A little slow on the response, but I do.

I uploaded my opto housing.
https://www.thingiverse.com/thing:4776015

TL:DR - The MMU2 is an interesting add-on that can do some cool stuff and is fun to play with. When it works, it's great. However as with almost all 3d printing, it has plenty of annoying issues that cause frustration and print failures.

To start, it is finicky to get setup and dialed in. Even after that, it's not as reliable as I would like. When it works, it does exactly as you would expect, but you have to accept the trade-offs of higher print failure rate, longer print times and extra wasted material. Also one thing I wasn't expecting was weaker prints. The boundary lines where the colors meet don't bond well and easily snap apart. This also contributes to print failures from the waste block falling apart mid print.

The main reason I bought the addition was to play around with soluble supports. So far I have only tried using HIPS with PETG. In theory HIPS should melt away after soaking in D-Limonene, but in practice (at least for me) it didn't work great. It eventually (hours) turned into a soft goo and I still had remnants to clean off, plus it made the print permanently orange smelling. I bought a spool of PVA, but haven't tried it yet. The Prusaslicer feature of only using the expensive soluble material just at the support interface layer is really appealing.

One thing the MMU is handy for is auto switching filaments during single color prints. It is nice to have 5 different spools basically ready to go especially if you do somewhat remote printing like I do. I find it perfectly reliable in auto load and unloading. One thing to note however is that running the filament through the MMU does limit the max extrude speed. For whatever reason, the control firmware doesn't put the MMU drive system in "neutral" so there is a always an added drag on the main extruder pulling the filament through the MMU feed system. There is also a problem when the MK3 has a reset, the MMU gets out of sync. If there is a filament loaded through the MMU, you have to extract that all out and do a full power cycle to get things back to steady state. You would think this doesn't happen very often, but if you use octoprint, it happens way more than you would like.

As for the spool feeding, the Prusa solution is to use multiple trays on the table top which takes up a ton of space. My solution is hanging spools above the printer and mounting the feed tubes on pegboard on the wall behind where the printer sits. This might not be an option for your setup, but thankfully load and unload feeding have never been a problem for me.

If I had to do it again, I would take a much closer look at Pallete 2s before deciding on the MMU.

Here are some multicolor game keychains. Homer was a 5 color print. You can see where some black artifacts show up in all the other colors. I forgot to mention it's really difficult to prevent color contamination. I think in this case, I might have a small goober on the hotend that I didn't get cleaned off before the print.

Let me know if there are any specific questions.

20190526_221841 (resized).jpg

Above all that was mentioned with drive tension and clogged drive gear, try bumping your first layer hot end temp up 5deg and see if that doesn't help relieve a little of the backpressure. Or perhaps if your slicer supports it, drop your first layer extruded width by 5% or so to see if pushing less plastic helps.

This is a better choice for what you are looking for https://www.smooth-on.com/products/xtc-3d/

My experience as well. I damaged a new smooth PEI build plate with a PETG print. Now I swap to the texture plate for all of my PETG prints. It is another nice feature of the Prusa to have multiple sheets configured for quick change over. Also have to be careful with TPU. That stuff sticks even better; to the point in the past I have added a filament specific positive Z first layer adjustment.

you have to print really slow, turn off retractions, hard to do with bowden extruders, sticks to the bed so shouldn't have any first layer issues (sticks too well sometimes...)

what kind of failures are you seeing and with which printer?

I would suggest to start with 1/2 or 1/3 typical speed and once you get it to consistent, dial it up slowly until you find the reasonable max. I am a big fan of TPU. The layer bonds are super strong and the prints come out near indestructible. Also note, that different formulations yield different durometer (harness) values, so you can find some variability in how flexible the end results are.

Like this?
https://pinside.com/pinball/forum/topic/ball-holders#post-3731175

Sure sounds like it's plugged up. What type of filament are your running and have been swapping types? Is the fan on the heat break running as expected? There are specific assembly instructions to make sure the nozzle seats correctly. Did you have any trouble there?

Not quite what you were looking for, but I made these for 4x6 sized replacements.

PXL_20210209_045831050 (resized).jpg

PXL_20210209_012833010 (resized).jpg

PXL_20210209_013139076 (resized).jpg

Yeah, the offset was needed to give enough clearance for the extra driver. That will vary a bit speaker to speaker.

It's pretty boring from the front.
PXL_20211105_154543912 (resized).jpg

I pulled it apart since you were going to ask anyway . You can refrain from comment on my hobo captive nut workmanship thank you very much! Haha.
16361277095016137280468982565323 (resized).jpg

What it would look like without the foam.
16361279535176399757111627752152 (resized).jpg

It was form factor reasons. I was thinking the shape of the speaker grill allowed for more speaker drive area with oval speakers. I picked the cheapest ok speakers I could find and hedged a little smaller (4x6 vs. 5x10) to see how they actually fit. If I were to do it again, I probably would go round. I think they are sonically better. This was more a thought experiment. These are the ones I picked if anybody is curious; https://www.amazon.com/gp/product/B06XBFM6CW/ref=ppx_yo_dt_b_search_asin_title

Do you know which is the off the shelf equivalent for the Stern upgraded speakers?

