Very nice! I plan to eventually modify my voxel ox into this setup. I opted for my Z screws to go at the sides though, similar to a previous build on here and at first glance I couldn't find a closed loop belt with the dimensions I need. This is my first 3D printer though, so I will probably revisit the idea once I've smoothed over my current printing issues. Out of interest, what size is the belt you are using?

On another point, Absolutely loving this printer. I've had minimal issues in setting it up, which is in no small part to its brilliant design. The only issues I've had have been made by myself to be honest where I have tried to skimp on expensive openbuilds parts. Unfortunately we pay a hefty shipping rate in the UK to get V-slot and simple things like the wheels soon becomes one of the greatest costs.

 

I havent finalized the size for the closed belt yet. I will likely add one more idler to help tension it.
Robotdigg sells many sizes of gt2 belt loops

 

I have run into binding and out of sync issues too many times with dual Z motor. thats why id prefer to go down to one motor. also this way I have a 2:1 ratio with the gears on the Acme lead screws vs the Stepper to help resolution / torque

 

most of the time. its fine. but often with microstepping the steppers are not moving full steps, and one might miss a step once in a while for some reason. these errors slowly starts to build up.
and you need to realign the Z's to get the X Gantry Level again. this is also a problem if you need to manually lift the Z with power off. and you are turning two motors manually.

 

Hi to everybody.

First of all I would like to THANK Marshall for this great design.
I was totally hooked and I am building one "OX" right now ;-)

I am living in Windhoek (Namibia/Africa).
Unfortunately even simple parts, like the flat head screws are not locally available, if at all.
The next "Wall Mart" is about 12000km away ;-)

Therefore I had to order everything from the store.
It turned out that shipping costs are as high as ~30-50% of the material costs!

Therefore, instead of going for a specific built volume ,
I opted for all V-slot extrusions with a length of 500mm and guessed the size build size to be about 30x30x30cm.

As I do not have a nice workshop ( cutting Alu was also not really an option)
I assembled the frame as it was delivered (~505mm cutting length)

After a few hours of trying out different things,
the ONLY thing to cut, was the x-axis to a length of about 460mm

I ended up getting the following moving ranges
X: ~280mm
Y: ~220mm
Z: ~400mm (this will be shorter when the build plate and hot-end is mounted
I expecting about ~320mm)
Y was way too short of what I expected!
So I had a closer look at the y-Axis design and found that the
motor and the idler mount restricted the Y movement by at least ~75mm.

The first thing I changed is the stepper bracket and idler mount.
Using the design from "Folgertech Prusa i3 2020 Y-Axis Upgrade".
Folgertech Prusa i3 2020 Y-Axis Upgrade (Openbuilds V-Slot Rail System) by sonnylowe

This gave me additional ~40mm but still short of the aim of ~300mm ( 12")
Unfortunately the belt tension-er was not designed for the big build plate and I could not find another design that works.

Then I saw the cut of peace from the x-Axis and extended the y-axis with it (not wasting any extrusion!)

To mount the extension with the parts I had at hand, I mounted it with an offset
I only had to cut of about 20mm on one side to make room for the motor cable/plug.

I also used this part to guide the extension:
OpenBuild V-Slot adapter for Tontec Raspberry Pi2 case
OpenBuild V-Slot adapter for Tontec Raspberry Pi2 case by MartinMajewski

This gave me additional 30mmm !!
At this stage I ended up with a moveable Y-range of about ~290mm

... but now I still had the problem with the idler and belt tension-er mount.

I went ahead and designed a new Y-idler part,
incorporating the ideas of the original design.
This included the belt tension-er

This gave me an additional 30-40mmm
The total Y-range is now at max. 335mmm
This is ~50% more than originally almost ~115mm more!!!

This should give me enough room to mount dual extruder "with a Y-offset"
and still reaching the whole print area with both heads!!

I also included a mounting space for the "Y-stop"
The mount is designed for Omron Microswitch "SS-5GL-F" (hinge lever - NO roller)
http://www.omron.com/ecb/products/pdf/en-ss.pdf
The micro-switch is mounted using M3 screws and I also cut the thread into the switch and block.

I ended up with the following design that I would also like to share.
So please feel free to use it ;-)

I should have stl/ iges & step files included

All holes are designed with threads inside except for the M3-tensioner screws and the bottom 5mm holes.
but i did cut the threads again.

The bearing screw (M5x30mm) is not using a nut and just using the thread in the block

As mentioned above keep up the great work and feel free to use this design ;-)

Best regards from Namibia and keep up the great work
Jurgen



... I still have to think about the x-Axis

 

Jurgen- OUTSTANDING SIR! For somebody who doesn't have access to a fancy, well equipped shop, you are doing great work. I'm frankly embarrassed to think about my tool driven building techniques and have never considered what our hobby would be like without them. I for one, am greatly humbled.

