Construction concerns

Gary wrote a detailed comment to this post, in which he described his experiences and frustrations with the mechanical side of building his Prusa kit. Rather than respond with another comment, I thought I’d write a post, covering his points.

Gary, I feel your pain ;) I’m no mech-eng, but I also had concerns about alignment. I think I might even have blogged about not being able to see a way of squaring the frame up precisely. And yes, tightening the nuts does shift the rods, it’s a trial-and-error proposition.

One thing that helped me during my build was a comment (on IRC, I think, which I don’t have time for now) by Joseph Prusa, who designed the variant we are building. He said it surprised him how many people decided to change his design before they’d even finished building it!

He reckoned (and I think he’s right) that there was virtue in building it as described, testing it for a while, and only then deciding which bits needed improvement. After all, the mechanical side of things is easy to disassemble and rebuild. The Prusa variant is one of the most popular RepRap designs at present, so it must work ok for most people.

Here are my thoughts in response to the issues Gary raised:-

Axis Perpendicularity

To the degree that your X and Y axes are not at right angles to each other, you will have exactly that degree of ‘parallelogramness’ in your final print, and increased friction. The software as written doesn’t correct for that.

So it is worth getting it as accurate as you think you will need your prints to be. While the stepper motors can position the print head extremely precisely (microns?), we’re printing with floopy extruded plastic about a third of a millimetre across (on my machine, with a 0.35mm nozzle). So the dimensional accuracy of the finished parts is not up to, for example, race-car engine turbo-charger tolerances anyway :)

Rods Parallel and Co-Planar

If the rods aren’t parallel, you’ll feel it as friction as you move the moving parts (extruder on X, printbed on Y).

I set one Y-rod exactly perpendicular to the frame and locked it in place. Then running the bed back and forth positioned the second Y-rod precisely parallel to it by feel, before I locked it in place too. Test and repeat, several times :)

For the X-rods, it is a similar process. I just stuffed the rods into their friction fit holes in the X-ends, and ran the X-carriage back and forth till things ran smoothly.

MakerGear’s printed parts are well made, so the alignment is pretty reliable. The bronze bearings take some bedding in before they get smooth, so don’t panic too much if they feel a little stiff at first. Just keep sliding the X-carriage from end to end til things loosen up. I did this both before and after mounting the X-end assembly onto the Z-rods.

When I tried to tighten the fixing/alignment screws to lock the X-rods in place, the screws pushed the rods out of alignment again, so that the X-carriage got sticky again. I eventually realised that there are exactly two places on each x-end where fixing screws directly oppose each other. When I gently advanced each screw by the same amount a bit at a time, I was able to lock the X-rods in place without shifting the alignment.

I never found a way to adjust for (or even measure) co-planarity, but it seems to work fine anyway :)

Bed Parallell to Plane of X-Rods

The bed levelling process (which you don’t do til all your wiring, motors and software etc are working) seems to deal with this.

Friction in Z-axis Threaded Rods and Nuts

I envy Gary his ability to improve the quality of his threads. I found the biggest problem I had with the Z-threaded rods was that the hole in the bottom of the X-ends was too small. The threaded rod would pass through it, but only with a lot of friction.

Since the point of the mechanism is that the captive nuts be the only part contacting the threaded rod, I reamed out the holes in the bottom of the X-ends so the rods wouldn’t touch the plastic. But not so large that the spring jams or falls through.

What Is Critical, What Is Not?

The instructions are, and always will be, a work-in-progress. That said, the more details in the instructions, the more likely that bit is critical. For example, getting the X-ends level on the Z-rods is critical, because otherwise you get binding that leads to skipped steps on one Z-motor but not the other, which leads to a positive feedback loop of increasingly mis-aligned Z heights.

The other clue is how many people have written about their own problems with a particular part of the machine. That will never be in the main instructions, but it makes me feel better to know that I’m not alone in my difficulties! And I benefit from their solutions.

Wiring and Electronics

Gary didn’t mention this. Maybe he’s smarter than me. The electronics stuff worried me more than all the rest of the build put together. Not because it is difficult – it isn’t – but because a stupid mistake can destroy a 50 cent component that I have to order from overseas.

I have 1970’s hobbyist level electronics experience, so soldering holds no fears for me, but this surface-mount stuff taxes my old-eyes, shakey-hands co-ordination. I paid extra for the pre-assembled RAMPS 1.4 board. Best $50 I’ve spent in a long time.

