PDA

View Full Version : Tank3d printer build



Boyan Silyavski
03-12-2016, 01:28 AM
Hi there,

Isn't it a high time to start and document a proper 3d printer build

These days the 3d printing bug bit me, so myself and my friend Jeff bought ourselves a Prusa MK2 printer. Not that i couldn't have made it myself but actually we both have some things ready to made so i did not want to waste time, plus i wanted to learn directly on a good enough printer and start playing directly with 4 colors and so on. So the Original Prusa obviously was the winner due not only to the innovative features, price, but when i analysed the design i saw and like its refined simplicity.

As its on a back order and i still dont have it in my hands i borrowed from another friend his Leon3d/ Spanish copy/ and these days i am starting to learn about printing. Frankly if you know me from the forum, you could guess that flimsy designs are not my cup of tea, so instantly i started to design a new printer.

I looked many designs from the web, and frankly i would have followed one, but there was no a single design which lived to my standard. I would say to "our " standard , as this forum is obviously the best for CNC design on the net :toot:, which is thanks to you guys!


Also looked at Chinese printers, copies of Prusa and other ones. Honestly for the price there were some that were tempting and then to fix the issues but the deeper i looked the more was disgusted by poor design. Arent these people even able to copy by the letter a working open source design? If somebody is interested just ask me about any Chinese printer and i will tell you its flaws on first glance. Not worth the money. Do not be tempted to save 300$ if you could permit it.


Lets say that i have the ambition to design the best truly universal semi pro printer for the money. Then simplify that design and make it so that even a child can put it together without any special tools or knowledge. That means many things had to be considered / thanks god for the multi core nuclear processor in my head/, needless to say last 2 weeks analysed all available on market. Lets say that most of the people making printers have no bloody idea of straightness, precision, speed, stiffness,repeat ability. And if some do, then the designs have so many mistakes, complexities or little details that spoil the deal, that honestly i stopped looking at any at all.

For the moment i did not dig too much into the electronics, and as i believe that guy Prusa knows his job and innovated some stuff, his printers work, are open source, so i will stick with his ideas about which board, firmware, etc. and focus on the mechanical and $$ side of things. With the only note that i am inclining towards servo motors or at least closed loop steppers. All will be Nema17. The reason is the long prints and very fine prints. But lets forget that for now and focus on the mechanical side.



One very important thing to note is that when i design something i prefer to start from scratch, then sketch the ideas, play with all possibilities and slowly purify the final idea. I have already done that. results, how and why- bellow; I will underline the main considerations.

Here is the plan:

1. Code name - Tank3d. First- it will look like a tank, second it will be strong like a tank, third- have you seen a fish tank :cat:

First i started with making the Prusa from aluminum profile. I found a nice printer designed by a guy and an aluminum version of the Prusa MK2
If you are lazy and want to stop here, there are the 2 designs i liked :
-http://www.reprapwilson3d.com/collections/frontpage/products/wilson-ii-complete-kit?variant=18734558981
-http://reprap4u.cz/navod-na-stavbu-3d-tiskarny-rebelix/


Then i said to my self: but it needs a double side beams for the gantry so its more rigid, does not vibrate and you could lift it and move it without need to adjust it again Then it looked like an almost finished half aquarium. So i said to my self- aluminum profile is not so expensive so better for 20 -30 euros more i make it so that if desired sheets to be bolted to profile and printer enclosed without any further modification or extra box needed.

2. The bearings.
8mm unsupported rods do the job if bearings are quality and shafts are hardened. The printer i borrowed is around 700euro and its rods are not hardened so they are scratched already, and my friend has that from a couple of months:beaten: On the other side 10mm bearings and rods could be better. Also all this talk about acceleration and printing speed. This printer here starts vibrating after 40mm/sec which according to me is the top speed that should be used. We will see the original Prusa, but what i am saying here is that 3d printing guys obviously dont know a s%^^t about rigidity and the only thing that saves them is not using the printer to make money or staff that must be precise.

Not the case here. I intend to make money using that printer selling prints , especially after spending any money on it, and i want micro detail precision and repeatability. Obviously supported rails Hiwin. I have Zero doubts here.

It was interesting to decide which Hiwin exactly the 9, 12 or 15 series. I am a big fan of the MGW series which i believe is the best for laser and plasma machines using single blocks

19769

As you see above at the snip, rail width and especially weight were main consideration as obviously 9, 12, 15 both normal or wide will work . But i believe 12H block and rail 12 size will work best as a compromise between block length, block weight, rail weight for printers up to 500x500x500 cubes.

3. Aluminum profile - 20x20

why:
-price 1/2 than 30x30
-enclosed structure
-rails size and especially the bearing block size is similar

Aluminum profile and linear bearing size and type choice was also influenced by the fact that this printer will be able to move at servo driven speeds and the Bowden type choice.


4. Bowden extruder.

I agree with mr. Prusa that this is the way. No double heads,increased fast speeds due to low weight as no need to move a motor, multi color easy, not unnecessary complications adjusting the 2 or more nozzles height and position, etc... We are almost year 2017. So at least 2 or 4 filament ability at the same time. many models will print better with support that can dissolve, especially if you sell them.


5. Self replicating.

In reprap tradition i decided to design it so that parts could be printed or milled from aluminum, without any further change to their design and withstand the scrutiny even of a hard OTT like myself, so thats a bit of a challenge here. Mostly cause i am still getting to know the plastics, but i printed here first parts today and i am quite happy. Needless to say - nothing flimsy or dubious.

Obviously once i finish the design i will open source it, but that does not mean that anyone can change it and achieve sth mediocre instead of a truly professional printer. Time will show if i will make it so good that the design itself discourage any changes as an unnecessary waste of time.

6. All parts must be self aligning to perpendicularity and maintaining same distances without the need of special tools when mounting. At most 1 precision square to be needed.
check pictures bellow.

