I am using my titan aero extruder that does NOT have the flament sensor in it. There is a sensor in the mmu2s but not in the extruder. I am going to see how far I can get before I upgrade. It would be nice to get a few prints out of it in this configuration. We will see. I am in the home stretch. Tonight after work I will start to design the screen mount. I appreciate the help CA.
I drew up the mount for the touchscreen and it is printing. Gonna take 13 hours and it is printing out on Printalicious..
https://i1050.photobucket.com/albums...0mount%202.png
https://i1050.photobucket.com/albums...0mount%204.png
some interesting info there - in particular the difference in quality from where it pulls the filament from.
This is something I've noticed particularly from my replicator clone. which pulls it all the way up and over from the back.
Causes real problems with flexible filaments. They stretch, which reduces the amount and causes issues.
It's interesting that the delta with a bowden setup can print flexibles almost as fast as the knp which is a direct drive. It seems that pushing flexible filament down a shortish tube is often better htan pulling it along a long one.
The klic'n'print has a single piece steel shell (beautifully made machine), and I have yet to work out a decent mount attachment to move the spools above the machine. I don't actually want to drill holes in the case lol Plus there's a sort of double skin which would make it difficult to get at the underside of any holes.
It does need to be done.
lol this is the problem with and industry like 3d printing, as soon as you embrace one good idea another three better ones come along a week later :-)
What I do find odd with the quadfusion is that they stopped at a 4 colour mixer, while stating that you need to change white for lighter colours, black for darker and transparent for more depth.
Now if you were to attach a mumu to the 4th spot for the 'key' filament - that could be interesting.
Also I can't find any indication of slicer software on the website. You go to the software support page for the quadfusion and it's blank.
Kinda weird, it's got to be coming from somewhere.
13 hours !
For a simple screen mount !
Damn, you really do need a faster machine lol
I don't have that kind of patience - I've been printing base mounts out for my rotary axis for my k40 (had to buy the bearings and rods and grub screws, but printed everything else including the rubber drive belt and gears) - 2 in 1 hour, and that's at 0.2 layer height, which is hi-res for me lol
You really need a delta :-)
150mm/s and still bang on spec for the sizes.
Might not be pretty - but then I don't really care as long as they work ;-)
Actually you need to start thinking about getting a k40 - it's a tinkerer's dream ! As well as being an industrially capable machine. Once you start cutting acrylic sheet, there's no way back. It's like the first practical thing you design and 3d print - at that point you've fallen down the rabbit hole :-)
is that the latest cura ?
Looks a lot better laid out than last time I tried it.
It is Prusa Slicer 2.0 or now it is 2.1 I think. It used to be called slic3r. The display mount is 9" long, 6" wide, and ~4" tall. It is a big part that will wrap around the side of the frame. It is almost done now.
There is a big part of me that feels like this whole Black Widow thread belongs inside of my thread for Printalicious. My 3 color mixing printer thread. Because that printer has made almost all of the parts for this machine. On the initial build and now on this rework. Printalicious truly is the unsung hero of this project. And it absolutely could not happen without this home made printer..
https://i1050.photobucket.com/albums...utoWiz/162.jpg
Here is the display mounted to the frame..
https://i1050.photobucket.com/albums...utoWiz/163.jpg
https://i1050.photobucket.com/albums...utoWiz/164.jpg
Ok guys the last part has been designed.. https://i1050.photobucket.com/albums...96%20mount.png From here I just need to print this out, install it and we can get started on understanding how to work with marlin 2.0, a 32 bit controller, LV8729 drivers, and an MMU2S all at the same time. It is going to be a while before this part is done printing. So in the meantime here is a selfie I took with my new machine.. https://i1050.photobucket.com/albums...utoWiz/166.jpg
looks like the bed should be wider ;-)
Well I am done building. Sadly the machine is now complete and my creativity must come to an end. Here is a proper 5 filament multi material machine as I understand it with 400x250x350 actual build volume.. https://i1050.photobucket.com/albums...oWiz/170_1.jpg
so go exercise your creativity and make something with 5 materials !
Just running that past a slicer is going to keep you occupied for quite some time ;-)
Yes CA. I have no idea how to use or program this thing. lol. But isn't the journey hafl the fun of any trip? :p I have some youtube tutorials to watch I think. lol.
you know it looks a bit like a transformer - a head at the top would finish it off nicely.
