While going through numerous posts about accuracy/flatness/wiggle/rocking of the work surface, it seems like there is a potential benefit for a platform replacement that would directly connect to the SBR16 rails and to the base of the machine. Keep reading
Disclamer for the GIF
The old version of snapmaker platform is used for illustration, I’m aware that new snapmakers are shipped with updated design platform.
This is a quick draft to showcase the idea and start working in the direction of production.
I would buy this!
No, I don’t need this.
The problem/solution illustrated in following topics
Current solution - attach SBR16 rails (or more robust version) via 3d printed shims to the platform.
3rd Party involvment
I’ve been in touch with a few companies who produce custom made parts for 3d printers like Voron. Some of them are very interested in helping our community by producing custom parts, like platform replacement, stiffness braces ect. Things that will hugely benefit from precision cnc.
Above I’ve attached a quick sketch of platform replacement that would be machined to order.
Why is this a good solution (IMHO)
Plug and play (only work is to attach rails to the base of the machine)
No 3d printed part, no variance in dimensions at 4 corners.
Full metal platform will be much stiffer on it’s own.
I would like to hear some feedback from interested members. This will help to understand the demand for something like this and potential modifications for the final product.
Should the SBR16 rails be included in the kit?
Best way to attach the rails to the machine base.
Would be great to hear from Snapmaker @Riskey on the upcoming quick change release design, if you could share details on how it is going to work with current setup, in case we have to adapt this platform for it somehow.
If you have other ideas for a 3rd party mods - please post. I would love to see bracing designed by @Elliot in metal. Hopefully this is something he would like to see too
HUGE Credit to @stefix for the fusion360 files of his Snapmaker A351
Yea. I’d like that if it was made. Probably would be worth a kit with everything. But, maybe allow people to buy just the bed? Special prepurchase thing or something.
As for quick release, I dunno a smart way to do it. If you had the holes in cad you could make pillars up where the screw holes would be and then use a lever with a tensioner on the front to lock the beds in place.
One issue I see at first glance is that having a direct connection to the SBR16/support rails is that it will cause any flatness variance in the base plate to be directly transferred to the platform, and that amount can vary from one machine to the next, and even from one machine being moved between different table surfaces. To complicate it even further, it can also change based on the position of the bed between front and back (one of the biggest pitfalls to the “bed slinger” design). One of the benefits of the 3D printed adapters is that you can adjust each corner individually (prior to printing) to compensate for quite a bit of that. In using a solid platform directly mounted and machined to a flatness tolerance, it would be good to have something like an adjustable “screw-in/screw-out” levelling foot. Keep in mind, the updated stock v2 platform for the 2.0 has a manufacturing tolerance delta of 0.15mm (0.0059in), but once attached to the Y-rails is pulled out of tolerance to between a 0.4 - 1.0mm delta. We would need a solid platform to have some way of adjusting for this per individual machine, otherwise we would still be left with the issue of having to shim between rails/platform/bed/build plate.
In some extensive bed level testing I just finished with, I found that the single biggest error in the bed delta came from the removal and re-install of the heated bed. With both a solid platform and a quick-change system, I would want them to remove as much of that error as possible. If they don’t, then to be brutally honest, I personally wouldn’t have much interest in them, as my current goal is to get my heated bed as flat as possible and then never again remove it, just attaching the laser and CNC beds on top of it. This is one of the pitfalls of this being a 3-in-1 machine; with a dedicated 3D printer you could “set it and forget it.”
It is a great idea though, and I am definitely interested in the development of it.
With a shim set and very good detailed instructions i could see it working. But like Mxbrnr said its allot of ask to lock everything together with out adjustments. since we cant guarantee the base to be any better then the current rails.
Personally I haven’t worked on flatness or z height yet but have gotten the x-y down to sub .05mm last cube was sub .03mm. But im finding the hardware is decent. Its the setup that is the hard part. Since its a make at home kit. It’s hard to control all the variables.
After thinking about it i could see it working if you where willing to give up say 20mm of z height.
The kit would be a flat ground base, the 2 linier rails and 4 trucks, the new platform. The customer takes off there current y axis and assembles them in this new kit. Which has precise locating tabs for the existing y-axis drivers. Then the whole new kit is assembled into the existing snapmaker. Then an aftermarket source could set the plate flatness and tolerance on there machines / systems with the rails installed. Sell it all as a matched kit. But that’s probably way to spendy and over kill for a 1500 dollar machine.
