Sunday, 16 October 2011

Progress Update 6

I have made some progress, i think i have enough parts to assemble the printer to a point where it will print some plastic YAY...

Last night and yesterday i completed the heated bed and 1/3rd of the hotend for the exxtruder, i have decided that it makes sense to get the hotend made up on a CNC mill/lathe.

 Doesn't look like much at all does it? lets look whats hiding below.
At the top of this image you can see the bottom of the bed, this is what mounts onto the linear bearings. the bottom section is the underside of the top.

 this is the underside of the bottom of the bed, you can see the two straps with the holes for the linear bearings.
 here you can see the M6 Allen key bolts that i have used as adjustment screws that level the print bed. these mate with some M6 Nutserts that i have place in the 3mm heat spreading plate that the glass sits on.

here you can see where i went a little nuts on the punch in my mates workshop, this is to allow access to the terminals for the heating element. this is 240v AC, need to make sure this frame is earthed and not touching the active or neutral lines...

 here are the parts for the prototype hotend. this is only 1/3rd of the parts you need. you need 3 x all the parts you see here.. 6mm OD,5mmID  x12mm brass tube, 2x M3 x 30mm screws, Nozzle and 4r7 3watt RWM series resitstor, you can see in this shot the hole on the left has a thread cut 10mm deep for the nozzle to mount.

 here you can see the PTFE mounting block.
 first slide the resistor in to the 6mm hole that does not have thread cut in to it.
 Then press the brass tube sections into the 6mm hole, one from each end, this does two things, it makes sure that the maximum surface area is contact with metal and allows easier replacement if/when the resistor dies. the reason its in two pieces is the resistors when you get them are not uniform in their diameter, so the two pieces allow enough movement to ensure it is held tight with no adhesives. when this resitor dies, knock out the old one with your calibrated hammer and replace withe new resistor and tubes if needed...

NextI screwed in the brass nozzle, to do this i cut a slot in the top of it for a flat blade screwdriver.
here you can see the hot end assembled for testing, i have placed a thermistor inside the nozzle where the filament would normally go, this is the green and white wire.. so i applied 12volts onto the resistor and watched the results..
after a little wait (2mins) the temp reading was 200'c after another min it was @ 270'c
so i think tonight if i dont find something more interesting to play with i will mount the thermistor to the bed and wire that up an then consider mounting the hotend to the extruder and see if i can make worms of plastic.. this nozzle only has a .55mm hole as all my smaller drill bits were snapped and i am yet to have more 0.25 or 0.3mm drills arrive...

Wednesday, 12 October 2011

Progress Update 5.5

I have been working on the extruder for the last few days. i have the cold end done and the parts for the hotend ready to be drilled, will need to order some new thermistors (will make calibration easy if they are all the same), then drill the brass heater blocks and make the nozzles, i am thinking aluminum  nozzles may be the go.

i have placed the extruder onto the X carriage, no holes or bolts have been placed in yet. just wanted to make sure it will fit between the two belts.

 here is the printer, the X belts are not attached to the carriage it's just sitting there with a gravity clamp.

here you can see the cold end of the extruder sitting in my most desirable orientation, but it only measures 55mm wide so it should fit across ways onto a normal printer if it ever made it there.

 side view
 and the other side as suggested i may be required to put a guide that pulls the plastic towards the idle wheel so it doesn't scrape away when not selected.
here you can just see the three springs that push down onto the top of the levers these are directly above the camshaft and vertical.

Triple Extruder Cold End Prototype BNZII

I have completed a yet to be tested prototype of the BNZII coldend.

this is what it has turned out like it. i have taken nicely out of focus pictures of it from all sides i will try and explain whats what.
here you can see the two stepper motors the one on the left drives the filament via the hobbed axle.

here is a top down shot, you can see the hobbed axle, and the three idle wheels that run in the hobbed can't really see it here but the springs push down on the levers which then press the idle wheel against the hobbed axle.

 not much more detail here, but you can see the same drive system for the camshaft as the filament. 20:1 reduction via the worm drive and spur gear.
 here you can see the camshaft on the left the top lever is engaged to grip the filament this is done via a groove making a low spot in the camshaft this lets the tension from the spring pivot the lever or arm and presses the filament onto the axle. roughly a 1:1.3 ratio (x1.3)

 the middle lever is engaged to grip the filament, i thought it was much easy to work on a low spot to release the spring tension onto the drive axle was a better way to do it, allows the filament not to be exact 3mm.

