The end ? maybe not

I've this blog to show that a 3d printer can be made by using relative easy to get parts and some ingenuity.This project took about 2 years working on my spare time to complete.The first thing i did was to search for the easiest printer to use, this didn't took too long becouse it was on my desktop a lexmark Z12.Then watch lots of videos on youtube to figure out how a 3d printer works, Then I did set a flat bed printer to make it move like it should and short after got the build plataforms made, but there was a problem, I did not had the software to slice and print each slice. I was thinking about writing a script to run inside a cad programm to slice and print the models but Talking to some friends about it got a better solution. Paulo cesar de abreu wrote a software to slice and print the .stl file(stereolithography). But more problems show up after that, I had to find a powder that works not easy since the commercial 3d printer makers don't sale just the powder to people that does not own their printers.knowing that the main powder ingredient is Gypsum I did start my research on it. Lots of time spent on this as it's the main thing to make it work. After trying many products I finally set on maltodextrin/gypsum mix.The ideia is to use maltodextrin as a water activated glue to bond the plaster particles together to build the parts.As pure dextrin is expensive to buy and it's sold only in large quantities I had to find a alternative . Searching for dextrin substitute I've found maltodextrin , it's sold as a energetic at some drugstores. that's how the first parts got printed.Problem when using high concentration(over 20%) maltodextrin in the gypsum is that parts feels sticky to the touch even after cyanoacrylate infiltration and it looks like it's made of grains of sand, so a even better powder needed to be formulated . At this point I've found that dental plaster + 5% Pva powder does a really good job and printed parts looks much smoother.The minimal printed feature is about 1.5mm as the Trex arms or hands.What am I using on the printer cartridge? it's a mix of 1/2 deionized water 1/2 regular inkjet ink.The lexmark printhead(cartridge) does not last too long before failling, only about 2000 layers or so after refilling it, some won't pass the second refill. The hardware and electronics is the basic setup to build the 3d printer for about 600 usd. There is a lot of improvement to be made to make it print really good parts. Critical components are, spreading roller needs to be concentric(better then 0.02mm T.I.R) ,good quality linear rails , and good quality shafts for the up/down build plataform. Now for the basic questions:

• Do you sale the printer? No
• Would you make one for me? No
• Would you teach me how to make one? No. if you can't understand what's here then don't bother making one .
• What If I make one that's better then yours? Congratulations! you have the time and money, why you didn't buy a used one instead.
• Do you sale the powder? No
• Do you sale plans? No
• What printers works for this? at this time only the one I've used(lexmark z12)
• Do you have more information on it? it's all here.
• What If I've more questions? try the yahoo tech group link at the left there is more information there, look at the files section there is a folder "Fogassa's 3d printer" with more pictures of the first build.
• Why you don't have more details on how to build one? Not even my clock works for free anymore!

Trex vs HM3DP

Got a better powder to try, the same statue printed with the new powder less time spent on cleaning up , just a air blow and it came out very good.The same goes for the Trex printed in two halfs (I'm running out of powder). the smaller feature was 1.4mm and I needed to be very careful to clean it up and handle it without breaking it.trex lost one of it's hands while trying to scape from the cage, I mean, from the powder bed :-)

Consegui formular um gesso melhor, Voce pode ver a mesma estatua impressa com o gesso novo, menos tempo para acabamento , usei somente ar comprimido.O mesmo para o Trex impresso em duas metades (o gesso estava acabando). O menor detalhe e' aprox. 1.4mm e foi preciso muinto cuidado para dar o acabamento sem quebrar.O Trex ficou deficiente pois perdeu umas das maos quando escapava da jaula, quer dizer, da impressora :-)

work

Sometimes everything goes wrong no matter how hard you try :-)

Did I say It prints 3D objects?

A statue I've found on the net. it's a 110 mm tall statue it took about 2 hours to print and 1 hour in the oven then another 30 min. to finish it using a toothbrush and a picker tool then sealed with cyanoacrylate. As you can see there is a excess powder to it's back becouse the powder was not dry . It looks better then the picture shows. the powder is the same and I've used only water and food dye in the cartridge. I could print it larger but I'm running out of powder need to get some more.

