How is MMU 2 an improvement over original multi-material?  

Re: How is MMU 2 an improvement over original multi-material?

Ask people who had MMUv1 first release. I don't think you will hear the word "flawless" a lot. Can't tell you if MMU2 will perform better then MMU1 but by design it should work with more variety of filaments and it has more tweaking options. So I'm confident PR and the community will improve it over time so it will get pretty reliable multi material unit.

Re: How is MMU 2 an improvement over original multi-material?

Ask people who had MMUv1 first release. I don't think you will hear the word "flawless" a lot.

I didn't refer to the original MM as flawless, you should read the last sentence of my post.

Can't tell you if MMU2 will perform better then MMU1 but by design it should work with more variety of filaments and it has more tweaking options. So I'm confident PR and the community will improve it over time so it will get pretty reliable multi material unit.

That is my point, why release something so unstable which turns the MK3, one of the most reliable printer on the market, to one of the most unreliable at over $1000.

MMU2 was advertised as an improvement, when in fact it's just different. Different != Improvement. They could have had a lot of very happy MMU1 users today if they just tried iterating over the original design while working on MMU2 in the background. Let those who want to be alpha testers (yes alpha, no way is MMU2 even beta) take on the effort and time, in full disclosure of what they are getting into.

Re: How is MMU 2 an improvement over original multi-material?

MMU2 was advertised as an improvement, when in fact it's just different.

Have you had direct experience with the MMU1? If not, I don't know if you can claim knowing that the MMU2 isn't an improvement over the MMU1 =P

They could have had a lot of very happy MMU1 users today if they just tried iterating over the original design

The original design had issues that couldn't be significantly improved by iterating on the same design. For example, the move to Bowden extrusion from direct drive extrusion on top of the increased resistance caused by the Y-splitter made it more difficult to print robustly with all filaments. Additionally, whenever filament got stuck somewhere in the system, there was no way for the MMU1 to detect the issue and pause the print because it simply didn't have the sensors for it.

Re: How is MMU 2 an improvement over original multi-material?

MMU2 was advertised as an improvement, when in fact it's just different.

Have you had direct experience with the MMU1? If not, I don't know if you can claim knowing that the MMU2 isn't an improvement over the MMU1 =P

No, I have direct experience with the MMU2, and I wanted to throw it out the window during assembly because its poor design made it frustratingly hard to put together. Looking back after a couple of weeks of using it I now realize that was the easy part of owning one.

I see this a lot in my profession as a developer, jumping on the new thing claiming it's better just because it's new. The reality is the idea may be better in theory, but the implementation of far from that. Before replying to my posts please READ them. The concept behind MMU2 may have more potential, but as is should not have been released because it's not ready.

They could have had a lot of very happy MMU1 users today if they just tried iterating over the original design

The original design had issues that couldn't be significantly improved by iterating on the same design. For example, the move to Bowden extrusion from direct drive extrusion on top of the increased resistance caused by the Y-splitter made it more difficult to print robustly with all filaments. Additionally, whenever filament got stuck somewhere in the system, there was no way for the MMU1 to detect the issue and pause the print because it simply didn't have the sensors for it.

Really?! How do you know it couldn't be improved significantly? Iterating design doesn't mean changing the part color from orange to red. And please explain how MMU2 detects filament not getting into the extruder today? Same problems repackaged in a different form factor. Different != Improvement.

Re: How is MMU 2 an improvement over original multi-material?

Really?! How do you know it couldn't be improved significantly?...

I guess because PR made this decision to change the design completely. It happens usually if you (as developer) don't see any chance to solve the core issues with the current design.
Your "Why" question could only be fully answered by PR developer. We as community can only guess.

Re: How is MMU 2 an improvement over original multi-material?

Really?! How do you know it couldn't be improved significantly?...

I guess because PR made this decision to change the design completely. It happens usually if you (as developer) don't see any chance to solve the core issues with the current design.
Your "Why" question could only be fully answered by PR developer. We as community can only guess.

And no one is immune to occasionally making a bad business decision.

Re: How is MMU 2 an improvement over original multi-material?

And no one is immune to occasionally making a bad business decisions.

How is it a bad business decision if the MMU2 unit is sold out? And I'm pretty sure it will be 2019 as-well. But again I think it was more a core developer decision.