 this is the bottom lever engaged. this extruder can be modified to take a 1.75mm filament really easy. change of bolt and idle rollers and your ready to roll..

 here all the levers are in the released state, you can turn the camshaft in this state by hand really easy as i have use 7mm OD bearings for each of the camshaft lobes to run on. costs a little and adds a little thickness to the design but i have doubts if the hotend can keep the three nozzles at three different useable temps, two plastics and a fill material may have wide temp differences! i don't know...yet

another shot from another angle, this thing was a PITA to put together in may lap on the couch so with a vice it will be much easier.. here you can really tell its a prototype you can see the hotglue that holding the bearings in place, its just keeps them from sliding out of the holes they live in when it was being handled... its not really structural when required i will epoxy in new bearings and the next one will have 6mm walls for the bearings to live in. when i next need to pull it apart i will take photo's that are in focus and that include the internals. cant wait to get onto the hotend, i cut the brass last weekend its a little over 10mm thick at the moment but i have 3/6 machined edges to go off, will need to get these CNC'd and or water cut but that should be no problem.

Saturday, 1 October 2011

The New PrintBed

I was talking with the peoples on #reprap and they brought to my attention that the PCB heaters work better than i had thought, this got me thinking do ii have to go to a HUGE mains powered heater for my new bed, this one is going to be light weight. And that has been the driving design criteria.

I was thinking that with a printbed that is 470x470mm that could be broken up into 9 squares of 156mm, looking at my local electronics shops website i see i can get 10 pieces of copper clad FR4 single sided for $6.95 a piece, and a resist pen for $9 this makes the heaters about $80 as i already have the acid and the spare time :)

there is a website that i sometimes use that will do 5 boards for that $80, so is my time worth $80 to me?? maybe but then there is the risk factor to deal with.. also there is the question is a 3x3 grid the best idea, what about a grid with 5 cells, think quarters then take a square in the center as well, this will mean bigger pcbs i might be able to make a layout that has a optional corner cut-out so i only need to order 5 boards the same and cut the corners from the 4 and leave the center one whole.

this will require me to have more analog inputs on my arduino controller, that will be 3 nozzels and 5 bed segments that's a total of 8, the mega has 16 available but they may be appropriated for other uses and may require re-routing,there are more than enough digital outputs available.

there is another pattern that i did think about using much the same as the 5 segment picture above but the square is rotated 45 degrees to make it easier for the glass cutters if i choose to go with individual tiles but that leads to wasted energy.. and thats what i was after by splitting the bed.

Update : after i made this post another suggestion was put to me for segments, I could use a similar pattern to the 5 section design but the idea was to have two hot zones ie the inside square which is where most prints get started from the middle of the printbed, and the outside square that only gets turned on for lager sized objects, this idea is interesting to me, it may reduce the number of thermistors required to monitor the bed. i am not really chasing efficiency just thinking about whats involved in removing the 240v AC heater from my printer.

BNZ's Hotend - Triple Extruder BNZII

Today I Finished my design for my new hotend this will take three power resistors and three thermistors or thermocouples, The power resistors will be fire cemented into short (15mm long) sections of thin walled brass tube, this will then be installed into the hotend, one for each nozzle. with a little heatsink paste.

the nozzles slip into place from above and will also have a interference fit, and some heatsink paste, i will be trying out a variety of nozzles so this hotend is fairly easy to remove once i complete the mounting plate for it.

Below are some screenshots from solid works showing the construction of the pieces, this section is by no means printable.

the height of the hotend is 29mm this is 4mm more than the height of SC10UU linear bearings that i have used for my X cariage. this wastes no Z axis movement.

 side view of the hotend,  the top greenish piece is peek 10mm thick as a thermal barrier and also to hold the slip in nozzles tight.

here is a cross section view from the side. you can see the nozzle section and the heater resistor cavity and on the right is the thermistor point. might not be drilled that deep in reality but meh.

 here is a bottom up view of the Hotend, the nozzles are 10mm apart and the white holes are mounting holes M3 will probably counter sink them to help clear the printed plastic. The grey holes range from 1mm to 3.5mm (except for the 5mm that power resistor's home) are to try and keep the heat from wandering between the nozzles, a probably feeble attempt but hey, this is going to be CNC machined from brass as there is no hope i could get all those holes right in a drill pres.