But if i was doing my own with my spare time at work that might be the path i go down.
The pre-cut rails as part of a kit are a good idea. I had to use an angle grinder on mine.
Selling just the carriage/base/bed, pre-drilled for the supported rails, should be an option. Not only is it a cheaper ption for those who have already done the rails mod, there are those such as myself who do not use the stock enclosure, and will have to cut the rails regardless or would in other ways need some flexibility in regards to this component.
I don’t think warping of the base is great problem. The rails are pretty stiff - in fact, the SBR-16 rails are of much more sturdy stuff than the base, and if anything will draw it into flatness. But even if they didn’t, this would be no worse than the 3D-printer versions, as the corners of the bed that are adjustable will be travelling over the uneven base and rail - so their adjustability becomes irrelevant, as what they are travelling on will be of uneven height.
I think the argument for adjustability at the corner mount points is a good one, though, for different reasons. It may turn out that the bed warps after a few heat cycles, or perhaps the linear module carriages it mates to aren’t machined to sufficient tolerance, or whatever. A length of M6 bolt with two jam nuts, one against the bed and once against the rail carriage, would serve well.
Thanks for reminding me of the Fusion model. I have been meaning to build one in FreeCAD, and if nothing else I can pull the measurements from that one.
This does seem like an interesting project, I had a few ideas myself for adding rigidity and helping to ensure flatness, but never got around to trying it. While I do CNC stuff professionally, I don’t have the home equipment to get down and dirty with prototyping.
I would say, given the potential deflection of other parts (namely the X axis, Z axis), or slight out of square (X axis left/right mainly) some form of fine adjustment on the four corner mounts would definitely be a good idea. I saw the suggestion of shims, which is a good start.
Something that comes to mind would be in 3D printing, if it’s a solid metal bed, deflection could be a problem at temps. Being a solid, thick, thermal mass would cause expansion, whereas the frame it has now (being as open as it is) allows a lot of heat to dissipate (and still has a little deflection as is). Maybe for the 3D printing bed, add some form of standing spacers, similar to how Prusa does it. This would limit the heat transfer to the metal plate, and reduce flex to just the heated PCB, which would be more repeatable.
This image is part of setting the PINDA height, but it’s the best view of the spacer I could find. With how many screws it takes for the snapmaker, threaded spacers might be the way to go so they don’t move around. (screw spacer into plate, screw PCB onto spacers).
I just purchased some rails from amazon to help take some of the pressure off the y-rails. I’m not sure this is the design I would buy for a better bed, but I like the general idea.
Also, my updated bed is slightly different from the one that is displayed in the rendering. Just FYI there are 3 different snapmaker bed designs that they’ve made since the original kickstarter. The newest one has a square frame.
@edf From front to back, you are correct, but from left to right there seems to be nothing on the stock machine that transverses across it for lateral stiffness and flatness. I did see a couple users who installed some angle-aluminum across the tops of the Z-axis modules to supposably do this to a certain degree. Perhaps something could be designed to do this across the baseplate; maybe even under the baseplate as a backing for the SBR16s.
@cunninghamand He mentioned that in the “Disclaimer” in the original post.
While I think this solid platform could be a good idea for use in slow speed movements as in CNC use I think it could cause problems with the extra weight at higher speeds seen in 3D printing. For my bracing mods I have tried to keep weight to a minimum to reduce wear. Even so when I recently opened up the rails for service I noticed a good deal more wear material on the leadscrews of the Y-axis(bed) rails which are the ones I have had supported the longest so my guess is the extra friction of the guide rails and weight of the glide block carriages had something to do with this as the X-axis(toolhead) which was unbraced looked almost new even though they have about the same amount of mileage.
I would be more inclined to just have a DIY kit that anyone could buy with all the needed supplies and instructions to add guide rails and simple shimming of the 4 corners and maybe an integrated clip system to add a glass bed if desired.
This is a thing indeed, One which I’ve seen with my mod of adding an MDF sheet in between. adds a significant amount of weight. On the other hand. Printing very fast with the A250 isn’t that great anyway. It’s mostly tuning the acceleration parameters accordingly. I just print slower in general (or use a 0.6 or 0.8mm nozzle for faster prints & prototyping)
The temperature stability would be the main reason for me to swap the bed. If it isn’t stable on higher temperatures, the point of using something like this is very limited.