Just spotted the temp readout - very neat :-)
It is voltage. The LM2596 step down converters I use to power the 12v noctua extruder fans have a display with a pot for adjustment and a button to toggle between input or output voltage on the display. In the picture it is showing input voltage which is the 24v power supply output at 23.9v. Here is a link to the step down converters I use: https://www.ebay.com/itm/DC-DC-Buck-...wAAOSwY45UMfSP
Knew it was familiar !
it looks like one of the leaping beasts in one of the metroid games :-)
So my first impressions of the SKR 1.3 mainboard are that it is incredibly cheap. I am very unimpressed with the lack of LEDs on the mainboard. The MKS boards all had them and I think even the GT2560 board that came with my cheap ebay GEEETech printer had LEDs onboard. And at the beginning of my 3 color mixing printer thread we can all see the leds on the mainboard I had with the 6 stepper drivers. No LEDs mean no conformation of when it powers on or off or when it starts or stops being written to or if it is even actually being written to. I would really like to have these visual confirmations because I have to use new software I am not used to.
Quick amendment to my last post, there is a power LED on the mainboard. The 5v power requirements of the MMU2S and the extra stepper driver expansion boards are overcoming the 5v supply and pulling it down. I will now try to supply 5v from a usb charger or something and report back on that success. In the meantime here is my machine all powered up less the MMU.. https://i1050.photobucket.com/albums...Wiz001/171.jpg
so go print something !
:-)
let's see what you can do quality wise at 150mm's and 0.4 layer height - basically what I use for rough prints on my £200 delta.
Mind you I do use a 0.5mm nozzle.
So I'll accept 150 and 0.3 layer height on a 0.4 :-)
let's face it, ANY machine can do quality at low speeds and thin layers. Where it's at, is the larger layers and faster speeds.
And - obviously get the mumu working :-)
Give me a model and we can have a print off ! :-)
IF and when all this goes well when I upgrade my extruder to the bondtech so I can have a filament sensor for the mmu at that time I will also install a volcano and a .5 or .6mm nozzle. I am totally with you on larger prints coming out decent and in a decent amount of time. In the meantime I have to figure out this new marlin that does not use the Arduino environment. I am all for a print off with you CA. But on this machine. My others all have this problem with artifacts or ghosting in the prints. They come out great dimensionally accurate and stuff. They fit and work great as specific parts I design and print. But if you look at one of my prints for too long I swear it will give you vertigo. That is why I had to stop the LS3 engine I was printing. I cant stand the ringing/ghosting/artifacts, or whatever it is that makes my prints look the way they do. This machine with these high quality parts should correct this. I hope.
Hey CA buddy, do you have any experience with the duet boards? I am now mulling the duet 3. It is just coming out and will be available for regular order on Novemebr 28th. I will buy one and upgrade a.s.a.p. if only i know i can program it. Is it the same as working with the SKR 1.3? or totally different software?
not a clue. Those thigs are way too rich for my blood.
I've met the people behind duet on several occasions. Very helpful, good people.
So best to just ask them :-)
holy crap its £222 !
That's bloody ridiculous.
I don't care how good it is, that's sheer daylight robbery !
No wonder they're helpful :-)
I think I once spent just over £100 on a motherboard for a clients gaming pc (£500 on graphics cards) - but never before or since.
What justifies that kind of money ?
More importantly, has brian printed anything yet ???????
I have a gaming PC i built for myself that i spent just over $4k building. The Mainboard for that was 4 bills. Asus Maximus X Hero Wifi. But that machine is absolutely radical. I used 3d printing to help make it. I spent over a grand just building the water cooling system. The fully modded D5 pump that can build up to 53psi cost $200 on it's own. I was empowered by monies made off of crypto currencies. I built that right here: https://www.superchargerforums.com/t...-machine.7445/
I have 2 Duet boards, Duet2Wifi and Duet Maestro. Both use RepRapFirmware https://github.com/dc42/RepRapFirmware and I don't think Marlin is available for them. But the hw is very good quality. Also Duet is very flexible in sw since everything is configured with gcode and macros so no need to compile fw.
There is some discussion about MMU2 with Duet boards here: https://forum.duet3d.com/topic/6155/...-s-mmu-v2-duet and it seems someone has got MMU2 working, at least somewhat (did not read all of the thread...).
I think we've already established you're not normal where tghis kind of thing is concerned :-)
I'm pretty sure i could build something at least as fast for half the money.
pc's are all about balance - just throwing money at them, is not the way to go :-)
It is very pretty lol
what processor and drives did you go for ?
