Status of the gigabot

What is the current status of the Gigabot? When I came in the other day it was being used as a shelf more than an available printer.

Giving this a bump, as I would like to know the status as well.

Tim W.

Gigabot is currently working, but not very reliable. Mike, Ryan, and Myself have gotten it to print decently, but it involves a lot of fiddling and patience – too much. The work to get it repaired and upgraded has not been fully completed. However, if someone wants to take this considerable task on, then it could be working again. I’d love to see it working, and can help, but I haven’t been healthy enough to do much. Likewise, other members familiar with gigabot have been low on time and motivation.

The biggest single problem has been the bed warping. RE3D acknowledges this problem, and now makes a machined cast aluminum plate bed. We could use similar material, but save some money by sourcing the metal ourselves. Using mesh leveling has gotten some decent prints, but it’s not reliable with so much warp and such a large surface area. Adding panels or crossbeams for support and rigidity would be good too, as well as truing up some of the frame and mechanicals.

Ivan, Bill, Mike, and others have gotten us much closer to the goal of a reliable gigabot. However there’s a lot more to do, and it really is a big job.

Does anyone want to take on this project?

I’d suggest starting by finding a good bed, cast aluminum tool plate, steel, or otherwise. Something that won’t warp. Also, RE3D has learned from their own mistakes and looking at the changes they’ve made from Gigabot to Gigabot 3 on their website would be worthwhile. I don’t think their $2750 upgrade is worth it, though, as most of those parts / features can be implemented for much less money.

I’d start with the mechanical, as no amount of firmware can compensate for these mechanical errors.

Any volunteers? No one needs to be a 3D printer guru, just motivated and willing to learn together and autodidactically.

If people want me to try to draw up the forensics of what’s happened and what to do now I can. However, I’m not able to do much now as I am recovering from a combo of multiple thoracic surgeries and neuro-surgeries.

LMK.

I know very little about 3D printing, and as of right now the Gigabot is foreign to me. However, I would like at the very least to get involved with this project in some way, shape, or form. Where should I start if I want to get up to speed on this?

Tim W.

I’ll gather my notes and all the resources I have for you and put them on the wiki.

First off, though, check out re3d’s documentation and compare their info to our machine.

The first thing I would do is check all the screws, nuts, belts, etc. Make sure things are square, that parts aren’t worn out, etc.‎ That’s not super exciting, but makes a real difference when getting around to other repairs.

Mike and Ryan can both show you how to get it online, or you can wait till I’m on my feet (4-6 weeks) or I can lead you through it by remote.

You don’t need to me versed in 3d printers. I didn’t know anything about them until I started working on them, building them, and now designing them. The learning is the best part.

Email me off-list if you’d like my phone # / skype/hangout

Lorin



From: Tim Wagner
Sent: Tuesday, December 5, 2017 3:31 PM
To: Hive13 Hackerspace
Reply To: cincihackerspace@googlegroups.com
Subject: Re: [CHP] Re: Status of the gigabot

I know very little about 3D printing, and as of right now the Gigabot is foreign to me. However, I would like at the very least to get involved with this project in some way, shape, or form. Where should I start if I want to get up to speed on this?

Tim W.

I’ll compare what we have with what re3d has on their website, and also take a look at the Gigabot and see if there are any glaring issues I can find out from the mechanical or structural standpoint. Next meeting, I’ll try to get a hold of Mike, Ryan, or whoever could give me a crash(free) course on the machine again. I remember having gotten the rundown from Mike during a beginner course from him a while ago (thanks again, by the way, Mike), but I need a refresher.

Anything having to do with the electrical portion of the Gigabot is going to be the hardest for me to grasp right now, but let’s see if I can’t at least give it a try.

Tim W.

Count me in for fabricating mechanical parts. I’ll join the refresher Tim has requested. Just let me know when.
Dave Velzy

Last I recall, they had upgraded it to octoprint.

Dave,

Re3d solved the build plate issue in later models by using ‎cast aluminum plate rather than rolled. Of course, they charge $750 for the upgrade. Considering that 24"x24"x1/4" MIC 6 aluminum tooling plate runs $100 or so:

Do you think we could fabricate a MIC 6 aluminum bed to fit gigabot? A steel tooling material would work as well, but would reduce efficiency. What do you think? I’d rather not spend $750 on an aluminum plate with 2 lead-nuts – couldn’t we make it?

Lorin



From: Dave Velzy
Sent: Wednesday, December 6, 2017 5:47 PM
To: Hive13 Hackerspace
Reply To: cincihackerspace@googlegroups.com
Subject: [CHP] Re: Status of the gigabot

Count me in for fabricating mechanical parts. I’ll join the refresher Tim has requested. Just let me know when.
Dave Velzy

Could we find any thing like that at Garden Street? Scrap price is pretty good price, even if it needs to be cut to size

Given:
This is all based on zero first hand knowledge of the thing in question. So consider it a SWAG.
the aluminum deck plate material ie. cast vs rolled vs machined aluminum isn’t much of a difference. Reclaimed or otherwise…

The issue is stiffness/stability over time and temperature. ie. Assure a flat surface independent of temperature.
If you put heat on a flat plate it warps. Duh.