Yeah thanks for the thoughts. These are all great techniques. I am familiar with them and how to apply them with PrusaSlicer. What you see was a first quick and dirty print just to see if I got the critical dimensions right. The captive nut holes came out a wee bit undersized and full of overhang garbage so the nuts didn't really fit well. I sort of cheesed the fit by hard pressing and overtorquing to draw the nut in as much as possible. That's why you see the stress and distortion. In the end, it all fit fine and aesthetically didn't look any worse with the foam in there, so I didn't bother redoing it. I didn't make to sell (or even to show?). This is what lazy functional 3d parts look like. I did ask you to ignore the hobo job, haha.

After seeing how they look without the foam, I might try to print some prettied up versions. Like you suggested I could print a cap or face plate to the hide the hardware. My preference would probably be to print a solid bottom, then space for the hex nut, pause print, drop in a nut and finish the print around it, making it completely captured. I also wonder what it would look like if I gave the base a pattern instead of a solid face.

Oh and I should also add, I have had pretty good results by increasing perimeters around screw holes as an alternative to adding a ton of fill. I think it could be a good tradeoff of reinforced strength versus excess print material and time.

Yeah I agree, always more that one way to print a cat. Specific tricks used are very design specific. Lots of good stuff shared here!

I thought the upgraded LE speakers were 5 1/4" or something like that?

This. Also make sure your print fan is off.

I would say, the last thing to try if this last experiment fails, mix up some abs slurry use that on the glass bed.

Second the Prusa recommendation. I have both the mk3s and a mini. It's as close to set and forget as a consumer level printer can be. When I started in the 3D print hobby 7 years ago, I bought a Mendel Max kit, and struggled for years to keep it dialed in. While the Prusas are not without problem and still require a small amount of fidgeting it is miles ahead of where I started. I would never want to go back to that.

For ABS, I can't remember your main motivation, but I think it was for higher temp and UV stability? If those are your requirements, I would suggest giving polycarbonate a try. Similar on those properties and much less prone to warping.

PC seems much more varied depending on the manf. This is the PC that I have printed successfully with at 270 (which the manf suggests 240-270).
https://www.amazon.com/gp/product/B078J3C9GY/ref=ppx_yo_dt_b_search_asin_title

Prusa Mini. It is worth picking up both the smooth and texture sheets.
https://www.prusa3d.com/category/original-prusa-mini/

Agreed. I print a lot with eSun PLA+ with good results. Although I will say, a recent white batch of this has turned really brittle coming off the spool and usually breaks into many pieces when changing filament. I am not sure if it is color blend specific, or a batch specific problem. This might be my 10th roll across 5 different colors and this has been my only issue so far. I will say it is worth doing a temperature calibration tower for each color as they do behave differently. I use their yellow for my High Speed/Getaway LED stop lights. The color is rich and it prints smooth.
DSC08734_1_sm (resized).jpgDSC08737_sm (resized).jpg

I used Octoprint for years. It does work well. Prusaslicer moved where you input the IP and Octoprint API key (which you get from the Octoprint config page). Go to the Printer Settings tab and click on the gear wheel that is just right of the printer name drop down. It used to be grouped with the main printer setting options, but would get overwritten with profile updates or loading other 3mf files. It is now sticky which is nice, but a pain to find.

There is a bug reported with the MK3S filament runout sensor not working specifically when using Octoprint. Keep that in mind if you rely on that printer feature.

A couple things I have learned; I recently switched to Repetier Server as Octoprint can't manage multiple printers. I also moved from a raspi to a NUC. I run my print server 24/7 and eventually (weeks/months) the raspi SD card will become corrupt and which require a full rebuild. I got tired of this.

Feel to reach out with a PM if you have questions on usage or set up.

I would suggest adding maybe three "pinchers" than fit over the spokes of the insert. It would not only add more gluing surface area, also mechanical support. It would complicate printing a little, but nothing that couldn't be managed.

Like I mentioned, should be easy to manage. If this is in F360 and you don't mind sharing the design files, I can take a crack at it and show how I would print.

Rats. No solidworks here

The added sections (thinking extruded rectangles on either side of the splines/spokes) would define the new bottom plane. The bridging needed would be really short and shouldn't need support if your printer and filament settings are reasonably dialed in. Or, you could flip it and add supports to the underside of the ring. Both should be easy methods to end up with a nice functional part.

Off to PM....

Yeah good suggestion. I watched a lot of his videos when I was getting started.

Neat! For the split, create an offset plane (or any other construction plane that makes sense) and put it at what ever angle you need. Then use the split body function with the new offset plan as the splitting tool.

Keep the extruded windows as separate bodies until after the split. In the splitting tool, just select the base as the body to split and it will ignore everything else. Also, there is a combine function where you can bring it all back together, but there isn't a real need to do that. When you choose to create your STL file, you can select the bodies you want to use. Or, you can hide the bodies you don't want in the STL and select the entire component. Another trick is that if you export a model with a lot of separate bodies and use Prusaslicer, you can have it separate out the individual bits with either the split to parts or split to objects (I forget which).