 here is a front view if the three nozzles the model has 0.30mm holes but we will see what works
 here is a semi exploded view of the three nozzles they have a 9mm OD flange (heheh i typed FlangE)
 here is a cross section of the exploded view

and the cross section view of the three nozzles, so far the rebuild of the extruder has saved about 10mm of wasted print height

Here is the New extruder  mounted onto the X carriage. its not fully mounted i will decide how that happen when i build the prototype,  i have to raise the axles up 3 or 4 mm more from the plate to allow for the 20toothed gears that will be used to drive the axles.

in this Design all running surfaces have been replaced with tiny 703, 7mm OD 3mm ID bearings that i get from the hoby shops at a ok rate for 2 at a time, this design will require 9 of these and 4 1608 bearings. but the wont wear too fast. this design is also missing the springs that will push down on each of the tension arms. the plate that the springs will press against is not in the drawing but will be added later.

here are the screen grabs.

this is a side view, looking along the X smooth rods.i will be able to bring the motor in a little more maybe 5mm before they will start to interfere with the filament too much. the area outside the worm drives is where the belts will run, i have to make a new idle axle and some extenders for the X axis to spread the belts to all this to mount up.

 this is a front view of the extruder mounted on the X carriage  i could have mounted it the other way but i prefer this way, i also get to shorten the X carriage length.

A 3d angle shot to help you get the picture. the two motor are not exactly in line with each other, this was done to give me a little more leverage on the camshaft.

i will be driving both the filament and the filament select camshaft with Nema17 2300g/cm stepper motors ( with a 20:1 worm drive reduction to get enough torque..)  they will be driven by a pololu stepper driver. with 1/16th the motors have 200 steps per rev and 2300g/cm so that's
64000 steps per rev, and ~46Kg/cm of torque if an can convert half of that into grip  i will be off to a good start. thats for both the feed and the camshaft which should be suffcient.

Update : The pictures here show a single piece heater block, the idea here was that there would be holes drilled between each of the three sections of the block to provide simple mounting and thermal isolation between each heater, this is too hard to manufacture for my liking, all those 0.5 and 1.0mm holes are a PITA to get right with out a decent MILL and plunge bits. so i have changed the design to use three separate pieces of brass, one for each nozzle, that will be 8mm thick this will give me 2mm of air gap  between each of the heated sections that can be packed with ceramic insulation tape if required, this does complicate the installation of the nozzles but simplifies changing of the nozzles as each one can be removed and replaced/repaired without disturbing the others, one drawback to this is that each nozzle can be installed into a slightly different position if the mounting holes are not the perfect size.

Thursday, 29 September 2011

Printing Surfaces

So there was a little discussion on #reprap the other night about what is the best surface to print onto.
I have tried a few different surfaces and have come to the conclusion glass seems to be the best, its easy to clean, but hard to work with when making your printer.. as you (should be)  printing more than assembling the difficulties in working with the glass should not really be an obstacle.

I have tried several surface materials with ABS (haven't tried PLA yet), l will list them for you..

MDF (plastic doesn't stick well, i suspect it sucks too much heat from the plastic for it to stick)
MDF with kapton tape (plastic sticks ok but will warp up once it cools.)
White masking tape ( sticks ok but my settings were not right and i was applying too much plastic so the nozzle knocked the cube free before it finished.
Bare Aluminium works well but can scratch up really easy.
Glass Unheated. Plastic does not stick!!!
Glass Heated plastic only sticks when its hot enough. ABS110'c. this works like a charm.
Glass Heated with sandblasting surface finish. (0.5 welders slag for sandblasting grit) plastic sticks to this like turd to a blanket.. and is just as hard to clean... when you scrape the top clean there is a little bit of plastic crud that is left behind and cannot be removes by scraping. need to soak in a chemical (acetone for ABS) this level of cleaning (IMHO) is not really required unless you have some real contaminant on the glass bed (eg lard used on crappy nylon bushes on your x axis.) then you may need to really remove the excess plastic that may be harboring some grease or something that will stop the plastic from sticking. there is a caveat to the sandblasted glass, it will efffect the finish on the bottom layer of the object being printed, you will get a surface texture that is the same as the sandblasted surface.
 you can by all means get smaller grits for sandblasting but the 0.5mm grit was what was at hand.