7. All must be inside the envelope of the frame, no motors outside, etc,. So if one desires to enclose it by himself very easily, and keep warmth, noise, etc. there. Obviously the electronics have to be placed wiselt, so i think for now to be under the bed, isolated from the heat of the enclosure.

8. Bells to be tightened to specifications easily, means each belt has to have screw tensioner. But hey, not tlike the ones i have seen, that are so flimsy and move the belt around when reaching speed.

9. machine will be fully parametric, so design will not change at all up to 600x600x600 cube. Just lengths.


10. last but not least: the type of design. After much consideration Core XY designs, moving Z table and so on. Let me tell you. The simplest is the best for me. Moving table, double gantry z . I am quite happy to argue and explain why. If you have questions. But its a fact and no changes.

11. Work area size 400x200x(200-300-400)

I was tempted to make sth enormous or sth miniature. After a lot of thought to me 400mm width and 200mm depth , is the best compromise. Small enough to make everuthing precise so ultra miniature models using 0.2 nozzle and even smaller, big enough to use 0.8mm and even the Volcano nozzle.

12. Hot end

no doubts here, the E3d and the Volcano. When i have a printed 100kg of various filament, then will think more in this direction

13. which Bowden extruders??
unknown for now, to be decided.

14. Bed

Aluminum 6mm thick at least. more to come


Thats all for now, i hope you enjoy the build and forgive me for the bragging :beer:



So here are some picture of what i have done till now, ignore rail lengths , that will still have to be adjusted so all to fit inside ,all is still in a sketch phase:



19770 19771 19772 19773 19774 19775

komatias
03-12-2016, 11:24 AM
Boyan,

there is only one extruder worth using. It is also worth going direct drive on such a large build.

The Bondtech extruders are E3D compatible so at least that is a good start.

https://emvioeng.com/shop/3d-printing/3d-printer-components/bondtech-qr-3-0/

You will need to heat the bed to allow adhesion and reduce warping. Think about how you are going to acheive this on a 6mm surface.

The rest seems legit

looking forward to seeing what you get up to.

Boyan Silyavski
03-12-2016, 05:15 PM
Boyan,

there is only one extruder worth using. It is also worth going direct drive on such a large build.

The Bondtech extruders are E3D compatible so at least that is a good start.

https://emvioeng.com/shop/3d-printing/3d-printer-components/bondtech-qr-3-0/



Thanks for pointing me there! I will probably contact you as i need some quality components.


I will greatly appreciate your input if you are kind enough to share info.


But about the extruder:

-First of all its a proprietary extruder, not open source, so no place in my build. Why? Obviously to make my printer i will use open source electronics, hotend, etc. So roughly said- i am basing on other peoples work half of the printer. Isn't it a better thing to buy their products to sponsor them continue the investigation? I think so.

-Even if its the best extruder on the market and probably very nice to resale as a high end extruder with a good mark up value, i dont think thats the correct extruder for the build. Though for special order build or for a desired upgrade i could eventually use it.

-Third but not least- See, i will ask you a question. if i am to make a printer for resale, could you resale more easily that printer it it has 4 x Bondtech or a printer that has 4xextruders that do the same job 95% and are altogether combined cheaper than one of the Bondtech?






You will need to heat the bed to allow adhesion and reduce warping. Think about how you are going to acheive this on a 6mm surface.

400x200 bed. I was tempted to go 400x400 or even 500x500, but i decided that quality is better than quantity. First of all 2x 200x200 heating pads or elements will do the job cheaply. Then silicone heating pad may be . What i am contemplating is to use even 8-10mm plate, pocketed seriously from bottom and using heating wire directly or silicone pad there.

One thing is a must: the bed must not "breathe" and when pulling sth from bed machine must not need recalibration In fact- i will not accept to say that i did sth good if the machine will need any kind of later software adjustment and squaring of bed. Thats why i did not like the "raising Z design and shied away from it though people praise it. No way, its not my direction of design and thought. Even if i have to move the bed with 2 motors or a real servo motor to achieve the needed acceleration.

I have 3d printer from 3 days at home and have not seen or contemplated 3d printer design earlier. And the first thing that bugged me was pulling away the piece from the glass. I thought i will brake the bloody toy.

I could and possibly will use the way the Prusa does it with the touch probe, but better design everything so that is not actually needed.

PS:

bed heat insulation from below is a must also! and i will not shy from thermocontroller if need be. its just ~20 euros altogether with SSR relay.




Questions:

-If the printer is fully enclosed, do i need to cool the things down inside the enclosure somehow? I am not talking about the electronics. But do the electronics support that function? or just leave a hole so air circulates?



-Another question as i have 30x30 profile here and have never seen in real life 20x20 profile. Will it be ok for the gantry only one piece of 20x20 at high speeds? will it not vibrate? Thats my only doubt from structural point of view.


-haven't somebody come with super light direct drive helper combined with Bowden extruder solution? In real life i see the 4 spols in a dry box 1.5 meter away at least from printer. I dont see them hanging around and sucking humidity from air, and i live 200 meters from sea and this is a serious problem.

In fact from a couple of days i dont stop the air conditioner at all at home, because i still have not made myself a dry box for the filament

if not- too bad but anyway i dont care much about flexible filament, so Bowden drive will be

nirmal
03-12-2016, 08:45 PM
There is a very neat extruder design on this thread. https://forum.e3d-online.com/index.php?threads/e3d-water-cooled-mod.53/The water cooled hot end design on the same thread might also be a good idea if you find that a heated chamber is necessary to keep super sized prints from warping.

You can order custom made silicone pad heaters from china in whatever size you require. I presume you will go for 240V AC and a solid state relay for the bed heater. I am not sure if any of the commonly used controllers will allow the sort of currents you will need to pass through them without overheating if you go for DC and on-board PWM heated bed control.