Intel core I7-8700k, Asus Maximux hero X wifi mobo (the mobo natively supports water cooling!), Samsung 1TB m.2 pro ssd, Nvidia gtx 1080ti, 4x8(32)gb rgb g.skill ddr4 ram, and it is driven with a 1000 watt power supply. And the video card is connected to a curved 27" Asus ROG G-sync gaming monitor with 1ms response time and 2560x1440 res @ 165hz. That monitor is not included in the cost of building the rig. But it did cost a chunk on it's own. But holy cow, it'll bring a tear to your eye just to look at it.
Hey so with 5 spools of filament and the filament buffer, this printer now has a rather large footprint..
https://i1050.photobucket.com/albums...Wiz002/172.jpg
So no ryzen processor then :-)
Oh well, lol
I haven't built an intel system since the first amd 486 processor came out, 30 years ago ? Something like that.
Faster, cheaper, more reliable and the new 7nm processors piddle on intel from a great height. Hell even the ryzen 3 benchmarks faster than most current i7's.
Intel are the apple of the processor market - they have a huge marketing campaign, but don't invest as much in manufacturing as amd/ati.
Anyway, so next you need a back frame to lift the spools up along the back of the printer, that'll reduce the footprint and reduce the filament drag on the spools.
At the time I built this machine AMD was getting into trouble for misadvertising the amount of cores in their processors and one day I got a check in the mail out of nowhere for my 8 core FX-8370 processor I bought a few years back. If that has changed then it is just the fluidity of the industry we are in. But so far as I always understood it Intel and Nvidia are where people went for the better quality and the folk who wanted to save money went with amd and radeon. Despite current benchmarks we should all wait and see what comes out next. Be it a better chip from intel or another class action lawsuit. One single product release does not undo decades of back and forth and Intel coming out on top. That is just us getting caught up in the moment I think. But time will tell. For now, why not go read this 411 about my paycheck from nowhere: https://www.theverge.com/circuitbrea...ulldozer-chips
I most definitely need to find a way to shrink this down. The trick is keeping the filament buffer and the MMU2S moving with those tubes up and down 330mm. There is a better solution for this. I like your idea of a back frame.
So I went to GitHub and got the latest firmware for the new Bigtreetech TFT35 v3.0 and I must say I am impressed. I can enter Gcode directly from the small touchscreen.. https://i1050.photobucket.com/albums...Wiz003/173.jpg https://i1050.photobucket.com/albums...Wiz003/174.jpg
And there are menus for auto leveling now, too.. https://i1050.photobucket.com/albums...Wiz003/175.jpg https://i1050.photobucket.com/albums...Wiz003/176.jpg
Here is the home screen on my new multi material rig..
https://i1050.photobucket.com/albums...Wiz004/177.jpg
that's pretty cool.
Seems like the screens are doing almost as much as the motherboards these days.
I thought the mumu just fed the extruder on the printer - didn't realise it had to move around as well !
But you 'could' build a simple up and down frame seperate from the printer, though if you have a direct drive extruder on the printer itself - why does the mumu need to move about ?
As far as processors go - both intel and amd do the cores with double threads thing.
As far as performance goes - amd have always been ahead.
What intel do is they pay the big manufacturers a lot of money to use their processors. At one point if you used intel cpus intel were paying 50% of all your advertising costs that mentioned 'intel inside'.
To be fair I have built a few intel machines recently actually.
Got some unused mini-itx boards that had been removed from unsold computers. So all brand new kit, but second hand.
The board, 4gb of ram and an i3 cpu were £25 delivered :-)
The boards are mini network server boards and when new - the board alone - would have been in excess of £100.
So the 'server' that runs my k40 laser is made from one and my current workshop disk cloning, data recovery machine, also uses one and I built a cheap base unit for a client with one as well.
Since amd stopped making the am1 cpus - my previous source of cheap mini-itx systems - this is the best deal I've come across.
But over the years I've seen and benchmarked thousands of machines and amd always come out best both for speed and reliability.
Ati cards do tend to be faster than nvidia and run cooler - but the nvidia software has always been a lot more stable. So On average for good graphics cards I generally go for nvidia.
But for normal useage the ati on chip graphics is much better than the intel on motherboard chipsets.
I've always found it interesting that amd have never advertised - well at least not that I've ever seen :-)
I am past the software and basic calibrations of my machine. I am now working on reworking the filament buffer. I want mine to be clear so I can see the filament. But also I need mine to grow a bit as my ptfe tube from the mmu2s to the extruder is still longer than on the Prusa machine. Also I want mine to be easier to load filament into. My filament buffer will come out something like this: https://www.thingiverse.com/thing:3681588/files