This is a Prusa specific part. I am not sure if there are generic equivalents.

https://www.prusa3d.com/product/double-sided-textured-pei-powder-coated-spring-steel-sheet/

I switched to these to handle the more abrasive filaments;
https://www.amazon.com/gp/product/B01BDL4N4S/ref=ppx_yo_dt_b_search_asin_title

I have a friend who has been fully steeped in FDM printing for 7+ years. A year or so ago he added a resin printer and did his best to fine tune the process. After a few months I asked him how it was going and this was his direct quote; "fuck resin printing, not enough time in my life for that horrible bullshit" I took that message to heart and haven't considered experimenting with one since.

I found this article really interesting relating to community developed 3d model trademark takedown actions;
https://www.thedrive.com/news/honda-orders-big-takedown-of-honda-related-3d-printing-models-from-maker-communities

I have both printers and completely agree. The mini is a great first printer. Not the cheapest, but probably the best bank for buck and easiest to learn on. I also have an XL on pre-order...completely ridiculous and unnecessary, but man it looks cool!

I used the exact same model on my DP and the damn SNIKT target went back to the leaning-target-of-pizza... Although strangely enough, I seem to have gotten a decent feel on how to hit it, so I REALLY don't want to touch it now.

My first thought is to remove the tabs from the design so it's essentially two cylinders. Print with TPU so it will never break. Make the tolerance really tight with the PCB holes. TPU is playable and a bit grippy so it will make a nice tight fit They shouldn't fall out even without the tabs. The material is tricky to print right especially with the small dimensions needed here.

TPU has outstanding adhesion. With a PEI sheet first layer sticks too well. Also....auto bed leveling. It's good for what ails you.

I stumbled on a new technique (at least for me) for reducing support garbage removal and thought somebody else might find it interesting. I designed a power supply mount for Spike2 back box. I wanted to enclose the screw terminals of the PSU to give a little additional safety to accidental shorting of the live wall AC and the 12v output. The end result included basically a 5 sided hollow box that would require support for the "roof". This meant the supports would need to start on the "floor" of the box and would require a lot of removal of the bottom interface layers which tend to fuse to the model more than top interface layers. A separate design choice I made was to add a pattern of hexagon holes anyplace that I thought I could save material without compromising structural need. Screenshot 2022-05-06 100854_sm (resized).png

While playing around with PrusaSlicer's paint on support tool, I had an ah-ha moment. I targeted my support enforcers to the holes in the "floor". This meant that the vast majority of support tower's bottom interface were restricted to the build plate. The picture is looking up from the build plate. The blue is where support enforcement is painted on. By manipulating the support painting tool, the brush size can be made to easily paint inside the hex holes. One thing of note I found is that if I made the paint dots too close to the size of the hex holes, the support engine would spill support up and over and would still include a bottom interface layer on the floor of the print.
Screenshot 2022-05-06 094450 (resized).pngScreenshot 2022-05-06 094657 (resized).png

The only potential downside is your printer creating a clean first bridged infill layer that builds off of the support towers In my case, this wasn't a problem, but it will come down to material used, general printer bridging performance and the spacing of the holes. If the bridging goes fine, there is a secondary support removal benefit with less interface points to the roof. This picture is looking up at the underside of the box and shows the end of the support towers (green) and the first bridged infill interface layer (dark blue). Screenshot 2022-05-06 094921_sm (resized).png

End result of the mount. The Spike games and the SAM vault games use a metal backbox, so I needed to design the mount to work with any available factory mounting screws.
DSC09128 (2) (resized).jpg

My first thought was PETG before looking at the part. But..oof, there is a really thin section that I think is doomed to break on most materials. I am going to guess that's where it snapped? I say TPU is your best bet. It's little finicky to print, but easier than the other options I can think of; polycarbonate and nylon. TPU will absolutely without a doubt never break If you prefer to have it less flexible, you can do a little digging in the different brands and find harder durometer versions (maybe like a 90shore?).

I'll break up the Ender lovefest by recommending the Prusa Mini. I say this with zero hands on experience with any of the Enders . I have been very impressed with how the little mini prints right out of the box with the only real compromise is the 180mm^3 build volume. This is a recent side by side comparison with real data. https://www.3dsourced.com/3d-printers/prusa-mini-vs-creality-ender-3-pro-v2/ Between the two main options, I don't think you can really go wrong.

Fudge. I have been trying to ignore this, but after the video I got overly excited and put in my pre-order without much more thought Is there any word on what the lead times for builds are?

Haha, whoops! What's a 'c' or 'm' between friends?

I also agree completley with the rest of your assesment. I also have both Prusas as well. One of the two common mistakes I have seen people entering the 3D printing hobby is to overestimate the benefit of the larger build volume and underestimating the learning curve needed for a printer expecting upgrades and tweaking to get the most out of them (I made both of these mistakes, but the options were much different 7 years ago). That's why for me the mini is the right choice for a first printer if the budget allows. But there are certainly other good options as described.