On the subject of control boards, the Duet WiFi with the TMC2660 stepper drivers would allow you greater a choice of stepper motors with it's high current rating and as an added bonus would be very quiet in operation. I have no connection with the Duet manufacturers. I am just the owner of a previous version of the board - purchased for a still unfinished project.

nirmal
03-12-2016, 08:47 PM
Sorry I messed up the link. Here it is https://forum.e3d-online.com/index.php?threads/e3d-water-cooled-mod.53/

magicniner
03-12-2016, 09:02 PM
I am not sure if any of the commonly used controllers will allow the sort of currents you will need to pass through them without overheating if you go for DC and on-board PWM heated bed control.

It's easy to do your control on the AC side of a transformer/rectifier circuit using a PID controller and solid state relay ;-)

Robin Hewitt
03-12-2016, 10:28 PM
WOW! I got my first 3D printer for my birthday back in 2012, wore it out by 2015 and I bought another.
I have absolutely no advice to offer because I believe 3D filament extrusion printing to be a black art and beyond earthly comprehension. Every time I think I understand it a new torment pops up and hurls me back to square one.
I would say, "Good luck" but I do not want to jinx you.
Break a leg :peaceful:
Robin

Doddy
05-12-2016, 04:12 PM
Similar to Robin - had a couple of 3D printers for a few years now. Just a point to Boyan - remember the nature of the material that you're dealing with here - it's a laudable design goal to have a tank-build, but there *WILL* be deformation of the extruded material that WILL cause head crashes - I've been watching mine recently and some of the flex in the design accommodates the fast traverses together with a clearance height that is intruded upon by the cast material. A bit of flexibility could offer protection to the extruder.

EDIT: (tongue-in-cheek) maybe in keeping with the nature of this site, Boyan should design a co-axially mounted fly-cutter to machine the build surface!

Robin Hewitt
05-12-2016, 04:42 PM
Boyan should design a co-axially mounted fly-cutter to machine the build surface!

and a 300degC air blower to soften the top layer before deposition should you dare to print any substantial area :distrust:

I'd also like a micrometer adjust on the Z axis so I can tweak it without reprinting. A mere 0.1mm too low on the first layer means no squidge and I lose adhesion :fat:

I prime the bed using an ABS wash over Kaptan over Pyrex over aluminium, with yellow dusters on top for a quick warm up. What a palaver. No wonder ebay is full of patent sticky beds :stung:

I used to use cheap ebay filament but now only the best will do, and small reels because you only wind up throwing the excess the moment it has sucked enough moisture out of the air. Pop fizz :culpability:

Neale
05-12-2016, 05:00 PM
I'd also like a micrometer adjust on the Z axis so I can tweak it without reprinting. A mere 0.1mm too low on the first layer means no squidge and I lose adhesion :fat:

Agreed - one of the first things I printed with mine (early Prusa now around 5 years old) was a fine Z height adjuster. Invaluable. I would also like auto bed levelling - I take my printer to exhibitions from time to time (good machine for this as you can see all the working parts very easily) and it doesn't travel too well. Always needs 10-15 mins of fiddling around to retweak the bed (heat-proof glass on PC board heater). I managed to find a ground glass trivet in a local discount hardware place which gives very good adhesion with no more effort than the occasional wipe with acetone.

Robin Hewitt
05-12-2016, 05:34 PM
I managed to find a ground glass trivet in a local discount hardware place which gives very good adhesion with no more effort than the occasional wipe with acetone.

I keep wondering about frosting it with a waft of hydrofluoric acid, trouble is HF really does everything H&S warns you about for everything else. It really does kill you and all the hospital can do is make you a cup of tea and contact your next of kin. Back in my youth the chemistry books told you how to frost glass by dissolving old teeth :tennis:

Boyan Silyavski
05-12-2016, 05:35 PM
Dont know, am i lucky or what? had no idea whatsoever before 1 month , decided a 3d printer before 2 weeks, 3-4 days to learn software and from one week ago i am non stop printing on my friends Legio / a Spanish knockoff of Prusa 3/

He told me he needs hours if at all to level the bed and that the machine is super flimsy/ it is/ . Last week he gave me the machine. I did not know how to operate it at all.

First thing i tightened properly the belts. Took from garage the hot glue gun and a precision square. Hot glued all that i did not like / ALL :hysterical:/ . Hot glued machine to a kitchen board/ not best but i am going to weld a base/ . Hot glued the belt ends to the belt plates. As i said -all.

Then leveled the bloody bed for 2 min and it would have been faster if people were aware that 3 screws leveling is easier than 4 screws leveling.

Did some prints and 1 print f%%%d up cause unglued from bed. So i sat down infront of google and after 10 min i knew the magic. Went to garage, took a bottle of C200/ white wood glue. diluted it 1:1 in a small hermetic jar and tapping gently a sponge primed the glass bed. From then on glues like a ...you know the saying. Only thing is i have to wait 5 min for the bed to cool a bit from the 55C i print. Then it comes off easily. I see the hotend or the hotend extruder combo is not perfect, but i print always from then with rafts and skirt so no problem at all with the crappy breathing bed.

I printed PLA and from a couple of days only pla 3d850 which i intend to be the main material as its stronger than ABS when cured, very detailed prints, and hi temp resistant when cured. Its Ingeo but made in Spain so at 19euro per kg is a no brainer that i master it.

The only problem till now was to master the material which i did in a day, had problems and it made me learn Simplify3D up and down , at the end it was bloody material being 1.69mm thick, but when logic was implemented it took me only 2 hours to find the bug.



What i am saying is PVA glue is the way. Rafts also. No need to hurry and try to save 5 min or 0.05 cents of material. No need for special stuff. Or to use my wifes hair spray.



Ok, i dont know if you guys are aware but on a properly designed DIY CNC when head hits the table or gantry hits the end or similar, if steppers are used the machine just stops and thats it. Plus if you use digital drives most probably stall detection will happen and alarm signal will stop the machine.

So if the printer is properly designed i dont see what will happen apart from stopping.