Could we add a layer or two to the existing plate, then make it flat and stiff relative to the attachment reference points?
Even, if the HIVE mill can’t do it, there are some very capable shops in a 1 mile radius.
I’m willing to explore a solution, weasel machine time from someone and come to a fix somehow.

First up, we need to out engineer Gigabot. Or is that re-engineer?
You’all who know the Gigabot and the problems need to give your insights toward the fix.

GAME ON

Dave

I wonder if we could go with a glass plate surface?

Agreed. The reason I suggest MIC 6 is not so much that it is flat and hard, but that it is annealed properly – the fact that it is cast is not really as important. Since the structure is stress-relieved, It’ll be more stable over a range of temperatures and less prone to heat cycling warp. Of course, I’ve done the same thing with tool steel, heating & quenching it with a torch. I can’t imagine heating a tool steel plate to 60 degrees C with silicone heater, though.

Another solution I’ve used, and which seems to be getting more popular is using thin spring steel as a surface, floating it above a larger mass Al to distribute the heat. A spring steel sheet is more easily replaceable or retempered than a big ol’ plate.

Of course, hand hammered copper would look handsome, wouldn’t it?

Glass would work too.

Can someone provide examples of the problem?
When problems happen have the solutions described on the Gigabot wiki employed?
http://wiki.re3d.org/index.php?title=Getting_a_good_print

Bed plate stiffness has costs. example; thickness, weight (cast iron?), complexity (composite)… Or is the cost simply time?
The wiki gives a solution to add a printed layer under the object and then build the layers slowly enough to allow for the existing bed plate to stay flat. If a printed layer is allowed to cool before printing the next layer the stiffness of the bed plate will overcome the contraction strain from the cooling of the printed plastic. The trade off is time.

Has this method been tested?

Has anyone done a stress analysis of the forces induced by contracting printed objects to determine how stiff we would need to make this puppy? It’s been 40 years since I was a grad student in engineering mechanics, so I’m a little rusty, but this stuff is still relevant.

Anyone on this thread going to be at the HIVE on Saturday? Could we talk this over?

Dave -

The build plate is cupped by a good few millimeters at room temperature. It’s enough to cause print head crashes during homing if Z height is set at the center, and setting Z at edge height results in air-prints toward the middle. I haven’t fired up the heat bed recently enough to remember if or how warming flexes the bed.

Bill compensated for this in firmware some time ago, but there’s constant hazard of dragging the print head when someone unfamiliar with that quirk operates the machine.

We we do have a bed-size sheet of glass down there, though it’s possibly too thin for safety.

I can be down on Saturday if you’d like a 3D printing primer. Be forewarned that I’m only physically present in the AM though. What times work for you?

  • Ry

I know people have done a glass bed on the smaller printers to get a flat surface. Some people use thermal pads to transfer the heated bed heat to the glass. Not sure if it has been done on the Gigabot scale though.
The new surfaces for the print bed (PEI) are awesome.

Could we get the boron silicate glass and mount that on top of the aluminum plate already there? Possible a really thin sheet of silicone much like what i have to order on alliexpress would give it traction and hold down clips could be easily made.

On a second thought as well i use the high temp silicone rtv gasket maker on all sorts of engine parts and casting to fill in voids. Could the glass be mouted via an ultra thin coating to account for deviation? Then to remove it, thin pianno wire worksd great!

A glass plate over the bed is a possibility–that’s how I printed a few things on the gigabot earlier this year, although it was a finicky setup.

Mounting the glass with some sort of thin coating and/or deformable thermal pads would be ideal to allow good heat transfer across the platform.
The other issue is that the gigabot has a capacitive sensor for finding bed level, which does not function well through the glass. Lorin suggested getting a more sensitive sensor as one solution…
Overall, I’m not sure if it makes more sense to go for glass over the current bed, or a spring-steel metal bed.

I haven’t been around much, but I have a break coming up 17th through the 29th. Maybe I can help get it fixed up during that time.

Mike

a) The current warp is really pronounced – it can be seen with the naked eye. +/- 5mm is too much when the print head is supposed to be 0.1mm off the surface. It does result in damage. There may also be cant issues between the two lead screws, etc. It needs to be evaluated as a whole (i.e. is the frame square, motor mounts, etc). It all adds up to a lot of error. There may also be bowing or twist on the gantry… Poor gigabot’s crashed pretty hard a few too many times…

b) Firmware Z compensation is extremely limited. The algorithms are really only designed to tram the surface and account for small deviations (<1mm). The large surface area compounds error by an order of magnitude. Bill did a good job hacking marlin to the extreme, but it has limits nonetheless. Ideally, you want to tighten a printer up so that it can work without z-compensation, and then use z-compensation to make the fine adjustments to close the gap.

c) There are sensors that work on glass. Bl-touch, mechanical probe, differential IR from filastruder – they all work fine. I use the diff IR on one of my glass beds and it’s quite accurate in most scenarios.

Really, it’s best to be moderate. Folks on the reprap forums often argue that you need NASA grade tolerances and rigidity. You don’t. But, everything should be reasonably square when measured with basic machinist tools.