I hope you will continue to keep all of us informed about your build. I'd urge you to start a build log here on the site and share the whole process with us all. It's just my opinion, but I wonder how many people are stopped from ever attempting a build because they don't have a lot of tools. Your efforts can inspire us and them.

Best wishes my Friend
Stargeezer
Larry

 

I have posted my Work in progress Voxel here. for those interested
Voxel 3D Printer on thingiverse

 

I have had my "OX" assembled for about a month. I bought a Smoothie Board for a controller. I thought it was going to be a great set-up, but both boards had thermistor reading problems. Nobody sure what caused it, but Smoothie took both boards back. Now I am back trying to decide what board to use. I really liked the Smoothie ease of setup and nobody else has had thermistor problems, so I am not sure what to do next. Any opinions???

 

Is the thermistor wired via the terminal blocks?
If so then you maybe think of routing the cable away from the heater cables and twisting the thermistor wires.

Reason:
The printhead is heated via PWM meaning ON and OFF switching.
This will cause high frequencies of all kind!
If the wires of the heater and the thermistor are close together than the thermistor wires may act as antennas and pick up that high frequency noise...

 

I have my cable management planned out and the terminal block is temporary until I can print my cover and other parts. The boards were returned and I am looking for an alternative controller.

 

I do not think that another controller will solve the problem because the input settings will most likely be the same
The thermoelement has it own IC doesn't it?
The wires will still pick that noise!
I would try one of the cheap normal 100K thermistors first.

 

... thanks for all the credit
Building was so fast that I forgot to take pictures.
I did some printing already.. but... I have the "first layer problem"

I am trying to print without any heated bed because I have to run the machine from a UPS.
With the heated bed on and a power consumption of 450 Watt, my UPS won't run very long.

I was printing on BuiltTak so far but I did pledge on kickstarter for FLEKS (http://www.fleks3d.com}
FLEKS is already shipping for 5 weeks so it should not be too long any more to arrive.

BuildTak is not as sticky as I wished (at least not with PLA and NO heater).
I used wood-glue mixed with water and sandpaper to get better adhesion and that worked surprising well.

Currently the heater is no problem as the inside temp s still above 28 degrees at night time ;-)
..but winter is approaching and it will get cold.

I also received a BLtouch sensor this week!
So far I used a capacitive sensor for bed leveling ....but my problem is the sensing distance. (18mm Sensor! = max. 3 mm Unless one is printing direct on aluminum, the distance is too small
...but with 2mm glass on top of the alu, the sensor is just 2-3 10th above the nozzle and if anything warps the sensor smashes into the print!
Maybe glass only with some steel washers would work better... hmm

So far I have used a GT2560 controller and have used marlin 1.1.RC 3-5 firmware but I do not get the automatic bed leveling right.
..... anybody else having these problems?

I am now trying to set it up with a MKS-Sbase V1.2 controller which is smoothieboard compatible and only costs about $50!
I have already identified how to connect the BLTouch sensor ;-)

So far I have used a bowden setup and an E3d V6 clone ... but I am not happy with this.
So I will change it to a "Wade geared Direct drive extruder" as soon as I manage to built a hobbed bolt.
...so much for today

cheers and happy building ;-)
Jurgen

 

The print quality of that sturdy frame is amazing ! im planning on building one aswell but im wondering what would be more accurate between a all lead screw vs all timing belt design ? would more accuracy be gained by building the same frame out of CBEAM Actuators ?

 

First good print. I have a Smoothie board and Prometheus Hot end. I have never printed anything before, but this is ABS and a idler for my printer.

 

Hey guys, I've been using my Voxel OX for a while now, so in the open source spirit I decided to upload my own modifications for this printer. I'll probably upload more in the future, as I've got more ideas for improvements such as bed mounts, adjustable endstops, extruder designs etc.
Find my designs at Youmagine here: Designs by Josh on YouMagine

Y axis motor bracket by OrigamiB posted May 13, 2016 at 12:45 PM

x axis idler and tensioner by OrigamiB posted May 13, 2016 at 12:45 PM

Z axis gantry plate + lead screw mount by OrigamiB posted May 13, 2016 at 12:45 PM

Y axis motor bracket by OrigamiB posted May 13, 2016 at 12:45 PM

 

Nice find with the RobotDigg belt sourcing. I did the single motor z with a homebrew Prusa built from plywood scraps & junked inkjet/scanner steppers a while back and it worked great even with an insanely undersized stepper. Marshall hooked me up with an economical hot end and the whole printer came in under $50. (teacup firmware on a $4 duemilanova clone, $4 engraver shield, four $1 stepsticks and the junked inkjet ps)

Riley Cassel - The black blocks are the x & z axis | Facebook

Riley Cassel - it was going well until the 35v power... | Facebook

 

Hi Marshall, seen your prints using Chimera!. Looks great. I am slowly collecting bit by bit of what I need. I also want to use E3D Chimera. Can you publish the STL files which you use for Chimera please?