I am still not there and will ook into that but: Arent these printers using some digital drivers with stall detection?????

I am much more worried why these toys dont use some kind of smoke detection or fire prevention??


Anyway, i moved further with the design. Managed to make the gantry very tight vertically, same with the hotend. Plus at the end i decided how what to do with the X motor, which will go in the clearance between the gantry sides and the table, which means keeping the structure height etremely low. That combined with the shallow depth will make it a real table top machine. Not some ugly cube. At the moment stands wide 620mm and high 370mm. I am still playing with the idea if i need or not the top and bottom cross bars, look last picture where removed. Looks more simple and better to the eye, I will decide that later . Their main idea was so if i sell the machine to sb all linear elements to be premounted here by me, Guess i have to receive the 20x20 profile and see how it will go. Their other function is if you fit servo drives and do crazy speeds so the machine to wobble. Though it wouldn't if side and top panels are fitted


Another thing that bugs me is the cable chain. Yesterday printed some cable chains and they are not properly designed for printing even if they tell so. I will check what price are Chines, design proper one or use what they use on the Original Prusa. Some kind of cable tie organizer???


have also started printing the parts and well, i am quite happy with the result. Till now had to redesign only one of the, All else is indestructible by hand. Which is how i like it to be strong. The hotend plastics are downloaded from internet. Just printing them now, cause i think i will have to redesign them to be stronger and ventilator adjustable somehow up and down in case i switch to the Volcano heater



19815 19816 19817 19818

Neale
05-12-2016, 06:13 PM
I keep wondering about frosting it with a waft of hydrofluoric acid, trouble is HF really does everything H&S warns you about for everything else. It really does kill you and all the hospital can do is make you a cup of tea and contact your next of kin. Back in my youth the chemistry books told you how to frost glass by dissolving old teeth :tennis:

I remember the school chemistry book with a picture of a hand that had been dipped in HF. I treated the bottle of HF with care after that; doubt if it would be allowed within miles of a school these days!

A friend of mind is a sculptor/artist working with glass and she tells me that sandblasting is the way to go. Certainly with PLA, it works well, and as Boyan says with PVA glue, you need to let the glass cool for some time before the print comes free. I typically print with a bed temp of 60C, but of course that depends on how accurate the bed temp measurement is. At least it's a lot less critical than hot-end temp.

nirmal
05-12-2016, 09:28 PM
I used to use cheap ebay filament but now only the best will do, and small reels because you only wind up throwing the excess the moment it has sucked enough moisture out of the air. Pop fizz :culpability:

I store filament in airtight clear plastic paint buckets with a rechargeable portable dehumidifier and a cheap digital hygrometer. It won't miraculously make poor quality filament good but it keeps it dry enough to use the whole reel for very little expense.

Robin Hewitt
06-12-2016, 10:39 AM
He told me he needs hours if at all to level the bed and that the machine is super flimsy/ it is/ .

I bought the UP! Plus because it has a robust folded steel base and ball slides throughout. If it accidentally stands on something it rocks, it does not deform. Minimal faffing about.

Chaz
06-12-2016, 03:15 PM
We used to do the DIY 3D printing kits a few years ago. We did the MendelMax design, then did our own design. Played with Rostock too. I was thinking to look at this again, now that I know more about cnc and have better tools available.

The extruder for me is the key thing to get right, then the bed / heating setup. The movement (XYZ) is almost the easy part.

The Ultimaker 3 shows what can be done in the 'DIY' market but I cant see the value in the cost of that.

Boyan Silyavski
06-12-2016, 08:27 PM
We used to do the DIY 3D printing kits a few years ago. We did the MendelMax design, then did our own design. Played with Rostock too. I was thinking to look at this again, now that I know more about cnc and have better tools available.

The extruder for me is the key thing to get right, then the bed / heating setup. The movement (XYZ) is almost the easy part.

The Ultimaker 3 shows what can be done in the 'DIY' market but I cant see the value in the cost of that.



The bed is easy. The extruder/hotend is difficult as needs lot of experiments and tiny adjustments. Thats why most go with established combos. You say the linear motion and frame is easy, but then where are the good designs? I dont see any, and if i see some they are >1500euro

There is growing market even with the cheap Chinese printers. Many people will buy one only to upgrade at a later time when they find they can not make money with it.

I did not enter earlier as they seemed quite under developed and funny. Material was expensive, hot ends were ugly. In other words- unreliable expensive toy with no real world purpose / at least to me/.



I think-- how to combine all this is the key and offer sth reliable, well supported and most of all easy. Even now i find all not so easy. Easy to me after designing a couple of machines and building them and being in the CNC world a couple of years. But things still are not "push and play", i would say far from it.

Right now i am into a couple of Facebook groups where printer guys speak, i see many real life problems with the machines priced 1000-3000$, that are ridiculous especially for the lack of support and warranty. To me its speaks: not well designed robust product.

Chaz
06-12-2016, 09:14 PM
The bed is easy. The extruder/hotend is difficult as needs lot of experiments and tiny adjustments. Thats why most go with established combos. You say the linear motion and frame is easy, but then where are the good designs? I dont see any, and if i see some they are >1500euro

There is growing market even with the cheap Chinese printers. Many people will buy one only to upgrade at a later time when they find they can not make money with it.

I did not enter earlier as they seemed quite under developed and funny. Material was expensive, hot ends were ugly. In other words- unreliable expensive toy with no real world purpose / at least to me/.



I think-- how to combine all this is the key and offer sth reliable, well supported and most of all easy. Even now i find all not so easy. Easy to me after designing a couple of machines and building them and being in the CNC world a couple of years. But things still are not "push and play", i would say far from it.

Right now i am into a couple of Facebook groups where printer guys speak, i see many real life problems with the machines priced 1000-3000$, that are ridiculous especially for the lack of support and warranty. To me its speaks: not well designed robust product.

I agree, its on my to do list to get something out. Time however is not on my side, too much to do.