Rapid prototyping technology

Guess what? it's still rainning we had 2 sunny days in 14 days. looks like the weather is going crazy all over the world.Metereologist says it's becouse the El nino fenomen I think it's becouse pollution. So turn that light off or the tv if you're not using it and don't take a plastic bag for one item only when you buy something. Yes you can make a difference, after all pollution is there becouse of what we do.
Is the HM3DP green? yes and no. It does not uses any material that pollutes but the industry that makes it does one way or another.
Products needs to be designed and manufactured much faster then 20 years ago and everyone is on a rush to put their products on market. That's when a rapid prototype machine comes handy.All prototypes were either machined or modeled by hand before the rapid prototype machines and it would take a week or two to get a prototype part for evaluation . With the rapid prototype machine one can design and have a prototype part on the same day.
There is more then one way to skin a cat. Different processes have being used to get parts done but still it's all about the same principle, get a 3d cad computer model and slice it into thin layers then build a part from those layers. I think that the first one to use this process was Stereolithography it uses a resin that hardens with a laser beam to build the parts layer by layer and it's called "Sl process" still the best process to make accurate parts.Another old processes is called "LOM" laminated object manufacture , it cuts the slice contour layers from paper or wood laminate then glue it together to form the parts.The new ones are "FDM" Fused deposition modeling, "3d printing"the name says it all, "SLS" selective laser sintering, and the list goes on.
FDM melts a thin abs rod and build the parts layer by layer . It's the machine to get if you want a fully functional part but resolution is not the best one, layer thickness is about 0.25mm.
3Dp gets you one of the best looking parts and it print in colors but parts are not strong as the the ones made by FDM. One could infiltrade the 3dp parts with epoxy but I've no idea how strong it gets.
SLS selective sinteres a nylon like powder to bond the particles together and parts are strong.
This are the processes I'm familiar with , there is a lot more processes out there like 3d printing using UV resins, a 3d printer that uses metal powder, 3d printing using wax and many others.The build material to use on this machines is not cheap it goes from 100 to 300 dollars per kilo and some of this machines uses a support material to build the parts taking the cost of build material even higher. It's like when you buy a new cheap desktop printer you pay something like 100 dollars for it and pay 80 dollars for a new cartridge when it runs out of the ink .Rapid prototyping machines are very expensive anything from 10.000 usd to half million add to that the build material cost and replacment parts and it turns into a very expensive printer to buy.
One thing that prevents it from getting out of the industry and into the every day home use is the 3d modeling cad software not easy to learn to use.
What's my intention then? since I've made the printer the next step is to try new affordable build materials.some that I still need to try is microcellulose, PVP, PVA, carbon fiber, ceramics and steel powder. maltodextrin and gypsum is cheap enough to keep using it and I'll add some PVA to it as it might increase green parts strength and controll how the ink absorbs into the powder layer.
Almost forgot to mention that I need to add a cover to my 3d printer as dust gets every where and into my nose and lungs, another thing is that some play on the printer's shaft is showing up I'll need to replace the cartridge bearings with a bearing that can run on a dust enviroment such as the ones IGUS have, anyway I think I has gone thru 5 thousand layers and 5 cartridges since I've started to work on this project and some of the cartridge was damage by trying some water based glue on it , besides that it's working very well.

The Software

The software to slice the Stl file short for stereolithography ( a 3d model made of polygones) was written by Paulo Cesar de Abreu another brilliant Brazilian mind( thank's a lot Paulo) as we talk about what was needed for it to work with a regular printer. It opens a .stl file ( 3d model draw in a 3d cad software), slice it out and sends it to the printer in sequence to be printed. It can print in one solid color only that can be pick from the color pallet, so no multicolor part.The 3d model needs to be a valid water tight model ( no cheating here).Parts can be rotate, scale up/down, repositioned to any position on the page, also the layer thickness can be set to any thickness. Once all the options have being set choose printer and hit the print botton. there is a delay in between pages that can be used to allow the printer to go thru the powder recharge cycle before the software sends another page but is not very usefull as some layers takes longer to print then others, so we need to find the layer that takes the longer to print and set the delay by that layer. but I've found a solution for that, what I've done was to use a relay to disconect the pin 11 from the parallel port cable by doing this the computer thinks the printer is busy and will not send another page till the printer has gone thru the recharge cycle, when it finishes the reload or recharge cycle the relay puts the printer back online and printing resumes.It's a done by a line on the code I've wrote.
The microprocessor does all the work to controll the recharging and motion but the printing part of it. It's amazing what a small chip can do this days. Bascom is such a good compiler that I've decided to keep using it along with atmel mcu's. I'll upgrade to the most powerfull atmega16 since it has more i/o pins and bigger flash memory after all I'm almost running out of the 90s2313 flash memory space and using all the pins available.
The stl. slice software is available to download at the yahoo tech group link for personal usage.