Would you believe my biggest mistake was accidentally tearing the insulation on a 3 inch long piece of high temperature wire? MakerGear Rick said to wrap it in some kapton tape – problem solved. Imagine if I’d had to wait 3 weeks for a single strip of wire to arrive from the other side of the world! MakerGear tech-support Rocks!

So take it slow, and test anything you can. There are a number of helpful guides on the web. Can I claim that my posts are helpful?

Software and Firmware

No worries. Seriously. You don’t need to program at all, in any language, to get a Mendel Prusa printing. If you can edit a plain-text document and install computer games, you can do this.

You will have to install a number of pieces of software. Arduino, Slic3r, Pronterface, & Sprinter would be my recommendation for the initial setup. And  you’ll need to edit some numbers in a couple of places, but it isn’t rocket science, and nothing catches fire!

Fire! Fire!

Well. On the subject of fire. Don’t turn your stepper drivers up too far – apparently they can catch fire. And don’t try to heat up the nozzle or bed until you are getting correct temperature readings from the thermistors, or you can melt your hot-end. And don’t home your axes until you’ve turned down your homing speed or you’ll scare yourself. All of this gets covered in more details in the various step-by-step instructional posts.

 

About these ads
This entry was posted in 3d printing, Assembly and tagged , , , , , , . Bookmark the permalink.

2 Responses to Construction concerns

  1. lateralg2 says:

    Andrew,

    Terrific reply and much appreciated.

    My biggest problem with X Axis was center distance of slider bushings being greater than center distance of holes in X end motor & idler parts … even with outboard adjustment at maximum. This necessitated enlarging the holes in the printed parts.

    The CD difference was great enough to require forceable entry of the slider-mounted rods into the printed part holes. I’m not looking for sympathy; just want to communicate an issue. Because of the many hours consumed on this, I encourage future builders to check slider action at Part 7 Step 8.

    It’s a huge relief to hear that software shouldn’t be a problem. I now look forward to that challenge.

    I’ve got the X axis slider to an acceptable level of friction. This is partly due to remembering the following quote:
    THERE COMES A TIME IN EVERY PROJECT WHEN WE MUST SHOOT THE ENGINEERS, AND BEGIN PRODUCTION.

    Before I leave, I need to make it clear that I consider MakerGear, their people and their products, to be first class. I will buy my next 3D printer from them.

    One other thing: I’m aware that the state of the art in 3D printing is much like that of personal computers in the ’70’s. I remember the difficulties sending printing commands to my Epson Dot-Matrix with my Franklin Ace 1000 Apple clone (but having 64K of RAM … double that of Apple) .

    Be well, and do good things. And remember: “If it ain’t broke, you’re not trying”
    Gary

    • BrazenArtifice says:

      Ok, Gary, that does sound annoying – might even have been a show-stopper for me. I’m surprised that the centres of the X-end and X-carriage holes could be that far out of alignment. It might be worth dropping MakerGear Rick a note about it. Not everyone has your ability to fix a problem like that. My parts line up fine, especially given the self-aligning nature of the bushings. I wonder if different batches off different printers could be causing problems at MakerGear.

      It is true that without having seen and touched a working example of a Prusa it is hard to know what level of friction is OK. I’ve only seen exactly one printer other than mine, so I don’t have much of an experience base to draw from.

      With the care you are taking with the build, you’ll have less trouble than I’ve had with layer offsets, and you’ll be able to run your stepper drivers at lower current, all of which is good.

      I have vague recollection of Apple clones. I started off on a TRS80 with 4kB of RAM. The tiny $40 arduino microcontroller on my bench (hooked up to a world war two era morse code key and a tiny LCD display) has more power and RAM than my first 2 computers put together, and is faster than my original 1980’s macintosh.

      Just keep at it. It is such fun when it works. Have you started collecting cool things off thingiverse to print when you’ve got the beast together?

      I’ve just printed off three sets of Tardis cookie cutter stamps without feeling the need to change anything from print to print. Must be getting closer to optimum settings. Pity I’m just about to have to change to a different coloured PLA, which will need different trial-and-error temperature settings, at least :)

      Thanks again for the comments. I didn’t do much 3D printing when I couldn’t write up my results due to lack of blog access. Knowing people are reading my words really helps.

Please leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s