Boyan Silyavski
07-12-2016, 12:25 AM
Today was a busy day, but while resting managed to finish printing a couple of popular cable guide chains and found that not one is working properly. Mehh, i was not into designing a cable guide but it seemed that i had to. Luckily having the printer at hand i did some prototypes and at the end i have sth that i am very proud of. Its strong, impossible to break and to break apart using normal force by hand and most of all moves freely cause i left the necessary clearances. if i further compromise a mm or two and make it bigger here and there , i can make it even to be printed assembled, though dint see reason to do that for now as its good as it is.

I wanted a cable chain not that i could not use just the proper flex cable, but because i would lke to be easily scalable, that means all has to be perfect from the beginning. So later just scale and do.

About the bed. Any idea what heating wire i could use to achieve the proper 0.4w per cm2? 12VDC , 24VDC or even better 230VAC ????? I would like to route the aluminum with channels and insert the wire from below. Certanly sb will know a silicone wire or similar that could do the job, instead of the ugly silicone pads?


19848 19849 19850 19851 19852




in case anybody is wondering that is the proper way to print a cable chain links, i tried all possible today and thats how they are printed strongest:

19853


Ok to be more exact like that:

19854



By the way i would like to remind you that this will be open source so if sb needs some CAD, meanwhile i am finishing the drawing just drop me a line. I draw in Nx and can send you solids.

magicniner
07-12-2016, 07:20 PM
It would be nice if the hot end could deal with printable wax, speaking as a man with a furnace or two ;-)

Boyan Silyavski
07-12-2016, 08:45 PM
It would be nice if the hot end could deal with printable wax, speaking as a man with a furnace or two ;-)

Of course, speaking as a man also with a furnace or two :-)

But isn't that possible with the e3d hotend? I thought of using wax PLA filament with my neew printer, so when heated in furnace is similar to wax and dissapears. Or yo mean extruder head where wax is melted and injected into nozzle? I think that bearing will support similar setup plus there is almost none overhang if you look at pictures from side

Robin Hewitt
07-12-2016, 09:24 PM
By the way i would like to remind you that this will be open source so if sb needs some CAD, meanwhile i am finishing the drawing just drop me a line[/B]. I draw in Nx and can send you solids.

Can you send me this as a .stl file
I don't need cable chain, I am just curious if my prints come out the same as yours.

Boyan Silyavski
07-12-2016, 10:04 PM
Can you send me this as a .stl file
I don't need cable chain, I am just curious if my prints come out the same as yours.


Here are how parts come from the printer i use now, meanwhile waiting for the Prusa MK2. I use PlA 3d850, setting same as normal PLA but 230C instead of 200C, 50% infill, 0.4mm nozzle, 40mm/sec as the printer is very wobbly. Outside rectilinear. Parts are very strong.

19914 19915 19916 19917 19918 19919


Here are the files that are most possibly final for now. Note that the link still needs work, for now i heat with hot water when mounting together, then use pliers to straighten and when cold works perfectly.


19922 19921 19920 19923

Robin Hewitt
07-12-2016, 10:25 PM
I am halfway through a print, I will do a couple of links when it finishes :eagerness:

Boyan Silyavski
07-12-2016, 10:33 PM
I am halfway through a print, I will do a couple of links when it finishes :eagerness:

Check that one too. I just finished adjusting the link pins, as they were too long i think and as i said assembly had to be done under hot water. And i broke some mean while . Due to the newly introduced slot which was supposed to help mounting it together.


I am just going to print that new variant and check if i resolved the issue.19929 otherwise it moves very nice and has bend radius of mere 55mm, chain is 25mm wide x 20 high outside

magicniner
07-12-2016, 11:35 PM
Of course, speaking as a man also with a furnace or two :-)

But isn't that possible with the e3d hotend? I thought of using wax PLA filament

I was thinking of this -

http://www.machinablewax.com/product.php?product=52

- Nick

Robin Hewitt
07-12-2016, 11:45 PM
Too late for updates, 40 minutes to go :wink:

19930

Boyan Silyavski
08-12-2016, 01:07 AM
I was thinking of this -

http://www.machinablewax.com/product.php?product=52

- Nick

Exactly. On my to buy list. Anything similar in EU? Or a supplier here?


Too late for updates, 40 minutes to go :wink:

19930

NIce. The update did not work well. To loose. Its a challenging little thing this link, either is too strong , either too weak. It breaks easily or is next to impossible to assemble. I manage heating it but sb else could find that impossible. So still not 100% right . Though works quite fine once assembled and if printed from nylon or a just a bit flexible plastic that will not be a problem.

Boyan Silyavski
11-12-2016, 02:50 AM
At the end after a couple of experiments managed to finish 100% the design of the chain. So now clicks easily, moves right, prints right and most of all attaches link by link flawlessly without any breakage. Anyways i used again hot water to click them, this time all went very nice.

Meanwhile managed to find realistic drawing of the 8x8x2 / 4 start screw in step, which was not perfect and after a lot of fiddling have the nuts 8mm 4 start and 2 start x 2mm pitch in 100 realistic drawing read to print. Very important aspect of all this project is that parts are 100% realistic represented.

Took me a whole day but managed to send a file from Openscad to FreeCAD and from there to ifus, import the Igus in NX11, sew the surfaces while importing and finish with an usable for customization solid model part.Phew... Nuts printed well as seen on picture. Also the chain.

19943 19945 19944




Finished drawing the Z motors and the mounts. Mounts are designed so that will space precisely the motors in all directions


19946


Table design include 2x motors and belts..etc. This is due to the whole idea behind that model to be able to scale bigger easily. Without any actual change of design. Apart from lengths of elements. So to be truly parametric design. Plus motors are cheap, Z plate will be cast and trully from what i see / the printer i am testing/ this is where all designs i have seen suffer. I dont see what is the big deal to include second motor and for mere 20-30 euro more to have no problem with acceleration of the plate. There are cheap signal multiplexers? that make one signal for 2 motors / if board will not support it/, etc.