First Parts out

We have had only one sunny day this week unfortunately it's rainning a lot again. so I had only one day to try it out and the powder was not dry but I gave it a shot anyway. this are all the same parts printed with diferent water/ink combination. the black one was with regular ink and the others varies as I try water/alcohol mixture. It's a 100 layers parts(10mm) and the smaller wall is 3mm thick. It's holding itself very nice after heating in the oven for 10min. the broken one is becouse I was doing strain testing by hand ( very sophisticated test equipment)Unfortunately I did not seal it as I was expecting another sunny day to try to get a better part and it absorbe humidity then to the oven it went again and again that's why parts looks warped but I'm very happy with the inicial results. the black one that was printed with regular ink seams to hold a better defined part but further testes needs to be made as ink is not cheap. Part Dimension is about +- 0.3mm and +0.5mm in Z direction ( height) but it's only the first parts we will see how it comes along. I did add a 40w heater to the aluminum plate on the build bin but that's not enough as it's bigger then before and holds lots of powder. Printing speed was 2 layers/min. The black part wa printed at 600 dpi and the others at 300 dpi.

Bem, como so tivemos um dia de sol esta semana so deu para imprimir por um dia, Ainda chove muinto por aqui. Mesmo com o po humido tentei imprimir algumas pecas para teste e nao selei com cyano porque estava esperando por mais um dia de sol entao as pecas absorveram humidade e deformaram. As pecas ficam resitentes o suficiente para manuseio depois de serem assadas por 10min. quebrou duas porque eu estava testando a resistencia das mesmas.Sao pecas com 100mm de altura, e a menor parede com 3mm a mais preta foi impressa com tinta da impressora as outras com agua e alcool. A impressa com tinta parece melhor na definicao mas, mais estudo deve ser feito pois tinta de impressora tem um custo muinto alto para esta aplicacao. As dimensoes da peca variam 0.3mm em x/y e +0.5mm em Z(altura) , mas sao so as primeiras pecas vamos ver como se comporta nas proximas impressoes. Eu adicionei um aquecedor de 40 watts no build bin como tinha feito na outra impressora mas como a area aumentou e a quantidade de po tambem acabou nao sendo o suficiente para aquecer o po por completo.

Faster Igor Faster!

I've got the reload speed increased from the last video by adding the ramping alg. to the code and that is the most I can get out of this motor.I did got the same speed with the standard printer motor but as I said before it could not start without the spring's help and was not reliable.
This video was a 400 layers test parts, but unfortunately it was rainning and humidity very high not a good day to print using maltodextrin+plaster, lost the parts becouse the plaster got wet by humidity and the parts end up in the garbage can.I will have to wait for a sunny day to try it again. the mixture used was:

• Plain plaster + 30% maltodextrin
• 30ml of deionized water, 1 ml of isopropyl alcohol, two drops of food die.( no sponge on the cartridge)

O ciclo de recarga do po ficou mais rapido em relacao ao ultimo video devido a ter incluido um algoritimo de aceleracao no codigo fonte e e' o maximo que da pra tirar deste motor. Este video seria de pecas para testar com um total de 400 layers mas, como estava chovendo muinto e o gesso absorveu muinta humidade as pecas foram para o lixo.Vou ter que esperar por um dia com sol(raro ultimamente aqui em ctba)para testar novamente. a mistura utilizada foi:

• Gesso rapido + 30% maltodextrina.
• 30ml de agua deionizada, 1ml de alcool isopropilico, duas gotas de corante pra comida.

Springs sucks!