here are some pics of the bottoms of objects that i have printed..

 the two pictures above are of things i printed on a high gloss piece of glass

the above three pictures are objects that i printed on the same piece of glass as above but after the surface was sandblasted with 0.5mm welders slag. this finish is not too bad if you ask me.. i kind of like it better than the full gloss finish. the result of the sandblasting was the plastic stuck really really well at one point i thought i was not going to be able to get the objects off..

for my new printer i am considering sand blasting one side of the sheet of glass that way i have the option to turn it over if i do or don't like one...

there was discussion of getting the glass chemically frosted, as well as frosted with a laser. the chemical treatment would probably work, but from what i learned about how the laser is used i am doubtful that this will have an effect. the laser treatment is performed by submerging the glass object in mineral oil and then directing a laser at the area to frost. the frosting (as i believe the process goes) is a result of the heat of the laser liquifying the glass (maybe to the point of plasticity ) and inducing boiling bubbles in the middle of the glass not the surface. I think the surface will still be flat and smooth. this is only an aesthetic change not a surface treatment. we need the surface to be flat and grippy.  if you look at the pictures of my first printer you will see the glass that i was printing on, there is two pieces one is sandblasted the other is not. they were scrap pieces of glass from some old kiosks at my work.

Sunday, 25 September 2011

Progress Update 5

Today I really really want this printer to do something other than take up space, And after the other nights efforts of Nuking a few stepper controller chips I have decided to install the old crappy Gen3 electronics set.

Oh the Gen3 is a Ugly ass creation, effective to some degree but its a mess!

In my travels today I latched onto some lubricant, Its a can of Spray lithium grease, it smells a little,but no where near as bad as the marine bearing grease that I had used on the old printers axis'
 and @ $10 a can its a bargain, this will make a thick layer on what ever it is sprayed onto, this will be great if i every have to park up the printer and store it for a bit while not being used this will prevent the steel from going rusty.

here is a picture of the printer as it stands, apart from the extruder the beast is wired up.

While i was attaching the endstops someone on IRC #reprap had asked the question about an easy way to square up the print bed, having seen this question pop up a few times previously i thought about it for a moment and devised a simple Z probe for my printer. this is what i have come up with.

 here you can't really see it through the glue holding the cable into place, but i have a tiny micro switch with a lever that hooks into a bit of brass tube that the bolt pushes up, there is a little slot that the lever fits into, this allows me to adjust the height with out it interfering with the amount of travel.
here you can me pressing the bolt head up to illuminate the indicator LED. now i am not 100% sure if this will work, or if its useful so i have also made a traditional micro switch endstop for the Z axis Min, just in case. I am thinking of putting in a switch that will allow me to choose between the two options. MIN or Z probe. the firmware will be fun.

Things left to do.

Need to go an purchase a piece of glass that is 2mm 470mm square, the framing shop quoted me $25 cut to size.
Will need to weld up the frame that will support the glass, heat spreader, element and heat shield. this will be place onto springs and held down with some screws to provide adjustment to get the bed nice and square to the extruder . ahh that reminds me i think i will get the glass cut with some 20mm 45' cuts on each corner to allow for the adjusting screws, getting the corners cut off is much much cheaper than getting holes drilled into the sheet glass.. not as pretty but who is going for looks, i want functionality.
Then all that is left is the Extruder...

but for now i am going to go and play with some firmware to see if i can get it to listen to the endstops. i was playing with Teacup firmware by Triffid_Hunter. but the version i had doesn't explain the endstops really well, so i have no idea if i am configured correctly.
So now i will try Sprinter by Kliment and see how that goes.

Ok Sprinter seems to be working  great, i managed to map all the pins to my setup and i can exercise all the axis' My X and Y axis' move upto 1700mm/min (283mm/second). this is GREAT its about 27 times faster than my previous printer, but the accuracy has been reduced i get 4 steps/mm (ie 0.25mm steps) i will be getting some new 10T pulleys and some new stepper driver chips that will allow better micro stepping option. this will give me 64 steps/mm :) 0.015625mm/step..  i could go even further and use a 8tooth pulley but the belts do not like going around those. the radius is too tight, and would require an idle roller to over come that, that sounds like too much trouble and will only give me an extra 16 steps/mm not sure if i need that level of accuracy.

so far i have tested all three axis' they all work fine except for Z it seems to bind in the upper 30mm of travel, not too sure where the problem is but i dont care about that. this limits my print area to 470Wx470Lx175H on the small bed. ahh things never work out how you plan.. :)

Saturday, 24 September 2011

Not happy!! :(

Just blew up two stepper controllers.. not happy.... have to update/create a separate  page for parts i have blown up, or broken during this whole fiasco..