Bottle opener?

It took quite a bit for me to get mine working semi reliably, but it was still a crap shoot when or how it would fail. It really stood out as my MK3 prints so rock solid. Multi color is neat to play around with, but prints take forever and there is wasted filament. I was most intrigued to try it with soluble supports. Only attempted HIPS and PETG and was not impressed with how that worked out. I bought a spool of PVA for water dissolvable, but never could convince myself to try it out because it's an expensive filament and I didn't want to waste it

The one area where I think the MMU is actually quite helpful is just having multiple spools ready to print in single color prints. Either for same colors so when a spool runs out, it can switch automatically and keep printing. Or if you typically print parts that require different colors or materials. It's nice for remote printing and not having to go and load/unload every time you print something different.

One drawback which I haven't seen mentioned in forums or bug reports, is that for some odd reason the MMU doesn't go into a neutral position when printing. Meaning, the drive gears are always engaged on the filament even if the motor is turned off. The end effect of this is extra drag on the main extruder which limits print speed. Sometimes I like to kick the print speed up manually and was finding that I was under extruding after a small increase in speed. All the MMU has to do is rotate slightly to remove the resistance and it would be fine. Honestly I think so many people abandoned using the thing, problems are just not reported anymore.

The upcoming Mega's tool changer will be a huge step forward.

I print remotely in my non-climate controlled shop. Humidity is decently high. I print plenty of PLA+ with the spools left open to the environment for months on end. I had generally ignored the need to keep things sealed as I never noticed print failures that I could attribute to wet filament. I changed my mind a little on this after I spent a considerable amount of time trying to tune out some minor stringing that was annoying me. I modified an old food dehydrator that I had collecting dust, and sure enough it completely cured it. So now, if there is anything I care about how it looks, I'll pop any exposed spools in the dehydrator for a couple hours ahead of time just to save on post process de-stringing. As was eluded to, it does seem specific formulations (brands and colors) tend to vary in susceptibility.

Hey I know that guy

In my experience yes they can. It isn't common and usually the difference is not much. However I did create a separate filament profile for eSun PLA Pro Yellow (versus all the other colors from the same manf). For some reason for me that color needs higher temperature and some changes to retract. The difference is consistent across multiple spools, so not related to a specific batch. I now print temp towers the first time I try a new color or manf just to be sure.

I could see it possible that you are right on the lower end border of your first layer print temps and switching to a different color that requires another 5C bump higher could cause specific color first layer bed adhesion troubles. All that said, it's way more likely as Pete suggested to be the bed surface needing a quick IPA wipe down.

I received my Bambu X1 a few days ago. I have to admit I am very impressed with the quality and speed I am getting out of the box with zero tuning. I have been really happy with my Prusas, but I think the X1 will quickly become my favorite machine. It looks like for the same multi part prints, I am seeing a reduction of at least half the print time at the same or better quality. I was not expecting that much improvement. What is really interesting is how the various filaments work with the generic presets with zero adjustments. For example, I bought some cheap PLA+on discount to try and was having difficulty on the MK3 with my tuned filament settings. Even after going through it's own temp tower calibration and specific adjustments I wasn't happy to use on prints I cared about and relegated the spool to just test prints. To my shock, the same spool printed perfectly in the X1 with the generic profile. Same goes for some silk PLA that I had stringing problems previously, printed 99% string free first try. Obviously these are just my results and there may be more contributing factors specific to my print environment and specific setup. So far, I am more than happy.

All that said, there are some annoyances that I have to get used to mostly with printing workflow. I print remotely to my detached shop and have been using Repetier Server for two printers on my LAN. The Bambu print through the cloud direct from the slicer isn't my favorite and I hope they can better detach the cloud requirement to the printer functionality. I would much prefer to slice to local file share and print directly from there as I have been doing, but I am not sure if that is even on the company's radar. I see there is an update today that allows you to slice and save gcode directly to the SD storage which is a step in the right direction.

I am printing a few of these gamecube controller stands for my son for Christmas and this was my first attempt at a multi color print with the AMS. It turned out awesome. There is a ton of waste which bugs me, but of course that and much longer print times are the tradeoff for single extruder muli print.
PXL_20221215_170342428 (resized).jpg

Good eye! Yeah it's a multicolor silk filament. How it is orientated through the extruder determines how the color comes out, either green, magenta or some gradient of the two. The way it this was loaded, on the color side of the model it came out mostly green.

There are ways to reduce the waste to some degree, but the nature of how you have to purge for a single extruder, it's just a fact of life. It's not unique to the X1.

Prusa through twitter shared gcode to try out a few presliced models with the new organic supports.

https://twitter.com/Prusa3D/status/1600521548053495808

Bambu Studio isn't quite on par with Prusa Slicer. In comparison you will find small bits of missing functionality, general UI chunkiness and plenty of minor bugs. If your workflow is built around remote printing through octoprint, or repetier you will be disappointed in lack of flexibility with the Bambu. As for quality of the slicing, the only thing I have noticed is much more pronounced seams with Bambu Studio. I haven't figured out if these are machine parameters that can be tuned out, or if it is some kind of sliced anomaly.