I've got tired of play with those springs setup to help the little printer motor to move. It could print 200 layers and not fail but the next day it fail right on the first layer.So I need a better motor. I did try about 10 motors before find the right one It's a Minebea unipolar size 23 and works fine with the printer's drive. Now, becouse it's a 1.8 and not a 7.5 degrees I had to make a new gear box and adjust the gear reduction ratio.I had a 32 pitch 18 tooth pinion and the gear came out to be a 115 tooth laser cutted out of a 2.5mm thick acrylic. Now it starts out without the spring to help it and have gone thru 400 layers so far without failing. Now I just need to redo the gear box with better gears.

Cansei de tentar varias molas para ajudar o "motorzinho" da impressora. Um motor maior seria necessario, tentei varios motores de impressoras mas nenhum funcionou.O que melhor funcionou foi um Minebea unipolar tamanho 23 acredite se quiser! , agora este motor e 1.8 graus e nao 7.5, entao seria necessario recalcular a transmissao. Eu ja tinha um pinhao de 18 dentes entao foi so calcular o tamanho da coroa que tem 115 dentes e que foi cortada na minha laser de uma chapa de acrilico de 2.5mm de espessura. Agora a impressao inicia sem ajuda de molas, e se tornou mais confiavel, 400 layers impressas e ainda nao falhou nenhuma vez.Agora e so refazer a caixa de transmissao com engrenagens melhores.

Testing it out

Testing it with plain plaster and ink .Made a new roller got right this time.

Things to do:

• Replace the timming belt
• Work on the ramping alg.
• Replace the print head ( crash it :-(
• Fix a few things .
• Make a acrylic cover for it.

It's a live !!!!

Not that I like red and blue but the cart was already painted blue and I had only red vinyl to cover the top . is it bigger? yes ! is it faster No ! does it works yes !. it works but Iwill need to make a new roller i did a shit job machining it and it's off center. another thing that I need to work on it is to add a ramp algorithm to the code to make the step motor to accelerate to top speed. it's slow to spread out the powder (about 15 sec.) I figure that if I can ramp the printer motor's from start then I can cut the spread time by half and that's the most I can get out of the tiny printer's motor.Anyway I'll be happy If I get 8 to 10 sec. spread time. Another thing I need to get is a new print head I crashed it when setting the printer up and a new timming belt.

Nao que eu goste das cores vermelho e azul, mas o carrinho ja estava pintado e so tinha vinil vermelho para forrar o topo de mdf. Ficou maior? sim! , ficou mais rapida? nao! , funciona? Sim!

Funciona mas vou ter fazer os arremates agora. Terei que usinar um rolo novo porque nao fiz um bom trabalho neste, ficou fora de centro.Tambem vou ter que implementar um algoritimo no codigo para acelerar o motor da impressora para diminuir o tempo que leva para espalhar o po'.

sem aceleracao leva aprox. 15 sec. para espalhar , calculo que com aceleracao posso diminuir isso para a metade do tempo.Mas vou ficar feliz se atingir 8 sec., de qualquer forma seria pedir demais do motorzinho da impressora.Mais uma coisa que tem que ser feito e' trocar o cabecote, bati ele quando estava regulando a altura e danificou alguns orificios de impressao.tambem tenho que comprar uma correia dentada no comprimento certo, emendei a antiga so para teste.

build bin assembly

Getting the build bin together and a view of the seal made from 0.8mm ps sheet and vaccum formed to the desired shape.I did not got a pic. of the lead screws but it's 14mm diam. x 4mm pitch acme thread and delrin nut.

Montando o bin e uma foto do selo feito em PS de 0.8mm e termoformado para se conformar ao formato do bin. nao tirei foto do fuso mas sao fusos de 14mm diam. x passo de 4mm com rosca trapezoidal e porcas de polyacetal.

new top

The new top layout works ok so far. next will be to put together the build bin.

O novo sistema de tracao da impressora funciona bem, montar o build bin sera o proximo passo.

build bin parts

Got the build bin parts machined, and the build bin lined up with a 0.1mm stainless steel sheet.

machining was the easy part but it was a pain to line up the build bin using contact glue. It's better to use a 0.05mm thick sheet but I could not find it.