Anyway most possibly my printer will be at least with closed loop if not with servos. maybe i will build one to see how it goes with cheap motors first.

19947 19948


Now the big question for me is the table plate.

i know i will be cast aluminum >6mm. I made it bigger so it accepts 2 cheap 200x200 heated beds. I will see if i can make it as i said using heated cable or resistors or sth similar. I will isolate it from bellow using special paper and then sheet.

But what small plates i use to connect the heated 100C plate to the bearing blocks???? Rigid connection, no fiddling with adjustable screws and springs? Something that will stand to 130 degree, that will not change dimension, so i make a precise connection. And most importantly willnot transmit heat to the bearing blocks.

Not that the bearing blocks can not be ordered for a couple of dollars to be heat resistant / bronze instead of plastic/ but i would like to be usable with normal bearing blocks, not special. Any ideas of materials? have to be easily obtainable and DIY at least by me.


The whole model starts to look deceptively simple but it was a lot of thought and different approaches tried, drawn and then discarded.

Boyan Silyavski
18-12-2016, 01:06 AM
Now some electricity quiz. I need some advice from sb not electrically challenged like me. As you know I am wondering how to make the printer bed plate as simplest as possible, rigid and cheap.

After examining many options i came to the conclusion that could be achieved using 6-8mm aluminum plate, routed from bottom with cable channels and insulated heating cable inserted.

230V, plate connected to Earth, thermal fuse for overheating, current fuse for over current, thermistor integrated, that for safety. Plus SSr and probably Pid controller or just using the boards PID and separate board for the thermocouple signal if i go this way.



So according to my research ~500W will be right for the power. I used this simple Ohm’s Law calculator (http://www.pelicanwire.com/resistance-wire-formulas-and-conversions/heating-element-calculations-ohms-law/)

I looked around and found the so called Minco Heat 12K Carbon Fiber Heating Cable 33 Ohm/m DIY Underfloor Heating Wires at ebay and ali express.
It sells around 15euro per 20 meters incl shipping which translates into less than 1 euro per meter

http://www.ebay.es/itm/10m-Infrared-Heating-Cable-2mm-Electric-FF46-Teflon-Carbon-Fiber-Wire-12k-33ohm-/291928479851?var=&hash=item43f84b186b:m:mfxtjRHzwqQq-NrOxSvmuIw

here is the description:

Product Name: Carbon fiber heating wire
Insulating materials: Silicone rubber
Temperature: 200 degrees celsius limit
Heating conductor: 12K
High voltage test: 3000V
Leakage current: 0.05mA/m
The product can withstand the power of power: 25W/M
Diameter: 30.2 MM
Optional colors: Red
Infrared wavelength: 8UM-18UM
Conductor resistance: 33 Ω / M
Length: (custom, not less than 10 meters)

PS: Rated voltage is ~300V


Now from the calculator it seems at 230Vac to achieve ~500w i need a wire with resistance around 100 Ohms more or less :

20023


Please correct me bellow if i am wrong before i burn sth:

1.Now as i understand it the wire resists temperature 200C. Obviously if use 3 meters of wire at 33 ohm per meter i will achieve the desired result.

2.They say :"The product can withstand the power of power: 25W/M" but thats for floor heating, so i assume if wire is connected to a temperature regulating device which does not permit it to go above 130C /+ the thermal and current fuses/ there will be no any problem????

3. I understand that also i have to push fit the wire in the channel/ 3mm/ so there is not any air gap between plate and wire which could cause overheating due to incorrect temp measurement.

I plan of: 6/8/mm plate=+ channel+tight fit wire+thermo conductive silicone flush with bottom surface. So no air there art all

Apart from that i will use some insulator sheet from bellow, still not clear but most possibly Teflon, Silicone, Mica ???



What do you think?

I know i can buy 40 euro 500 W ready heater from China for 40-50 euro, stick it to the aluminum bed from below and thats it. But that is expensive compared to what i propose, Chinese will not care to make trace separation properly and so on. I mean its expensive and i dont find it more reliable than sth i could make. Plus machining all will cost only around 10 euros with the plate, which is 1.3 kg

Calculation shows that 3000mm cable/400mm bed =7.5 which is roughly 7 channels on the long side. So ~200mm/7=~28mm separation between channels which seems quite right to me.

Doddy
18-12-2016, 09:27 AM
Please correct me bellow if i am wrong before i burn sth:

1.Now as i understand it the wire resists temperature 200C. Obviously if use 3 meters of wire at 33 ohm per meter i will achieve the desired result.

2.They say :"The product can withstand the power of power: 25W/M" but thats for floor heating, so i assume if wire is connected to a temperature regulating device which does not permit it to go above 130C /+ the thermal and current fuses/ there will be no any problem????

3. I understand that also i have to push fit the wire in the channel/ 3mm/ so there is not any air gap between plate and wire which could cause overheating due to incorrect temp measurement.

I plan of: 6/8/mm plate=+ channel+tight fit wire+thermo conductive silicone flush with bottom surface. So no air there art all

Apart from that i will use some insulator sheet from bellow, still not clear but most possibly Teflon, Silicone, Mica ???

What do you think?

I know i can buy 40 euro 500 W ready heater from China for 40-50 euro, stick it to the aluminum bed from below and thats it. But that is expensive compared to what i propose, Chinese will not care to make trace separation properly and so on. I mean its expensive and i dont find it more reliable than sth i could make. Plus machining all will cost only around 10 euros with the plate, which is 1.3 kg

Calculation shows that 3000mm cable/400mm bed =7.5 which is roughly 7 channels on the long side. So ~200mm/7=~28mm separation between channels which seems quite right to me.