Again, what I experience with my mdf mod: works, but after a few months, there is warping and it becomes more of problem over time. Which is a bit annoying.
If you want to make something that isn’t dependent on the base plate, enclosure, or whatever, you could make something that is more like this: Snapmaker A350 attachment additional linear guides by gojoetofly - Thingiverse Where the rails are attached to the linear modules.
Personally, I let them rest on the base plate and use 3D printed braces to keep them aligned. Seems to work well too in my case (rails lie on the bed, not these brackets, the brackets just align).
Other questions I have: If this is made of metal, aren’t you creating a giant heatsink for the heated bed? i.e., won’t it take much longer to get it up to temperature? (and require more heating to keep it there)
And finally, for this to be attractive for me, a quick change mechanism would need to be included as well. That still is my biggest annoyance with the SM2.0 model. A possible solution to this would be to keep the original bed as is, and have this extra layer that “slots” into the original bed and can be secured very easily.
Having a real flat surface would definitely be great, yet this thread might create a lot of extra wishes and different requirements as everyone has their own specific desires and needs
A massive/full metal base will act as a heat sink for the heated bed - it will take much longer to reach target level temperatures. Unless you introduce spacers, but then you’re back to the heated bed having room to deform. On the other hand: Once heated up, bed temperatures will stay much more stable, the bed having some buffering capacity.
SBR16: Adding linear guides is on my project list, I see the benefit in avoiding lateral forces that wear out the linear modules and make the bed flex. However, my plan is to investigate for linear guides that I could attach to the screw holes in the linear modules itself. That should make the alignment issue mentioned above less difficult, and you would not need to drill holes into the base plate.
If you read what other makers that built their own CNCs or hunt for maximum accuracy in 3D printing take care of, one thing is getting the X-axis really parallel to the bed. So I’d second that there should be some fine adjustment mechanism for the bed.
One thing this suggestion wants to address is the often bad bed flatness. I myself have a very bumpy bed - but I guess before I take further steps against that, I’ll wait for the announced quick swap mechanism. Snapmaker is aware of its bad bed flatness, and I’ve a bit of hope that the’ll address this with the quick swap. Perhaps it makes sense to postpone your idea until the new platform is out?
Just thoughts - in general, I’d appreciate some easy available upgrade to address all issues you mentioned! Would save me a lot of time
This would be a fine addition to the proposed kit, and would make installation much easier. Kinda wish I’d come across the idea before I chucked the base on the Bridgeport table for drilling - would have saved a lot of setup time and I could have just used a plywood table on a drill press.
This could be coupled with @Hauke’s idea of bolting directly to the linear modules: provide four spacers per side that would bolt to the linear module, and butt up against the base of the SBR-16, as an additional allignment aid. If 3D printed, there could even be disposable drill guides, like provided with the MJ Splitter for table saw inserts, which bolt to the linear modules.
A few probably unnecessary words on the actual drilling. The rails have to be in their final position when the holes are drilled, but the standard-fit M5 clearance provides quite a lot of leeway, and the mounting holes in the rails even more so (I should mention here I didn’t use SBR16s which I think use smaller fasteners). So you loosely fasten the bolts holding the rails, then power up the A350 and run the Y axis to the extent of its travel a few times, back and forth, tightening the bolts each time (outer, then inner). The tricky bit is that we are drilling through-holes, not threaded holes, so there has to be a nut under the base for the bolts to thread into, and that nut has to be held when tightening. Putting the A350 up on 2x4s will allow a socket wrench to get in there; I used a Workmate with one of the table boards removed to allow acccess from below.
One of the biggest frustrations I had with the mount-points of the linear modules is that the dual rows of holes are only on the bottom: the sides provide a single set of holes on the far side from the carriage. But the sides are where you need to mount rails and such, so you end up having to use an L-bracket which either goes around to the bottom (not possible in this case) or over an end cap.
In your particular case, this means two pieces of angle iron on the end caps, and the rail mounts to the end caps, meaning it is only supported on two mount points. Not a big deal. The greater concern is getting the carriage of the rails to extend to the edge of the bed so they provide support there. I can see how it could be done (aluminium bar with a pocket cut out for the angle iron, with the bar and angle iron both using the mount holes along the linear module, and the angle iron using the end-cap mount holes, and the aluminum bar providing the necessary distance to place the carriages under the bed corners, with further angle iron for connect the carriages to the bed), but phew, that is even more work than mine
With all the feedback I gathered so far, seems like finding happy median will be the hardest. But not impossible.