Finalmente consegui tempo para usinar as pecas do build bin e revestir com aco inox de 0.1mm de espessura. Nao e facil revestir usando cola de contato, um opcao melhor seria usar uma chapa de inox de 0.05mm de espessura mas nao encontrei por aqui.

build

This is the build bin drawing, I'll be using 16mm diam. linear round bearings and shaft and a acme screw coupled to the step motors. support plates will be 13mm thick aluminum. I'll get the parts as soon as I can.As you can see there is openings at the plates to avoid powder build up at the botton if the seals doesn't work as planned.

I'll line up the inside of the chamber with a thin stainless steel sheet.

Esta e' a plataforma onde sera impresso os modelos. usarei guias rolamentadas de 16mm de diam. e fusos trapezoidais acoplados aos motores de passo para os movimentos. As placas de aluminio sao de 13mm de espesura, e como vcs podem ver existem aberturas para o po nao acumular no caso do selo nao funcionar direito.Assim que possivel vou fazer o pedido das pecas.

Build bin

The build bin,

Made from 6mm MDF, notice the epoxy filling at the corners to form a 15mm radius, that makes the box stronger and will make it easier to seal the corners, it's harder to get a good seal with sharp corners.You can see a jig that i've made to spread the epoxy.

it's 540 x 210 x 350mm , the bigger one is about the size of a A4 paper sheet, the smaller one is half of the size.I'll paint it after the epoxy dries.next week I'll work on the mechanism to complete it.

How it works

I've got too many e-mails asking me how it works,so i will try to explain it here.
whats needed besides the printer:

• 3 step motor drives
• 2 step motors size 23 with about 200oz/in
• Power supply(24volts x 5amps)
• The Pcbs and Bascom
• All the hardware (Iwill talk about that when I finish it.)
• The software to slice the .stl file and send it to the printer

The software was writen by my friend , Paulo, and you can get it here http://tech.groups.yahoo.com/group/diy_3d_printing_and_fabrication look at files section. don't think you'll make the next miracle 3d printer, mass produce it and get rich using the original software, becouse you can't! you can not use it for any commercial application nor sale it.It works better with .stl files from Rhinoceros CAD, open your .stl file, set the layer thickness and print it.There is delay option that you can use to pause when sending data to the printer , use that to allow the printer to spread a new plaster layer. As I'm rebuilding the printer I've used one of the relays pcb to put the printer in pause by disconnecting the parallel port busy signal, it's the pin 11 on the db25 conector, so , the computer thinks the printer is busy and won't send another page . When my lexmark finishes printing a page, it keeps moving untill the paper sensor detects that there is no more paper inside the printer(purges the paper out), the sensor is nothing more then a lever that blocks a light sensor.I've removed the sensor from the printer's pcb board and relocated it to the side of the printer sensing a aluminum plate not the paper. when it senses the plate, it keeps moving about 80mm and stops ,thinking that it purged the paper out and there is a new paper in ready to be printed, it won't look for the paper edge since it thinks the paper is there already, So It will start to print as soon as it gets the data to be printed.

Now , when it reaches the end (after the 80mm back end) it triggers a NO switch(I'll be using a opto switch) that starts the powder recharge cycle, this is where all the magic happens, the microprocessor reads when the switch triggers and sends a step and direction signal to the step motor drive to lift the feed bin by about 0.2mm, switches the relay on to turn the spread roller motor on, switches the printer's step motor from the printer's drive to the external drive and pauses the printer by breaking the pin 11 on the parallel cable using the relay pcb Board, next it moves the building bin about 2mm from the previus position then the printer moves forward to spread a new 0.1mm plaster layer on top of the just printed one, after spreading the powder, it lowers the building bin and the feed bin, the printer moves to the start position turns the relay off bringing the printer online and waits for the next page(layer).

What build material do I use?

