Page started April 2014
I've resumed work on my large format 3D printer idea.
I started mid 2013 but put it on hold to put time and money into my para-gilder.
That is now on hold waiting for good weather and rules to change.
The project started as freecad project but freecad was too buggy and I've gone back to using sketchup.
Quick status - 13th Nov 2015.
Now over a year old Huey is doing well.
Head over to the huey log to see what I've been doing.
The link to the build log here
Overview.
This is not a general purpose 3D printer. It is the size of a small chest freezer and has a expected build volume over 200 litres. It uses eight drive motors and may require a lot of setting up compared to printrbot or similar.
The parts are mostly printed on my printrbot.
Unlike most repraps Huey has a stationary bed. Therefor the printhead must move in X,Y and Z dimensions.
Provided motors have enough motor torque and power this should scale up to be even larger.
It is experimental and could fail to work.
Name.
I wanted to avoid adding another "bot" to the world.
I wanted a "big" name and toyed with naming it Moa after the extinct New-Zealand bird.
Extinct Australian birds names like Dromornis stirtoni don't exactly role off the tongue.
I wanted a big name not necessarily a long one.
So I picked another big bird from my childhood - http://en.wikipedia.org/wiki/Baby_Huey
Frame.
The frame for Huey is made from 25.4mm square tube. I'm using galvanized steel.
It is old stuff - maybe 25mm is easier to get now.
The vertical tubes have bearings traveling along them and aluminum extrusion could also be good here.
The horizontal tubes can be rougher.
Using steel may have a slight advantage in having a better thermal expansion match to the hardened steel rods.
The horizontal tubes as 20mm longer than the rods.
The frame is a rectangle prism. The twelve tubes are in three groups. They could be all the same but mine aren't.
My rods are 1000mm (call it X, left right ) and 700mm (Y - front to back).
This means the tubes in the X direction are 1020mm and Y are 720mm.
The footprint size was chosen to sit on a table I have.
I wanted the height to be low enough that the printer can be turned on its side and pass through a 820mm doorway. The Z tubes are 650mm.
The tubes were cut using a metal cutoff wheel.
The cuts were surprisingly square (meaning 90 degree right angles).
The ends were cleaned up both inside and out using a dremel.
The expansion plug.
The tubes are fitted with plugs which each house a nut. This spread plug when the nut pulled by a screw jammed the nut in place. This allows easy attachment into the corner blocks.
The nuts are M8. Metric nuts are more standard than imperial but some of the one I have don't fit together well - badly made I presume.
It is almost 50 years since we became metric and it is still a struggle to get a good range of metric screws here. Some 3/8 nuts and screws might fit the plugs - my local 3/8 nuts don't.
I was planning to leave the plugs recessed a few mm and use some sort of compressible washer to allow some fine adjustment on the distance between the Z tubes.
My screws were too short so I've kept them flush.
With 12 tubes there are 24 plugs needed - that is a day's worth of printing.
It is worth testing them before printing the lot.
Files.
This one is for 1.2mm walled tube.
http://www.thingiverse.com/thing:72164
This one is for 3mm walled tube.
http://www.thingiverse.com/thing:72164
Rods.
There are six rods - two lengths in my case.
They are 12mm hardened steel.
There are four tubes near the edge of the printer. These are travel up and down with the Z assembly but don't move in the X or Y direction - I will call these "edge" rods.
All the edge rods are at the same Z height.
The two other rods form a crude cross with the head assembly at the cross point.
I will call these the "cross rods".
The cross rods obviously have to be have to be at different Z heights - the are separated vertically 6mm.
The edge rods and cross rod for each axis are the same length.
The cross rods have a single linear bearing at each end which rides the edge rods.
The head assembly has two linear bearing ride on the X rod and also two on the Y.
Bottom corner thing.
Version one is pretty basic and will probably be upgraded at some point.
The photo should make it pretty obvious what it does.
What may not be obvious is the extra holes and divots.
The square/diamond shaped holes are calibration marks to check the part has been scaled properly.
My print was out a bit in Z calibration and I decided I may as well add these marks to check.
The captive nuts is not to lock the tubes. They are optional attachment for extra bits such as cross braces.
You could also use this for a screw facing outwards and the hole on the other side of the tube hole is for a driver.
If braces are fitted they need to be at least 40mm out from the frame and added this to the corner piece wold make it too big to print on my printrbot.
If Huey works I may use it to print a large version.
Have just a hole doesn't commit me to a particular bracing angle.
Bracing.
Bracing on the Beta Huey will be ad-hoc or absent.
To the eye the unbraced frame looks fairly true.
I don't have to decide on a bracing method now but 16mm steel curtain rod is a on the short list.
Files.
http://www.thingiverse.com/thing:311770
Belts
Huey beta uses T5 belts from beltingonline.com
Qty. Products Total
1 ea. T5 Standard Breco® Open Length Timing Belt
Width in mm - 6
Length in metres - 10
Nylon covering - Not Required
Tension Cords - Steel
Material - Standard 590 White PU
£29.06
Files.
http://www.thingiverse.com/thing:316821
http://www.thingiverse.com/thing:52155
May-30th - Z-bearing assembly
June-22nd - Go between PCB.
I was originally going to try running huey off a printrboard with no extra drivers. This is the reason for the small motors on the prototype. One drawback of wiring the motors in parallel is there can be quite a bit of mismatch between motors which work together to move the stage. This doesn't matter on the Z motors - for example the two parallel Z motors on the printrbot work OK.
To get X and Y at 90 deg would need some adjustment and they might not home properly at the start of a print.
Giving each motor its own controller allows automatic homing with accurate independent control of all motors.
It also allows bigger motors to be fitted if needed. Driver PCBs are down to $6 or so from places like dx.com so not a big cost.
The reason for the go-between is printer controllers such as the printrboard only have one driver per axis and only limit switch input per axis.
I need four motors and switches for Z and two each for X an Y.
The go-between sits between the printrboard and motors etc.
In normal use it will just redirect the step pulse to all the motors for that axis.
However it will be smart enough to block the pulses when the carriage is homed.
It also send the printrboard a "home" signal when all the switches for that axis are closed.
There is one go-between per axis (ie three boards).
The brains of the go-betweens are arduino pro minis - also about $6 from dx.
The job could have been done using fpga,cpld,gal pal or even a rom. But I'm very familiar with AVRs and pretty sure I can get sub-microsecond response from them.
Using a micro also allows things like timeouts to be added to the code.
Unfortunately printrboard doesn't have easy access to the signals I need so either I have to hack it or use something else.
(update - I used a teensylu)
Given time and inclination one could make a printerboard with more drivers and fork sprinter/merlin/repetier to use them. I have enough work cut out for me already without going there.
October-26th - Day 3 after 1st print.
I've been working through some problems - most are minor.
I made six usable parts so far but motion in the Y direction is not as smooth as on the X.
The parts are quite good but the Y roughness is a mystery I'd like to solve.
I've had to cool the motors and will upgrade them at some point.
For now I've ordered more 40mm fans.
See - http://nerdipedia.com/tiki-index.php?page=huey+log#October-26th
for more details.
November-9th - Z-bearing assembly self rep.
By eddie