Re: How is MMU 2 an improvement over original multi-material?

I've had mine for 2 weeks, ordered it for my MK2.5 but it wasn't compatible so ordered a MK3 that arrived and worked perfectly. Added the MMU2 and 2 weeks later and £70 worth of filament lighter still not one single successful print. Each print skips one or 2 layers (on different colours) so its impossible to track down. Support just keeps pointing me to youtube videos made by users not themselves.

I feel like an unpaid beta tester that had to pay for the privilege plus since adding the MMU2 it has been a downgrade losing the filament sensor and the power panic

Re: How is MMU 2 an improvement over original multi-material?

And no one is immune to occasionally making a bad business decisions.

How is it a bad business decision if the MMU2 unit is sold out? And I'm pretty sure it will be 2019 as-well. But again I think it was more a core developer decision.

Yup, that will be a lot of "happy" customers.

Re: How is MMU 2 an improvement over original multi-material?

I hate to jump on the bandwagon but the folks at Prusa really shot themselves in the foot with this MMU. I don't want to insult the good people at Prusa - they turned out an epic printer in the MK3 - but this MMU release is going to hurt their reputation. I would agree that this release is pre-beta and really unfit for sale to unsuspecting customers. Every discussion board I turn to has nothing but frustrated customers desperately searching for fixes for the MMU - so I feel comfortable that its not just me. I would wonder if there is anyone out there who has had good, reliable experience with this MMU? Is there some "trick" or "patch" that makes this thing usable? The concept is amazing and I really do commend the ingenuity of the people at Prusa - but they should never have released it in this condition. The worst part is that I very much doubt that a software patch can fix this and I doubt Prusa will be sending out free hardware to fix it - assuming they every find a fix... For now I'll convert back to single filament mode and conside the price of the MMU a learning experience. Please do better next time Prusa!

Re: How is MMU 2 an improvement over original multi-material?

To the gentleman who suggested that it is not a bad business decision if they are sold out... They sold out because of the hype and anticipation which was based on the Prusa reputation for excellence and reliability. I think the more people who buy this MMU the worse it is for Prusa. It takes years to develop a reputation and only seconds to destroy one. I would bet that the vast majority of the people who have purchased this MMU regret it and have significantly lowered their opinion of Prusa as a result. I honestly feel badly for Josef Prusa. He brought us something special in the MK2 and MK3.

Re: How is MMU 2 an improvement over original multi-material?

I've had mine for 2 weeks, ordered it for my MK2.5 but it wasn't compatible so ordered a MK3 that arrived and worked perfectly. Added the MMU2 and 2 weeks later and £70 worth of filament lighter still not one single successful print. Each print skips one or 2 layers (on different colours) so its impossible to track down. Support just keeps pointing me to youtube videos made by users not themselves.

I feel like an unpaid beta tester that had to pay for the privilege plus since adding the MMU2 it has been a downgrade losing the filament sensor and the power panic

I'm right there with u guys. Had some success, but mostly failures. I even thought I turned a corner as far as not having failures.....nope wishful thinking.

The failures are even worst because of the added time mmu prints take....filament changes add up to 2 or 3 times longer prints. Then theres the waste. I'm not happy about the purge block, but at least if the model finished I could convince myself it was worth it. So model with layer skip - a waste, double the print time for a failure and now all the waste from purge block as well.
Then it's users complaining about hobbed gear inconsistencies that prompts a firmware change? No ramming guide at launch. Mk2.5 users can't even use their mmu's because they don't have the right cable or firmware? What where they doing/testing for 6+ months if this is the result.
And don't get me started about it can't detect lack of filament in mk3 extruder issue...
You can calibrate Bowden lengths and tweak ramming settings to the cows come home, but until Mk3/mmu2 can catch/prevent layer skips this design is broken.
The out of box experience is horrible, not the polished Prusa experience at all. I haven't read 1 post where:
"I just built my mmu2 and look at this awesome mult material print" That certainly wasn't my experience. Instead it's "I can't get this thing to work, please help"

Re: How is MMU 2 an improvement over original multi-material?

I would bet that the vast majority of the people who have purchased this MMU regret it and have significantly lowered their opinion of Prusa as a result. I honestly feel badly for Josef Prusa. He brought us something special in the MK2 and MK3.