For a Prusa power user, the hardware will impress, the software will slightly annoy. I still would recommend adding one to your stable. You can plow through prints so much faster at the same print quality. The AMS unit is much better than the MMU2 and doubles as a four spool dry box.

Pretty sure it works the same way in BS, but maybe it doesn't think you have an AMS and hides the option? I am away from my main PC, so I can't confirm, but a Google search hints at example usage (along with current bugs).

@Mr_Tantrum: Here you go. Image taken from a recent bug report
https://github.com/bambulab/BambuStudio/issues/843

210933787-b17a5579-a686-4bd9-827e-3f0e7e81a9f7 (resized).png

Lidar based flow calibration, bed leveling and first layer inspection which can all be turned off add to the prep time.

I think the hot plate is the smooth PEI plate right? I print PLA on that. I started with the cold plate and used glue stick infrequently. I thought it was going to be much more obnoxious going back too glue after being spoiled for years on the mk3, but it really wasn't that bad. I have also printed TPU direct on the engineering plate and that worked great.

I think the easiest way if you weren't comfortable with doing it in cad would be to design the fan mount area as a plain solid rectangle. Then in the slicer add a modifier for that area and set it to zero top and zero bottom solid layers. Then choose the infill pattern you like the looks of the best. Voila, free grid. Add perimeters to make it strong enough. Use zip ties instead and don't worry about the mount holes.

Janky, I know. But there are a lot of quick and dirty things you can do in the slicer if you don't want to monkey with the model or more likely when you don't have the source.

Another good application of this technique is for speaker adapter plates.

For me anyway, PLA on the smooth PEI sheet (Hot Plate) and TPU on the Engineering sheet require zero prep (glue or otherwise). The cold plate when I was using it was very minimal use and mostly as a release agent. I haven't tried the textured plate yet for comparison.

Bambu's default purge values seem to be very conservative. Dialing these in will save a decent amount of waste. I haven't gone through this, but it is on my todo list;
https://www.printables.com/model/390414

There is a purge into infill but I am not sure if there is a purge into a waste object as well. PrusaSlicer has this for the MMU2. Both of these also cut down the waste.

I am not sure this is right. I have seen errors during slicing when the wrong plate was set for the selected filament (i.e. PLA on the engineering plate). But it had no problem allowing me to send a print meant for the hot plate when the cold plate was installed. Maybe it's supposed to catch it and I just ran into a bug.

Oh, and on the topic of Bambu additional build plates; I looked at multiple plate holders similar to the one you posted. My solution is to store them inside the enclosure on the "floor". I haven't double checked if there would be an interference issue with the bed all the way down, but I would rarely, if ever print to the full Z height. Having them in the enclosure saves a bit of bench space and the potential to knock them on the floor.

Yeah I am the latest as well. That is good to hear and interesting that mine allowed the wrong thing to happen. I'll have to see if I can recreate the scenario. Thanks for the info.

-Rob

Same as you. Remember our PM conversation when I convinced you to buy yours? haha

If you haven't already, I would suggest looking into the potential health concerns around the ultra fine plastic particles and fumes that can be emitted. I personally would not choose to have a printer in my kitchen because of these concerns. I only mention since you say you are new to the hobby and maybe not obvious to look into.

You definitely do not was sliders made out of TPU. It is way too grippy.

If someone was hell bent on trying different material off their printer, why not just make "caps" that fit over the metal sliders to get a feel for friction differences and test wear properties? Or probably better a cap over the part on the playfield?

I think you will find all materials are inferior to Delrin for this application and that time spent playing around goes way beyond these $25 parts. I am still not convinced there is an actually a problem to solve with the metal solution that already is on the game.

I can't help it

How about this....It would be an easily justifiable project for nothing else than something to practice cad design on. But I think coming up a design that is a cover over the metal rail would be more interesting and less risky. I mean, do you really want to be testing strength limits of your PETG part with the weight of your unsecured playfield standing straight up 7 feet in the air?

This example is only applicable for the few multi-extruder machines. For the Bambu w/AMS, Support-W filament is meant behave in the same way (low layer adhesion at the support interface layer).

I ran into a similar looking slicing error recently on BS. Mine was on a known good model that I created and have printed before and seemed very specific to support settings and adding a different support filament to the interface layers. At first I thought it was just a display artifact but confirmed when I tried to print, it actually was screwed up gcode. Since updating to the latest 1.06 beta build, it has gone away. Also maybe try the OrcaSlicer branch and see if it behaves differently.

I wonder what would happen if you cut your model somewhere above where the weirdness starts and only kept the upper half. If it's a model issue, I would expect the artifacts to remain in the exact same place as the full model.

PM sent.

I am prepping my games for NWPAS and made keychains to help the techs find mine easier.

PXL_20230531_002940270 (resized).jpg

If these are the same across all sys11 games, then I do. I made them for my high speed. I'll post a link when they are loaded up on printables.