I'm using dental plaster + 30% maltodextrin mix and distilated water instead of ink. If you try to use just plaster it won't work, you need a water activated glue(maltodextrin)to hold the plaster particles together.You can get the maltodextrin at your local drugstore, it's sold as a energetic, there is lots of brands some work better then others.Mix it well with plaster. Parts comes out very fragile and I'm using Cyanoacrilate to infiltrate the parts after printing to make it stronger .One could use epoxi to infiltrate it and get stronger parts. Printing quality was not very good, mostly becouse it was a "I think its going to work" project and there was hot glue holding parts every where, that's why I'm rebuilding it with a better hardware setup. The minimum wall thickness that can be built is about 2mm, but it dependes on how tall it is and what kind of magic powder do you use, if you find something better then what I'm using don't be shy, e-mail me.It took me almost two years working on it(on my spare time)till figure it all out.I've started by looking how printers work, and I choose the lexmark z12 becouse it doesn't make lots of movements before start to print.

I've set this blog to show you what I've done, I do not sale anything you see here nor plans on how to do it,that's up to you to build your own, you can use some of my ideas to build your own printer but don't blaime me if your's doesn't work, the only printer that I know that works is the lexmark Z12(the one I'm using) but if you're using a diferent printer and it works let me know.

Now this all I can tell you before get the new build chamber done(I'm working on it), I've got the new longer rails (Drawer slides) 700 mm long, just perfect.Next is to build the new building chambers, stay tunned.

Project changed

Project changed, I've being thinking about rebuilding the printer and why not make it bigger?so the new specs for it is:

• Build size 300mm x 210 x 125mm tall.

The electronics still the same but I need to get the material and build a new building chamber, also I need new rails about 600mm in lenght.It's hard to get drawer slids with this lenght so, I might need to buy comercial linear shafts.So stay tunned, it can only get better.

Mudanca no projeto, pensei em reconstruir e documentar a impressora mas, pensando bem vou reconstrui-la maior, o tamaho sera maior 300x210x125 de altura. Com esta mudanca , sera possivel imprimir pecas do tamanho de uma folha A4.

A eletronica continua a mesma, mas sera necessario construir novos depositos de construcao e tambem tenho que comprar novas guias lineares, talvez nao consiga as guias de gaveta e terei que comprar guias industriais.Entao espere mais um pouco porque ficara melhor ainda.

let's talk about programming the mcu

You'll need a compiler(A software that writes to the mcu) I think the best one for beginners is BASCOM AVR http://www.mcselec.com/ it's a basic language compiler, it translates the basic language into the more complex hex code that the chip understands ,it's Easy to learn and the demo version allows you to program your mcu(at90s2313) with the basic code that I wrote.
You also need a hardware programer connected to the parallel port of your computer to send the program to the chip, I'm using a Stk 500 compatible programer , you can get that from the Bascom site.Connect the cable from the hardware programer to the mic. pcb board using the ISP pins and download the compiled code to the chip.
Now, depending on how you have set your hardware( screws, step drives and transmission) you will need to modifie the source code to fit your hardware, I've set this as variables in the source code ,so it's just a matter of change those numbers. this is the code and it's free to use for personal use only, by using the code you agree that you're not going to use it on ANY comercial machine nor resale it.


E necessario um compilador(software que escreve no chip), acho que o melhor para iniciantes e o Bascom AVR, http://www.mcselec.com/ que usa a linguagem basic facil de aprender e traduz a linguagem basic para a linguagem hex(linguagem maquina) que o chip entende. Voce pode usar a versao Demo do Bascom para compilar o source code que eu escrevi.
Tambem e necessario um programador fisico conectado entre a porta parallela do computador e a PCB do chip , para fazer a programacao propriamente dita. Estou usando um programador compativel com o stk500, entao, conectar o cabo do programador aos pinos de ISP da pcb do chip and faca o download do arquivo compilado.
Dependendo de como voce fez a sua impressora(fusos,step drives,e transmissao) sera necessario alterar alguns numeros no codigo fonte, para tornar isso mais facil, eu coloquei estes numeros em variaveis, bastando mudar os valores dessas variaveis e nao o programa inteiro. este e arquivo fonte somente podera ser usado para uso proprio, fazendo o download do arquivo fonte voce esta concordando em nao usar o mesmo para fins lucrativos de qualquer forma nem usar para comercializar nenhuma maquina que use este arquivo fonte nem comercializar o arquivo fonte.