It's funny you say that about the MK3, mesh.f, because for the first 4-5 months of its release, the MK3 was a disaster too. None of the new features worked reliably, the firmware rebooted often during prints, you couldn't use Octoprint or USB to print until they disabled linear advance, crash detection caused random layer shifts when enabled, etc. It wasn't until around April or May that there was a solid firmware release that mostly worked. Even now almost 1 year later, some of the major features like the filament sensor still don't work reliably, and many of us are still waiting for the promised powder coated sheets. Not to mention the dreaded "602" inconsistent extrusion issue, which Prusa says is caused by the Bondtech gears (and I believe it) but there is no official solution yet. The original R1 printed parts had a number of errors and misalignments which contributed to bad prints. The R2 parts fixed some issues and the R3 fixed even more, but they took a lot of work to install. Many users received non-hardened steel rods which quickly wore out and had to be replaced. The first revision of filament sensors had a design flaw that could cause them to get hot enough to melt through the extruder. The first batch of PINDA probes had bad wiring and broke after a few dozen hours of printing.

My MK3 was one of the first off the assembly line (frame is signed "First MK3" by Prusa) and I suffered through every problem listed above, and there are probably some more that I forgot about.

I'm not trying to be too negative, I'm just saying that if you feel the MK3 is special (in a good way) now, then the MMUv2 will probably get to that level over the next year or so. Or maybe next year Prusa will announce the "MMUv3 which solves all the problems of the v2" and we'll be back to the starting line again. 🙂

Re: How is MMU 2 an improvement over original multi-material?

No, I have direct experience with the MMU2, and I wanted to throw it out the window during assembly because its poor design made it frustratingly hard to put together. Looking back after a couple of weeks of using it I now realize that was the easy part of owning one.

I see this a lot in my profession as a developer, jumping on the new thing claiming it's better just because it's new. The reality is the idea may be better in theory, but the implementation of far from that. Before replying to my posts please READ them. The concept behind MMU2 may have more potential, but as is should not have been released because it's not ready.

Okay, but I don't see anything to indicate the MMU2 didn't improve on the MMU1. It's not like the MMU1 didn't have assembly issues either, such as holes being too small, splitter tubes having bad alignment, splitter tubes missing machining, etc. I don't see how assembly issues proves that the MMU2 didn't improve on the MMU1, particularly in the aspect of functionality.

And what makes you think I didn't read your comment? I know you suggested working on the MMU2 in the background, I'm just telling you there was nothing further to improve on the MMU1. It was a dead end design.

They could have had a lot of very happy MMU1 users today if they just tried iterating over the original design

The original design had issues that couldn't be significantly improved by iterating on the same design. For example, the move to Bowden extrusion from direct drive extrusion on top of the increased resistance caused by the Y-splitter made it more difficult to print robustly with all filaments. Additionally, whenever filament got stuck somewhere in the system, there was no way for the MMU1 to detect the issue and pause the print because it simply didn't have the sensors for it.

Really?! How do you know it couldn't be improved significantly? Iterating design doesn't mean changing the part color from orange to red. And please explain how MMU2 detects filament not getting into the extruder today? Same problems repackaged in a different form factor. Different != Improvement.

Okay, and how *exactly* do you propose iterating on Bowden extrusion? How do you propose iterating on the friction caused by the Y-splitter? It's easy to say "just iterate on it", harder to actually provide any solutions to those issues.

You seriously think the Prusa team didn't try to think of ways to improve issues? I worked with them on a number of MMU1 issues and can confidently say that they tried.

The MMU2 currently detects feeding problems on the multiplexor side using the FINDA on the filament selector. It could theoretically detect feeding issues on the extruder side using the optical detector, but that has been disabled for now.

Re: How is MMU 2 an improvement over original multi-material?

I would bet that the vast majority of the people who have purchased this MMU regret it and have significantly lowered their opinion of Prusa as a result. I honestly feel badly for Josef Prusa. He brought us something special in the MK2 and MK3.