Yeah I think we are talking the same parts. I found a good color match PLA+ to the Williams red. They were pretty much indistinguishable from the new replacements. One of the sets were new replacements, the other set has the 3d printed targets. For reference, the game was played heavily at NWPAS show last year (560 plays for the weekend) before I sold it and they had no problem holding up to the abuse.

https://www.printables.com/model/508421-williams-pinball-target

DSC08860 (resized).JPG

DSC08861 (resized).JPG

Haha. It's an embarrassing picture to be sure. I sold the game so it was the only picture I could find. Don't worry, all targets were precision aligned after completing the hardtop install and shipping off to the new owner

My suggestion is to not jump into 3d printing with the idea of multi-color right off the bat. Spend some time building experience on single color/filament type. There is a lot to learn and simplifying will make your early days in the hobby much more enjoyable. If you must have something in different colors, don't discount the ease of painting.

It's much easier to go from single to multi color printing if you start with Bambu vs. Prusa. The new MMU3 update hasn't been released yet, but I can't see any world where it can match Bambu's AMS. Also keep in mind that the P1P price just dropped $100. That is really hard to ignore.

It's fair and I understand what you are saying; switching per layer is pretty straightforward. My main point is, jumping into 3d printing for the first time there are plenty of new things to learn and refine for your first 50 or so prints. If the mentality is to start with multi-color print, that means to me, someone is focusing on a final product on the first print. I suggest that getting through the initial learning curve and quality refinement without swapping spools each print will be quicker and easier. I generally try to eliminate variables to establish a solid baseline and then go from there.

If you are getting the P1S, ASA or polycarbonate will be "easy" for you to print. Oh, and keep in mind that the P1S does not come with a hardened nozzle so abrasives (like any of the CF blends) will rapidly wear out the standard nozzle.

I have printed these for later WPC games, but honestly don't remember which material I used, probably ABS.
https://www.thingiverse.com/thing:2619071
preview_featured (resized).jpg

Nice! I didn't know you could do this. Thanks for sharing the tip. It would be handy to have paint on ironing like there is for support and seams.

Thanks for taking the time to post that up. It's how I imagined from your description. But I wonder if it works in all scenarios.

I have found weird unexpected dependencies in BL. For example, you cannot use variable layer height with a prime tower. One side effect of this is the slicer basically ignoring your support interface Z difference when using multi materials unless it is a multiple of the print layer height. For example, I was experimenting using a spool of HIPS as a cheap substitute for Bambu's Support-W material for PLA based prints. I found at 0 interface height (what the slicer recommends for support-W) the HIPS support interface would bond a little too well to the PLA. If I changed the Z interface to say 0.1, the slicer would automatically round that up to 0.25 or whatever my layer height was. The only way around it is to disable the prime tower which is not good for multi material printing.

So I am thinking modifiers that manipulate layer heights (like ironing) may run into the same strange limitations when multi material or more specifically prime towers are enabled.

I haven't seen this specific behavior in Prusa Slicer, but I haven't used the MMU in years. A little more background on my bug report to Bambu here; https://forum.bambulab.com/t/missing-support-layer-height-option/12324

Prime tower isn't for flushing (although it does server that purpose in a small way) it is for priming the nozzle pressure before continuing with the print. It is meant to mitigate oozing and under extrusion at each new start. It's possible that without the prime tower, the slicer attempts to prime into the part infill, but success of this approach would be model dependant.

Good data point.

Which materials specifically? Yeah, it is good to hear your experience and will give it a try, but given how little material a prime tower uses I am not sure why it is worth the risk for a multi-hour colored print? I found this interesting video of prime tower experiments on the X1.

My meta point wasn't really to debate prime towers, but that I have seen Bambu Slicer behave strangely when asking it to do multiple complicated things outside of default behavior and it's good to watch out for.

How do you turn off the motors on the Bambu's? The couple of times I looked for the function, I couldn't find it. I am assuming you had to do this in order to do the cold pull?

Decimals are just decimals, true. But there are more advantages to sticking with metric for modeling/printing. A few I can think off the top of my head;

- When working on fitting prototypes, for me at least it's a lot easier to estimate one mm here or there versus 0.04 of inches. Even if you just work in mils, 10 mils seems too fine to estimate and 100 mils is too course. 1mm seems just about right.
- Some slicers or other tools that import STLs, assume they were generated in mm. Most have the ability to easily convert or scale, but it's an extra step that adds potential for getting something wrong or loss of accuracy each time you need to transverse tools.
- If you deal with hardware, it's so much easier to understand how to model around metric hardware. For example M4 and M3 versus 1/4" and #4. Thread pitch is also way easier to work with as well. 1.25mm versus 28TPI
- If you need to design with other commercial products that have mechanical drawings you are much more likely to run into drawings only in mm.

Ha. Would you believe me if I told you that I did that on purpose to prove my point on trying to avoid stupid conversion errors?