"HM3DP source code"
$regfile = "2313def.dat" 'at90s2313
$crystal = 10000000 '10mhz crystal
Dim Distance As Word
Dim Volta As Word
Dim Feed As Word
Dim Feedreturn As Word
Dim Build As Word
Dim Buildreturn As Word
Config Debounce = 30
Distance = 2150 'distance the printer moves forward to spread the powder in steps
Volta = 600 'distance the priter travels back to the start position in steps
Feed = 640 'lift the feed bin by 2mm
Feedreturn = 592 'returns the feed bin by 1.85mm
Build = 608 'lift the building bin 1.9mm
Buildreturn = 640 'returns the building bin 2mm
Config Pind.0 = Output ' roller
Config Pinb.0 = Input 'sense switch
Config Pinb.1 = Output 'enable/disable drive
Config Pind.1 = Output 'step X moves the printer forward
Config Pind.2 = Output 'dir X direction signal
Config Pind.3 = Output 'stepfeed
Config Pind.4 = Output 'dirfeed
Config Pind.5 = Output 'stepbuild
Config Pind.6 = Output 'dirbuild
Config Pinb.2 = Input 'dirfeedmanual
Config Pinb.3 = Input 'stepfeedmanual
Config Pinb.4 = Input 'dirbuildmanual
Config Pinb.5 = Input 'stepbuildmanual
Portd.6 = 1 'build direction
Portd.4 = 0 'feed direction
Portd.2 = 1 'printer direction
Portd.0 = 0 'roller
Portb.1 = 1 'set pin high to start

Direcaomanualfeed: ' start the manual move of bins
Do
If Pinb.2 = 0 Then Exit Do
Portd.4 = 0
Pulseout Portd , 3 , 5000
Loop

Manualmovefeed:
Do
If Pinb.3 = 0 Then Exit Do
Portd.4 = 1
Pulseout Portd , 3 , 5000
Loop

Direcaomanualbuild:
Do
If Pinb.4 = 0 Then Exit Do
Portd.6 = 0
Pulseout Portd , 5 , 5000
Loop

Manualmovebuild: ' ends the manual move
Do
If Pinb.5 = 0 Then Exit Do
Portd.6 = 1
Pulseout Portd , 5 , 5000
Loop

Inicio: ' here is where the cycle starts up by sensing the trigger switch
Do
Debounce Pinb.0 , 1 , Feedbin
Loop

Feedbin: ' moves the feed bin up
Portd.4 = 0
Portd.3 = 0
Do
Pulseout Portd , 3 , 1000
Waitms 2
Feed = Feed - 1
If Feed <= 0 Then Exit Do Loop

Buildchamberup: 'moves the building bin up
Portd.6 = 1
Portd.5 = 0
Do
Pulseout Portd , 5 , 1000
Waitms 2
Build = Build - 1
If Build <= 0 Then Exit Do Loop

Rolo: 'turn the roller on
Portd.0 = 1
Wait 1
Portb.1 = 0 'enables drive
Eixox: 'moves the printer forward
Portd.2 = 1
Portd.1 = 0
Waitms 1000
Do
Pulseout Portd , 1 , 1000
Waitms 2
Distance = Distance - 1
If Distance <= 0 Then Exit Do Loop

Buildchamberdown: 'lowers the building bin
Portd.6 = 0
Portd.5 = 0
Do
Pulseout Portd , 5 , 1000
Waitms 2
Buildreturn = Buildreturn - 1
If Buildreturn <= 0 Then Exit Do Loop

Feedbindown: 'lowers the feed bin
Portd.4 = 1
Portd.3 = 0
Do
Pulseout Portd , 3 , 1000
Waitms 2
Feedreturn = Feedreturn - 1
If Feedreturn <= 0 Then Exit Do Loop

Retornox: 'return the printer to the start position
Portd.2 = 0
Portd.1 = 0
Waitms 1000
Do
Pulseout Portd , 1 , 1000
Waitms 2
Volta = Volta - 1
If Volta <= 0 Then Exit Do Loop

Portb.1 = 1 'disables drive
Wait 1
Portd.0 = 0 'turns the roller off

PCB Boards

controller

Here you can see what the controller looks like with 3 drives, the relay board,microprocessor board,and power supply

Este e o controle da impressora com 3 drives, a placa de reles,placa do microprocessador e fontes.