And by happy median I mean - weight, flatness, minimal mnf steps (read cost of production), friendliness to not sophisticated users out there.
In general, I think this is a great start to get some interest for aftermarket manufacturers. The platform deserves the attention.
SPOILER - Summary In regards to Platform
Adjustable mount points to the SBR-16 Rails - YES - this is something that we have to have as many of you mentioned. The adjustment has to allow for parallel to X axis adjustment - how do you tram X axis prior to shimming the platform to parallel?
Shims - the hardest and less user friendly for users not spoiled with precision machining. But I think it’s cheaper include shims, then
Adjustment screw - much more user friendly approach but has to be developed in a way that will keep the price down.
Can we attach rails to the Y-axis and bound axis together? As per @brvdboss , @Hauke , @edf.
This should help with
Complexity of install - no drilling required
Establishing flatness of the assembly
I’m not sure if there is a rail that can be mounted on the side and accept load from the top. Point me in a direction of one.
Both Face and Sides of the Y rails would have to be used here as mounting points. Any ideas on how to do this?
Heat flex - cast aluminum should help? After all bunch of people jumping on the option for 3d printing with Vorons. They do edd a silicone pad to heat it aswell. This is something you can already purchase for snapmaker via Aliexpress, just 3d printing bed with silicone pad I mean.
Weight - this is something that has to be kept at bay for sure. There was a discussion with @edf on this. And many of you shared some valuable thought on this here. Would be great to get some data on the wear and tear vs added weight from people who did the rail mods.
Platform turning into heat sink - this is potential downside, I agree, @Hauke has a good point on stable temps here I think. Not sure how to fight this without replacing the current bed with silicone pad. This could be designed around the silicone pad being under the platform, adding magnets (like they do for other printers) would allow to mount the printing plate directly on top. And CNC/Laser beds would just go as usual.
Drop in design@brvdboss has an Idea of drop in design, utilizing current bed, this makes sence at first glance, but at increasing the production price and complexity I think.
SPOILER - Final Platform product has to:
Be available with or without rails.
Rails should be precut - but to what length? I’ve seen reports of 800mm being perfect.
The instructions should be made as simple and as user friendly as possible.
Other things people mentioned to me that would be a nice accessory to produce:
Dont get me wrong, im not trying to put cold water on this.
But at what point in your use case is it better to just get the dedicated machine built for the specific job, then to endlessly upgrade. Only like @Mxbrnr said in his long write up published today. The changing out of the heat bed is the greatest factor in repeatable accuracy. if I digested the long write up correctly that is.
Personally after reading the issues and doing my own adjustments (still in progress) i think to me this is a 2 in 1 machine. especially since we can strap a laser board to the bed as is. I think for all purposes for my use case. I’m going to stick with the 3d printing and laser. I really don’t think the snapmaker 2.0 is actually robust enough for good CNC. For the money to upgrade it to be-able in my opinion to do good CNC is same cost as a dedicated machine.
But like i said that is my own opinion and if you want to work on the machine to do CNC have at it. Its part of the makers life style and its half the fun of owning something like this. I would be interested in seeing what people can do with it.
This is exactly why I have already decided that at some point I will transition to a dedicated 3D printer, specifically a Voron Trident. The case for not having to constantly remove and reattached the heated bed and the constantly shifting lack of accuracy that creates, as well as doing away with the inertia and vibrations of the bed “slinging” back and forth, are a huge upside, among several other reasons I find very beneficial over the SM. At that point I will keep my SM only for laser work, unless I can find something equivalent to replace that function as well.
Funny enough, I just bought an XTool D1 Pro yesterday to get rid of the gantry limitations of a bed slinger. I have a Prusa MK3S and Mini for 99.9% of my printing, the snapmaker wears its 3dp gear only on items I need the larger space. (tbf, I never got the snapmaker for 3dp, the prusa machines were first, the option for a bigger bed is nice tho)
However, it’s by no means a replacement. With how much work I’ve done honing in on the repeatability and Lightburn control, the snapmaker is still going to be my primary laser. (and the location/ease of access).