Bambu has now three separate teasers for Sept 20. For the first color printing teaser, here are a couple fun speculations;

TPU doesn't snap. It gets wound up in extruder gears. It's a regular pain to deal with this when it's an easy to access hotend. I can imagine it being a giant pain if this happens inside the AMS. The more flexible the material you are using, the more risk there is to this happening. I have had my share of these issues on the prusa that required tweaking extruder tension to get right. Surprising, the Bambu has printed the same and even softer (ninja flex) TPU without issue, but I have never been brave enough to put it through the AMS. You can slow down print speeds and turn off retractions so that the extruder runs really slow and steady. I don't know of a way to adjust the AMS feed motor speed and I imagine that's where the trouble would start.

Interesting. They both seem to be 95A. Is the Priiline their high speed variant?

How did it fail inside the AMS and how did you fix it?

I would encourage you to not give up on it. The best practice is from trying to make something you actually want. I am not sure how you are approaching the modeling, but I see it as pretty much three steps;
- sketch half the side profile off of the z axis line
- rotate the sketch around the z axis
- use shell to hollow it out and to give the desired wall thickness

That will at least get you a button without that notch fairly quickly for test fit.

Thanks for the added input. Agreed, your description is the better approach, although I would consider drawing the notched ring in a separate sketch later in the timeline and adding a second revolve with a cut operation to the first Sure, it's more steps, but I find this more flexible would make it a ton easier to go back and make adjustments or use as a template for other like designs. The great and sometimes frustrating part of CAD modeling is there are always multiple ways to come to the same final design. The steps I suggested was to meant to simplify things down to basic elements at the expense of accurate reproduction. A couple insights for consideration based on my experience learning F360 over the years.

- One of the trickiest parts for me has been understanding how to correctly constrain sketches for what F360 expects. After the getting the basics down, next was learning how to do this in ways that leads to easy and flexible parametrization with user set variables. In this regard, I was hesitant to suggest sketching the profile shape with thickness and notch ring details as it would be a more complex drawing. Sketching using Offsets seems the natural way to get the shell thickness, but I find this tool tricky when not a simple chained drawing. Sometimes it does weird things (like reversing directions for positive and negative values) and requires each sketch line to be handled separately which makes it a royal pain to go back and change if adjustments are needed. Also, I don't remember for sure but I think Revolve requires full constrained sketch in order to define the element to execute on. Back to my comment, correctly constraining a multi-line sketch might be frustrating to a beginner.
- Shell isn't needed as you mentioned. I find it a forgiving and straightforward tool that at times can be used as a time saver. It's good thing for a beginner to get comfortable with
- I ignored adding in the notch to help speed up prototyping. Sometimes I like to rush to the basics of a design so I can validate general shapes and dimensions. I am also not sure if the ring notch is a functional requirement. Something else I have learned when trying to reproduce a part is that I spent way too much time to get every detail of the original in the model, only to find complex and time consuming details were unimportant.

There is no right or wrong. There are only tips and tricks to pick up along the way to make life a little easier. I always appreciate learning other's approach to solving the same problem.

I recently stumbled on HueForge software which opens a whole new world of colored prints. It basically uses the light transmission properties of overlaying colors to create shades and gradients (referred to as filament painting). With the tool you can import any image, pick the filament colors that you have available and adjust how you want the final image to look. It spits out an STL and a list of the layers that require a color switch. This means any printer can be used with the addition of pause commands and manual swapping. The Bambu AMS makes this easier of course as it can be set to switch automatically.

The biggest hurdle to get the results you see on the screen, is that you need to have calibration numbers for the specific filament you are using. The tool has a pretty big library of pre-defined values across multiple brands, but it is not that difficult to create new profiles from what you have on hand. The software starts at $18 for non commercial. It can also do lithophanes, but I haven't played with that at all. There are test models to download and play around with just to get an idea how these look in person;

https://www.printables.com/@StephenLaveda_686551/models
https://shop.thehueforge.com/collections/hueforge

I finished my first print and am pretty excited by the potential. In this one I used black/red/yellow/white. It is 200mmx200mm with a total thickness of only 1.36mm.

PXL_20231024_181708222 (resized).jpg

PXL_20231024_181715927 (resized).jpg

I followed pretty much the same path. Before Prusa, I started with an open source reprap kit and fumbled through that learning curve spending more time modifying firmware and updating the hardware than actually producing quality prints. I moved to the MK3 and it was a welcomed step up in consistency and reliability. Over time I added the MMU2 (which was not good) and the Mini. Conversations in this thread convinced me to give the X1C a try. I had a deposit on the 5 tool Prusa XL when it was first announced, but after the long delays and time spend on the X1C I canceled that. Since then, I have added a P1S and multiple AMSs. I had put an order in for the MK3.9 and MMU3 upgrade kits hoping to find good use for my old Prusa hardware, but again after more delays, I canceled those as well. I just can't logically justify the price for performance in comparison.