Your figures are broadly correct, but there's a bunch of information missing from the cable spec - max current rating (you can guess at 0.8A given the 25W/M - but that's likely derated given the intended installation buried in a concrete scree) - to be honest, I'd just try it and see. Yes, you'll get a bit over 500W heating which should give a respectable amount of heat energy for the 1kg block of ali that you're using. Temperature control you've mentioned, and a thermal fuse (you mention) a necessity in this system; zero-crossing SSR also a good idea as well as a current protection device (fuse). All good stuff, but very little in the data on the cable to give support to your design simply because you're using it for a different purpose. Try it and see - my thoughts are that it should be workable.

My only reservations are: No indication of the thermal expansion of the cable - if you bed this too tightly in the ali block you could sheer the cable strands - I'd look to a loose fit and infill with thermal paste when you look to productionise.

Experiment, is the name of the game - you've nothing to lose (provided you keep a fire-extinguisher to hand :) ) In case the cable fails catastrophically I'd certainly use a test bed well away from anything flammable and be able to isolate the supply easily. Once you have your thermal control in place then test worst-case performance with it shoved in a cardboard box crammed full of thermal insulation (cloth sheets) to make sure that the control system still works.

The one area I'd be concerned about with this cable is making the ends off correctly - looks like you are supplied with copper ferrules to crimp - that to me is the area that you're likely to see failure if not done properly.


I'm not sure I'd agree that this is a cheap solution, though - in the UK I can source heated beds off ebay for 6 each (214mm x 214mm) and these will be being etched in their hundreds or thousands in china with the precision from decades of investment in PCB manufacture. Only 120W, but two in parallel for the 400mm bed and you're not a million miles from your solution.

Boyan Silyavski
18-12-2016, 01:32 PM
Doddy, many thanks!


Your figures are broadly correct, but there's a bunch of information missing from the cable spec - max current rating (you can guess at 0.8A given the 25W/M - but that's likely derated given the intended installation buried in a concrete scree) - to be honest, I'd just try it and see. Yes, you'll get a bit over 500W heating which should give a respectable amount of heat energy for the 1kg block of ali that you're using.

My research on similar setups shows that in real life 500W non insulated bed this size will heat very fast, and insulated will heat probably under a minute, so no worry as you say.



Temperature control you've mentioned, and a thermal fuse (you mention) a necessity in this system; zero-crossing SSR also a good idea as well as a current protection device (fuse). All good stuff, but very little in the data on the cable to give support to your design simply because you're using it for a different purpose. Try it and see - my thoughts are that it should be workable.

Good info for the zero crossing SSR, i did not know what that means. Learned something. So are the cheap Chinese SSR relays zero crossing? I have used them for my kiln but they just work, i did not have an idea that they are different types.



My only reservations are: No indication of the thermal expansion of the cable - if you bed this too tightly in the ali block you could sheer the cable strands - I'd look to a loose fit and infill with thermal paste when you look to productionise.

I chose the carbon fiber cable exactly for that reason that the carbon fibre itself is very stable and does not expand or shrink. At least that is what they brag about it. But will look into further info about the whole cable.




Experiment, is the name of the game - you've nothing to lose (provided you keep a fire-extinguisher to hand :) ) In case the cable fails catastrophically I'd certainly use a test bed well away from anything flammable and be able to isolate the supply easily. Once you have your thermal control in place then test worst-case performance with it shoved in a cardboard box crammed full of thermal insulation (cloth sheets) to make sure that the control system still works.

I will do exactly that.


The one area I'd be concerned about with this cable is making the ends off correctly - looks like you are supplied with copper ferrules to crimp - that to me is the area that you're likely to see failure if not done properly.

I briefly looked into that and saw how people connect carbon fibre tape into clothes. More or less they use some kind of silver glue. So i am thinking i have to buy some ferules and test, if needed fit inside some extra silver compound for better conductivity. One thing is clear that everything should ve additionally fixed mechanically to the bed, meaning pressed between plates or sth, and from bed only 3 kevlar teflon double shielded thermo resistat wires will protrude.


I'm not sure I'd agree that this is a cheap solution, though - in the UK I can source heated beds off ebay for 6 each (214mm x 214mm) and these will be being etched in their hundreds or thousands in china with the precision from decades of investment in PCB manufacture. Only 120W, but two in parallel for the 400mm bed and you're not a million miles from your solution.

I will explain why i like this solution below. Mind me that the actual bed in the drawing is 440x220mm, so it will take 2x MK2 heated beds 12/24VDC on top of it bolted, which 1 of these is 7 euro.

1.The overall design of the machine.
-The printer must be rigid, fast and able to Scale Up. From that the result will be precision, nice prints and easy to build a second one, which probably most of the people will do, once they have one and are serious into printing.

I meditated long on the various possible designs and chose this one due to the simplicity to build and later adjust, compared to all of the other designs. Some of them are so ridiculous honestly, that even i can not imagine what hassle will be to straighten it to some acceptable precision. If i that build a 3m machine to some standard, can not be sure how to put together precisely a 50cm machine, then imagine what the result will be for a person that has no idea at all.

When you start scaling the machine up, the so called Core designs start to loose in moving weight compared to this one, though i dont know why the guys are so confused and praise them so much.

Thats another story but suffice to say that from what i see scaling the machine up, if i do the custom bed then the bed weight will keep in line with a Core XY design moving mass weight very simple. 20x20 beam is 0.45kg per meter even a 1000x1000mm bed could be done from same 6mm precision aluminum cast sheet and a simple beam bolted from below if it starts deflecting from heat. At the same time Core XY will have very serious lengths belts which will start back slashing/ from stretch/ on high accelerations, or motors have to be moved, which is additional mass...


And anyway all other design suffer from the problem called twisting the machine out of precision when prying out a printed object from bed Or separate glass bed that is insulator, ... and so on.

2. The lack of additional under bed carriage that is connected to the bearing blocks on which the bed lies, or whatever they call it. If that thing is made from plastic it bends, if its from aluminum it is heavy. In fact the aim here is to go without it.

