EEVblog #327 - Makerbot Replicator Troubleshooting

Dave fixes his Makerbot Replicator, and rants about a few things....

EEVblog Main Web Site:
EEVblog Amazon Store:
Donations:
Projects:
Electronics Info Wiki:

[EEVblog]

After two prints out of the box?, and with both grub screws completely falling out?

[EEVblog]

How did I screw up exactly?
Do you think it's acceptable that the grub screws should fall out of an $1800 commercial printer after two prints? Was that my fault somehow?

[janbarthelmes]

Beeing an engineer myself, I agree, that this not a great solution. There's a couple of better methods (of whom I do not regard adding some loctite as far better). However, give thema break, it's still a nice machine - hope they learn from it. Mechanical engineering is not a piece of cake. Those of you (engineers) who never made any mistake, cast the first stone :-)

[johneastmond]

In the U.S. we call a fastener that fastens a collar/gear/sprocket/similar to a shaft with out a head then we refer to them as set screws. "Grub screw" is a fantastic term! In South Africa (which is a British colony) they also called them set screws.

[jondoe]

As an owner of a rep 1 and a self-built prusa, I have to say the replicators are not as plug and play as people think.  Dave's a smart dude and even he had issues in the beginning.  But I will say, once you get them dialed in and make a few tweaks (i replaced the filament guide with a self-printed, spring-loaded guide) they work great.  Mine prints 100% flawlessly three and a half years later.  Still going strong.  I wouldn't buy another makerbot though.  The amount of troubleshooting I put into the MB was equal to the amount of time it took to build my prusa from scratch. 

[treborrrrr]

"Homing" the machine means that all the axes return to their default position. In the case of this machine that means that the print platform is raised until it barely touches the print head and then it parks in the back right position of the machine. To detect that it has reached the correct position it uses the limit switches which open when the printer bumps into them. After homing the software has a known position to work from, since there is no position feedback from the motors this is key.

[nikushim]

One major issue i have noticed with these machines from a CNC standpoint, is the belt drives (the speed advantage is not required for that amount of travel, and you will get much better accuracy with a screwball). There is also the complete lack of motor feedback, or any other from of position sensor (outside of a maybe a optical interrupt acting as a stop point).

[FXGreggan.]

Someone has probably mentioned this but if not - the cogs are _supposed_ to slip from the axis so not to break anything if there is a mechanical blockage...
But the screws should not come undone so loctite the suckers in place :)

[EEVblog]

Well, to be fair, I did actually find the problem quite quickly. But others may not fair so well if things go wrong.

[maikeydii]

Problem here is that the shaft of stepper is hardened steel and the pulley is aluminium. It would require so much force from the set screws to bite into the hardened shaft that the thin wall of pulley hub would probably brake down. Putting medium hard locktite to setscrews might help the problem of screws shaking out a bit but to be sure I would also apply some bearing glue to the stepper shaft and assemble the thing.

[McGeias]

I'm currently, and have been for some time, been working on a proper milled version of a reprap printer. With proper ball screws and nuts for accurate power transmission. Also using gecko stepper driver hardware for reliability. 3D printing like this isn't available for the mass market yet. Just like cnc mills and lathes are not for your average hobbiest. Perhaps they will be soon. It looks like they just might be heading that direction.

[jaapweel]

There were oodles of different new 3d printers at Maker Faire this year, ranging from the lasercut plywood variety to stuff that looked almost like consumer-grade tools from Home Depot. It would be really awesome if you could somehow finagle samples and do teardowns/reviews on some of them...

[EEVblog]

Yeah, the uCurrent case came out excellent, and many things to follow I'm sure. Within it's limitations of course.

[treborrrrr]

The machine did exactly what he told it to do, it homed the print head. Doesn't matter if it's a one million dollar industrial machine or a homemade machine, that's exactly what it's SUPPOSED to do in that situation. It's like saying that putting a car in reverse isn't supposed to make the car move backwards because the car was home built. If he read the docs (which he often admits that he doesn't do) he would've known that there is another option, jog, to move the head out of the way manually.

[mydbase]

Due to vibration, shouldn't these be secured using e.g. Loctite?

[EnergeticAdvantage]

Cheaper than trying to spline the shafts would be to drill a small indentation under where the grub screws go. Could even do that as a running fix if you can get it to an accurate enough drilling rig.

[EEVblog]

I tried to get a commercial $1K one, but no response from the company.

[McGeias]

As far as the lug nuts comment, check your owners manual for your car. Your tires should be rotated ever so often. Usually more than once a year soooo I'm guessing your lug nuts would get re-torqued. Kinda invalidates your metaphor but I do agree with the poor engineering of having grub screws on a polished chromed super low friction shaft. That's a hell of a fail.

[Andrew_Sparrow]

The fact it comes with a repair kit says something :) Imagine if your inkjet came with tools!? for the price and the output quality it's not a very good advert for this thing... there have been years of manufacturing with this kind of hardware (printers/plotters/etc) the expertise is out there and tried and tested, they are not breaking new ground with this. I was building this kind of thing at school 20years ago. I'm wondering if I would even buy one if I saw it in a store for $100 probably not

[siliconwitch]

How did you stop the print curling from last time?