I Never have any luck with stepper drivers, the only time i get to actually use them is when they come in a kit like the gen3 :( now i am waiting for a RAMPS pcb to arrive in the mail, will start collecting the parts for that, will need to get a arduino mega 1280 to ensure its a hassle free operation.

its not been a good week, the post man took three days to deliver my parts that i just blew up in 30seconds, this hobby doesn't really make ecconomical sense in any way shape or form, but Meh neither does drinking, and we all keep doing that....

have to order up some more steppers, i needed to order another two for the extruder, but now i need another two, and i probably should get a few spares, hmm think i am going to need to get just the IC's and gets the hot air gun out or it will be noodles for me real quick...

Friday, 23 September 2011

Progress Update 4

Last night i was at my mates workshop to build something to connect the belts to the tension springs without the belts wearing. these were really simple they are made from 3mm aluminuim, with a hole for the spring and a slot that the belt can go through. the belt is then bent back and ties off with some cable ties. 

I Also  made some brackets for the X carriage to hold the springs and anchor the belts, these are now installed and ready to go.

I had to add a spacer between the angle and the anchor to raise it above the bearing bolts.

I just visited jaycar and left with lots of goodies, i bought some multi core cable to extend the stepper motor wires, and a whole stack of various connectors and heat shrink, I have been tossing up on weather to upgrade my electronics or not, so i thought i will do some thinking and see what i need from my electronics.

so i want to have 3 nozzles, this will require 6 pins 3 analog inputs and 3 pwm outputs, i will need to control the heated bed this will take another analog input and 1 digital output, could be pwm but it doesn't need to be. i need 1 digital pin for the fan. 12-15 pins for 5 stepper controllers. and 6 pins for the endstops.

6   +    Heater + Sensors
2   +    Heated Bed + Sensor
1   +    Fan
15 +    Steppers ( S,D,E x 5)
6         Endstops

that be a total of 30 pins, that's just what the ATMega644 has exposed on the Gen3 motherboard, yes i will have to use a USB - TTL converter but that's easy. I made one from a pic18f14K50 chip and crystal, used the sample firmware provided by microchip (there is a cdc emulator that is perfect for this no mods needed).

 X carriage you can see  the belts anchored on the left and attached to springs on the right, this keeps the belts nice and tight, the anchors for the belts should help prevent them wearing, if they fail i will get the corrugated plates made for anchors.

 The Y Axis belt attaching to the springs. i bent the belt back again just to keep the excess from interfering when the carriage goes all the way to the end, the excess would jam the pulley.

 just a couple of shots showing me the measurements so i could make the brackets the right size.

 I have started to layout the track board RAMPS type carrier for my nice new shiny stepper motor controllers they are pololus without the 5 volt regulator. that's no problem i will add one its easy.

i am planning on soldering the gen3 motherboard to the track board and run wires to the bit on the under side, this will be much neater than the previous setup. as you can see there is 4 pololu stepper controllers there, these are fro X Y and 2 for Z, I could use one driver for Z but my motors are Larger Nema23's so i will use one for each. i will then reuse 2 of my MB2.3 stepper controllers to drive the two stepper motors on the extruder. one is for selecting the filament and the other feeds the selected filament. i am thinking of re-designing the extruder, and i may change the filament selection stepper for a smaller one with a much much higher gear ratio, the camshaft doesn't need to go really fast so i can use a smaller one and save weight, i might even switch to a smaller stepper driver to suit. I have a stack of PM14 and PM20 stepper motors and some 1amp driver chips to suit, i have drawn up a control board for them, yet to get them etched, i will one day soon, if anyone wants to look let me know and i will post the schematic and pcb files, i have used a pic to convert the SDE signals to grey code that the stepper driver wants, the pcb is really small, i have drawn the board up to take 5 of these but can be separated. the plan is for a micro printer one day, will be slow but very very accurate, a friend of a friend is a jewellery setter and  can drill me some micro nozzles, this micro printer would have a 50mm^3 print area designed for wax casting.