3. Eliminate adjusting the bed once its adjusted when building the machine.

This practice is absolutely ridiculous to me. From day one i saw a 3d printer. Imagine having to do that on your mill or router every time you start a job :hysterical:

4. Ability to bolt it and to make bolt holes from bellow where i like. You can not drill PCB heated bed. That will additionally provide support to bed against bending from temperature, though from reports 6mm bed is alright.

5. If you glue additional silicone plate which is the only viable solution for big beds for now, first you can not drill it-means additional plate from below, second as i said its heavy, expensive and not any safer or cheaper.

6. I found a nice 230V heated PCB project which also seems stack-able but it will suffer from need of additional glass on top, hence weight, low reliability, loss of precision surface etc.

7. I think an additional sheet of thicker PEI on top of the aluminum bed will allow things to be removable and exchangeable, without f"""g the precision of bed and machine. Glass i dont like for 2 reasons- first its an insulator, second i dont like one shiny surface and all else different. Anyways i learned to print 100% with rafts.

I am not in a hurry, i prefer the quality result.


Here bellow is the latest update on the design. At the end when all is clear, printer will look very simple and clean. Bed will be thermo insulated from bellow. Electronics and motors will be below. Heat flow will be up , so all is ok. After long search at the end i found some people sharing that when bed temp 100C in enclosure temp is around 45C, which is ok for the plastic parts and the motors, so i dont believe they will need additional cooling.

Meanwhile i looked most of the extruders and hotends. The way i see it- E3d6 plus E3d Titan extruder or i have to make sth similar, which i dont have the time to investigate. One company, great products. No super duper over complicated extruders that cost 200$. These days i printed 24h and many different things, and i found that the weakest place in the 3d printer is the extruder itself and the measurement of the heat in the block- hence the thermistor type, position and especially way of attaching it.


All else could be Chinese or mediocre and will work more or less, but not feeding reliably the filament and not knowing reliably the temperature of the heating block= bad prints, unreliable prints, messing with adjustments that is not needed to mess with if these 2 are working right.

And these 2 could make you crazy if they are not perfect.


20028 20029 20030

Robin Hewitt
18-12-2016, 03:34 PM
Totally unsuitable. You need an electric element or elements.

This was fun to CNC but probably overkill for your application :cool:

20033

Edit: Just ignore me, nothing works on this website, I can't see what you are typing, if I upload a picture the world goes away, famine and frustration.

Boyan Silyavski
19-12-2016, 08:37 PM
Some electrical help here please.

If the heated bed consuption will be 2.1A at 220V, and 500W, then what cutoff fuse should i choose? I am looking at 150C 2.5A and 150C 5A. Which one will be better? I assume the 5A? As it will cut off at 150C in both cases, i am not sure about the amps.

Doddy
19-12-2016, 08:51 PM
Some electrical help here please.

If the heated bed consuption will be 2.1A at 220V, and 500W, then what cutoff fuse should i choose? I am looking at 150C 2.5A and 150C 5A. Which one will be better? I assume the 5A? As it will cut off at 150C in both cases, i am not sure about the amps.

If this is a thermal cutoff that you're talking about then either is sufficient - either will handle the current draw, and disconnect. Of course, the over-current protection needs closer consideration; I'd not go as low as 2.1A, but 3A would be adequate (just make sure that all cabling and switchgear is rated at 3A).

Doddy
19-12-2016, 08:53 PM
...and if you want to know what pees me off about cheap 3D printers - I've just had to replace the x-axis cable form on a near-new QIDI; what an unnecessarily tedious faff! Boyan - get a decent energy chain in there and use silicone cable. The cable form should be designed never to need replacing.

Boyan Silyavski
19-12-2016, 11:07 PM
What i am looking is a thermal fuse, which apart the thermal protection, has Amp rating. So i thought before that i need 2 separate elements but it seems is 2 in 1. All cabling will be industrial servo cable rated for flex drag chain instalations. No crap stuff. I am just having also problems with the cables of the printer i borrowed.

here is link to the cheapest best quality cables i have found after long sleepless nights searches. These are not for the printer, i will look further, but just to see prices. have bought from there flexible thermoresistant cable for soldering gun also, which is superb for the price.
servos

http://www.tme.eu/en/details/str-c-p...lukabel/49637/


shielded cable for the proximity sensors:

http://www.tme.eu/es/details/str-c-p...lukabel/49632/





The cable chain is ready and tested. It had some final adjustments and i am 99% happy with it.. Its printed but i have designed it as a commercial quality chain. So its 2 in 1. What do you think? It seems to me bnetter than the one they use in the printer i have here





here are the files http://www.thingiverse.com/thing:1963745


https://www.youtube.com/watch?v=8EyVTwjgCQQ

Lure
17-03-2017, 06:55 PM
Is there any progress om this project? Should be really intresting to see the final product, considering your great skills.

Boyan Silyavski
18-03-2017, 09:00 PM
Is there any progress om this project? Should be really intresting to see the final product, considering your great skills.

Thanks,

Yes, there is. But as it seemed that there is no interest here about the printer, i kept silent a bit. Also i am very short of time and had to design each element including the hotend assembly, blower fan, etc. I have already uploaded some of the designs http://www.thingiverse.com/silyavski/designs
but still have not had the time for the separate assemblies.

In short, i am putting all together 2 printers, while meanwhile revising the drawings and populating screws, nuts etc. Very soon will be more info on all with pictures

All parts are here minus the controller, as i have still not clear which will be the controller.

for now the final design :

21154

Lure
18-03-2017, 09:22 PM
Looking nice. Can't wait to see the rest of the project. Hopefully there will be a detailed and educational build log like the others you have made.

Desertboy
29-04-2017, 12:17 PM
Excellent work, puts my 3d printer to shame but also overkill lol.

I guess you have commercial uses for this printer as seems extreme for a hobby machine.

Will be super interested in seeing the end product this is going to be something special.