 

HI Marshall, I have another question for you. I am considering of using Anti-Baklash nut block, would you recommend ?

 

Anti backlash nuts are only really needed in CNC machines, where cutting resistance can push the X-gantry upwards with the backlash. Since there is little resistance in a 3D printer, gravity will take up the backlash. Anti backlash nuts arn't really needed and will give little, if not zero, improvements as far as I understand. (unless of course you are using screws in X or Y, then there may be a good reason for using them)

 

Thanks. I was thinking using this machine for occasional CNC use. On second thought, I may build a separate one for CNC use.

 

I have noted you have moved the lead screws to the side. Can you also publish a photo of the x axis from behind please? I wanted to see where you have mounted the motor.

 

IMAG0757 by OrigamiB posted May 30, 2016 at 10:47 AM

I just use an openbuilds plate, but placed on the side. I'd suggest this over printed motor mounts any day. My Y axis motor still shows some signs of bending, despite being very chunky. I think you could use a printed mount, if you could support the motor from behind (eg, by the rear mounting holes, the body itself, or if you have a motor with a rear shaft), or by using a small bearing on the end of the shaft at the front (assuming you have some protruding out, I don't).

The high tension will always causes pivoting around the mounting holes in a plastic mount, the metal openbuilds mount stops this nicely, and much smaller than any printed mount

*EDIT* The silver bar at the front is for bracing, as the 20x20 rails will spin around the tapped end. I simply used a leftover spacer from the Z-axis motors, and some leftover fixings.

 

Thanks. That's pretty good you moved the belt to the middle. Could not you move the 20x20 to the middle of plate and brace it using L plates on both sides of 20x20 ?, given that you have the freedom as you don't have the 8mm lead screw obstructing it. Understand it will also push the print head by 40mm towards front. Just wondering; I am sure you had your reasons.

 

You could, yes. This is actually something I'm considering doing, as otherwise you can't place too much tension on the screws holding the wheels and the gantry. Using separate screws would solve this instantly. I don't have any of the low profile openbuilds screws though, and so I'd have to add a spacer under the gantry plate so the screws don't catch on the extrusion. This method works for me for now, its all a work in progress

I'm also interested to know if there is any noticeable bend in 20x20 extrusion, I can't say I've noticed personally. normally this shows when you level the bed, as it appears the bed will bulge in the center (when really its the x-axis bending towards the bed). Perhaps this is only an issue at very small sizes, but to me this would still be an issue as I've been trying to push the printer to its limits. so far I've achieved a very nice 40 microns layer height, which is probably as far as you can go with the acme leadscrews and a 0.4mm nozzle.

 

I am planning to use the 20x80 just like Marshall's original build. But keen to move screw to the side either using your STL file ( as I also have cheap Chinese lead screw nuts) or openbuilds second plate with spacers. Other thing I was wondering to support the lead-screw from the bottom. Just can't imagine having one end exposed. It probably would not do any thing at all!!.
You have probably noticed I am new to this. I am more than happy if I can 40 microns . That is awesome!.
Major problem I would face is hooking the electronics and calibration. I purchased cheap all in one RAMP 1.4 compatible board. No idea how to go about it. Did you use repetier? Any information is helpful.

 

I'm not sure how much difference it makes to use bearing blocks to bind the lead screw. I've actually got them on my lead screws now, and the best thing about them is that they allow you to get the screw really straight and true. I may just use them as a guide and then remove them later on. Problems occur if you use cheap screws that are slightly bent, and using a bearing block to trap the floating end of the screw will cause it to bend in strange ways. this is why a flexible coupling is used, to allow for a bent or non-true screw to move around. I'll upload a picture soon of my bearing blocks

*EDIT* added pictures

IMAG0758 by OrigamiB posted May 30, 2016 at 1:46 PM

 

That is

That is really cool!. Thank you. I didn't see the STL files for bearing blocks on Youmagine. Can you upload them please? If you don't mind, can you also publish the STL for extruder, if it is different to Marshall's.

 

I'll add the bearing blocks at some point today, The extruder is also still a work in progress, So I won't be uploading it just yet. It is a direct drive extruder, using no gearing. I have never liked the idea of printed gears.
I've been using this now for roughly 3kg of filament, so it is nearly ready for release I think. It's based off the direct drive extruder by ffleurey (Direct-drive hinged extruder for E3D/J-Head hot-end (Prusa i3) by ffleurey), but heavily modified and almost completely redesigned to be stronger, more compact and easier to disassemble for maintenance.

*EDIT* added a picture of V1 of my extruder. V2 is in the works at the moment

IMAG0759 by OrigamiB posted May 30, 2016 at 4:14 PM