Make no mistake, there are things not to like about the Bambu printers and ecosystem. First and foremost I am not a fan of the cloud first approach for various reasons. One specific example is that I lost use of my P1S for two weeks after being unplugged for several days. I was unable to re-bind the printer to my account and bambu tech support was pretty much useless, even after providing them network traffic captures. The short version is that I accidentally resolved it by switching my ISP back from Verizon 5G home to Xfinity. The point is something stupid and completely unrelated can unnecessarily cause wasted hours of frustrating downtime. In a closed system, you are at the mercy of the manufacturer to figure it out. I also find Bambu/Orca Slicer not quite on par with PrusaSlicer specifically for seam print quality and certain workflow usability. I'm also not a fan of the excessive plastic waste. Even understanding the AMS implementation, there are ways this could be reduced, but it just doesn't seem like a consideration. Most of these fall under annoyances and lack of refinement, but it is healthy to say they are not perfect and can be improved.

But in the end I completely agree. Without hesitation I would recommend Bambu over Prusa.

Yes I understand. My printers are in an outshop, so it's an inconvenience to walk back and forth with a micro card. I also think I could have forced it to LAN mode, but was working on troubleshooting the issue. My X1C that was sitting right next to it continued to function 100%. The bug in their network code is confined to the binding handshake, but once associated with an account normal printing is fine.

Yep. They have (begrudgingly) improved LAN only mode for sure. An interesting footnote with my interaction with Bambu support is that they provided test FW to experiment with. It was simple to load the file on the SD card and the printer easily recognized it and updated. This means a robust offline update mechanism is already in place, but they still want to restrict it to cloud only for whatever reason.

The driver or the box?

I have seen plenty of homebrew games with 3d printed ramps, so sure it has been done. How are your CAD skills and how much do you value your time? I would guess at least 10hrs of prototyping and you will still be fighting designing and building it in multiple pieces that fit your build volume restrictions. My opinion is $130 sounds like a pretty good deal for perfectly fitting, brand new metal ramp. But if excited about giving it a go, don't let me talk you out of it

This would be overkill. The shapes are simple to model. One minute with a digital caliper is all you need.

I am assuming the trouble is with the support areas? Layer adhesion is one of my favorite properties of TPU, but it makes any kind of support really difficult to deal with. A couple thoughts that may make it better;

1) Use a dissimilar material for the interface layer. I have never tried this with TPU, but maybe just swapping in PETG for one layer would do the trick? Of course this a small pain to manually swap out. I wouldn't try with the AMS, but I know jrcmlc is much for adventurous

2) Redesign slightly so that the bottom edge has no groove and is flush top to bottom. Then print fully vertical with no support required. It may look bad since it could not cover the face of the metal flap completely.

3) Design into two halves. Use the center section as the key with the bottom half presenting the shell (female) and the top half with the interior of the the shell (male). Use any adhesive to combine them.

One could also reverse the mounting sides easily.

I have printed a few of these drink holders from Thingiverse with the same two piece design. https://www.thingiverse.com/thing:2268960 Maybe this is the model that you based yours on?

Besides the swapping around sides, I have been guilty of breaking a couple by bumping into them. Indeed it is nice to only needing to just reprint the mount.

Also, as a small bonus, I reused the same mount and designed a headphone holder to match.headphone (resized).png

It's not clear to me what you are trying to get to and how it is different from the example. Also, is kind of sounds like you trying to manipulate the STL in F360. If so, I would strongly suggest to instead design it from scratch. You will be happier with the result with less frustration and probably learn a bunch of new F360 modeling techniques along the way. I think the key piece to getting what you want is spending time with the repeated pattern along path function.

My opinion is that you will have a learning curve and frustration either way. But in modifying the STL, you will be frustrated with swimming upstream of trying to manipulate something to look like something else in a format that isn't intended to be manipulatable. By designing from the ground up, you will be frustrated by the learning curve of being new to modeling techniques and how the tool works. The difference is, for the latter you will learn skills that will benefit and grow with every model after. The former, I see the learnings as limited and somewhat throwaway in that the next STL you want to manipulate will have it's own unique challenges to get what you want. And in addition to getting generally lower quality results, you won't learn best practices.

I am sure there are other opinions and likely to point you to starting with Blender for STL manipulation. That's not the route I went, so better to let somebody else speak to the best alternatives.

Nice! Pretty cool someone worked the example for you. It is exactly how I pictured doing it in my head; make the one banister shape, make the rail, pattern the one banister along path of the rail. Once you get comfortable with that example, play with the F360 timeline and modify the banister shape/height. Then go to the end of the timeline. If designed with best practices (parametric references and fully constrained sketches) the final design will update automagically and you will get a sense of the power of the tool if used correctly. Also play with the pattern function and you can easily change the number of banisters and their spacing. You can also exclude individual instances from the generated pattern. Lots to play with and learn!

Be sure to share how it turns out. If you did not know, with F360 you can easily share a link to the model for others to view (under File -> Share -> Link).

Maybe try leaning the model as far back as your build volume allows and force all the support to the smooth side?

Finally uploaded this to printables. https://www.printables.com/model/803310-pinball-headphone-hanger

Make sure to also print the mounting piece from https://www.thingiverse.com/thing:2268960

PXL_20240217_224740264 (resized).jpg