Schematic

Prototype PCB boards finished.
J3 conector is out put to step drives(3) this is how it works;
When the cycle start switches triggers the processes starts, sends a signal to:

Switches the relay on(Pause printer, turn the roller on,switch printer step motor to the external drive)

1. Enable external drive
2. Move Feed bin up 1mm+0.15mm from the last position
3. Move Build bin up 1mm- 0.1mm from last position
4. Move printer forward to spread powder
5. Move Build bin down 1mm
6. Move Feed bin down 1mm
7. Move printer to the start position
8. Disable external drive
9. Turn relay off
10. waits for the next page to be sent.

Placas de pcb prontas.J3 e saida para os drives de motores de passo.Funciona Assim:

Quando a chave de cycle start e acionada o processo comeca e manda sinal para ligar os reles que vao, Pausar a impressora,ligar o rolo,e mudar a alimentacao do motor da impressora para o drive externo, entao:

1. Habilita o drive externo
2. Move o feed bin para cima 1mm+0.15mm desde a ultima posicao.
3. Move o build bin para cima 1mm -0.1mm desde a ultima posicao.
4. Move a impressora para frente para espalhar o po
5. Move o build bin para baixo 1mm
6. Move o feed bin para baixo 1mm
7. Move a impressora para a posicao de inicio.
8. Desabilita o drive externo
9. Desliga os reles.
10. Espera pela proxima pagina.

motor switching pcb

I've changed how the printer moves, I'll be using the original printer motor to move it back and forth, no additional motor needed any more, I've built a opto coupled pcb board with two relays, one with 2 reversible contacts and another with 4 reversible contacts.
The one with 2 contacts will put the printer in pause during the powder recharge cycle and turn the roller on.the one with 4 contacts will switch the printer motor from being driven by the printer driver or the external drive needed to reposition the printer to the start position.

Eu mudei como a impressora move, agora nao precissa do motor extra para reposiciona-la, uso o mesmo motor da impressora, para isso fiz uma PCB com dois reles um com dois contatos e um com quatro contatos, o de dois contatos poe a impressora em Pause para o ciclo de recarga do po e liga o rolo, o de quatro contatos muda o motor da impressora de ser alimentado pelo drive original da impressora para ser alimentado por um drive externo, necessario para reposicionar a impressora no inicio da pagina.

gear box mods

The original printer gear box, I've cut the roller that move the paper, and kept the gear at the end of the shaft, then I've pressed a timing pulley over the shaft to make a new traction gear box that moves the timing belt thus moving the printer back and forth

Esta e a caixa de engrenagens original, cortei o rolo que traciona o papel mantendo a engrenagem e entao prensei uma engrenagem sincronizada para formar um conjunto que vai mover a correia sincronizada movendo assim, a impressora.

top assembled

top assembled

Parte superior montada

Roller

The roller assembly

Rolo montado

the sliding rails

Custom made sliding rails from a drawer slides, but any linear rails could be used.

Guias lineares feitas apartir de guias de gavetas,mas qualquer tipo de guia linear pode ser usada.

hardware needed for the top

Here is what is needed, a roller and motor to spread the powder,two sliding rails,some aluminum L shape plates, and a timming belt from a old epson printer

Necessario, um rolo e motor para espalhar o po, duas guias lineares para movimentar a impressora e chapas de aluminio , e uma correia sincronizada retirada de uma epson velha.

The not so happy printer

The standard Lexmark Z12 Printer

another one

A video off it

The homemade 3d printer rebuild started

Today I've started to rebuild the 3d printer to make it easyer to build with standard tools and off the shelf hardware, it started out with a regular lexmark desktop printer, as I'm doing it on my spare time it will take some time to get it done, so check back often to see how this project is going.Here is the specs for it.

• Parts Bulding size 6.5 x 4.5 x 4 inches
• Driven by 2 nema 23 step motors
• 3 step motor drives needed

Hoje eu comecei a reconstruir a impressora 3d para que possa ser construida com ferramentas normais e pecas disponiveis.Como estou construindo no meu tempo livre vai levar algum tempo para terminar, visite meu blog para manter-se informado em como este projeto anda.