It's funny you say that about the MK3, mesh.f, because for the first 4-5 months of its release, the MK3 was a disaster too. None of the new features worked reliably, the firmware rebooted often during prints, you couldn't use Octoprint or USB to print until they disabled linear advance, crash detection caused random layer shifts when enabled, etc. It wasn't until around April or May that there was a solid firmware release that mostly worked. Even now almost 1 year later, some of the major features like the filament sensor still don't work reliably, and many of us are still waiting for the promised powder coated sheets. Not to mention the dreaded "602" inconsistent extrusion issue, which Prusa says is caused by the Bondtech gears (and I believe it) but there is no official solution yet. The original R1 printed parts had a number of errors and misalignments which contributed to bad prints. The R2 parts fixed some issues and the R3 fixed even more, but they took a lot of work to install. Many users received non-hardened steel rods which quickly wore out and had to be replaced. The first revision of filament sensors had a design flaw that could cause them to get hot enough to melt through the extruder. The first batch of PINDA probes had bad wiring and broke after a few dozen hours of printing.

My MK3 was one of the first off the assembly line (frame is signed "First MK3" by Prusa) and I suffered through every problem listed above, and there are probably some more that I forgot about.

I'm not trying to be too negative, I'm just saying that if you feel the MK3 is special (in a good way) now, then the MMUv2 will probably get to that level over the next year or so. Or maybe next year Prusa will announce the "MMUv3 which solves all the problems of the v2" and we'll be back to the starting line again. 🙂

Consumer 3D printing is still frontier territory if you were to compare it to the plug and play reliability of your average $300 laser printer. But that is understood, so the onus is on the end user to calibrate, test, tune, etc to get the printer working reliably. 3D printing is high maintenance, but a solid printer will give you reproducible results.

I've had my MK3 since December, even if there were bugs and hardware issues, I can say the machine was solid, expectations were met, I was happy. There will be unfortunate customers who had really bad experiences, but I feel the general consensus is the MK3 worked mostly like it was supposed to from the beginning with a reasonable amount of issues to be fixed.

When I compare my experience with the MMU2 to when I first used the MK3 (which BTW is also my first 3D printer), I am baffled how this was even released. These are not edge cases or obscure use cases that cause failed prints, the unit just does not print reliably because there is so much work that still needs to be done on basic functionality. The only way some are able to get a significant print to finish is by modding, which is beyond calibration and tuning. That is not a reasonable experience or expectation.

Re: How is MMU 2 an improvement over original multi-material?

I would bet that the vast majority of the people who have purchased this MMU regret it and have significantly lowered their opinion of Prusa as a result. I honestly feel badly for Josef Prusa. He brought us something special in the MK2 and MK3.

It's funny you say that about the MK3, mesh.f, because for the first 4-5 months of its release, the MK3 was a disaster too. None of the new features worked reliably, the firmware rebooted often during prints, you couldn't use Octoprint or USB to print until they disabled linear advance, crash detection caused random layer shifts when enabled, etc. It wasn't until around April or May that there was a solid firmware release that mostly worked. Even now almost 1 year later, some of the major features like the filament sensor still don't work reliably, and many of us are still waiting for the promised powder coated sheets. Not to mention the dreaded "602" inconsistent extrusion issue, which Prusa says is caused by the Bondtech gears (and I believe it) but there is no official solution yet. The original R1 printed parts had a number of errors and misalignments which contributed to bad prints. The R2 parts fixed some issues and the R3 fixed even more, but they took a lot of work to install. Many users received non-hardened steel rods which quickly wore out and had to be replaced. The first revision of filament sensors had a design flaw that could cause them to get hot enough to melt through the extruder. The first batch of PINDA probes had bad wiring and broke after a few dozen hours of printing.

My MK3 was one of the first off the assembly line (frame is signed "First MK3" by Prusa) and I suffered through every problem listed above, and there are probably some more that I forgot about.

I'm not trying to be too negative, I'm just saying that if you feel the MK3 is special (in a good way) now, then the MMUv2 will probably get to that level over the next year or so. Or maybe next year Prusa will announce the "MMUv3 which solves all the problems of the v2" and we'll be back to the starting line again. 🙂

You make some interesting points. Perhaps I was lucky enough to get into the MK3 at just the right time as mine worked pretty well from the start. Of course we are all a bit on the leading edge and some level of personal work with the kit is to be expected. As an engineer I just don't see how the MMU can work reliably. I think the concept is creative and has a bit of genius in it - but the implementation would probably require a vast change to the quality and sophistication of the hardware - and likely a much higher price. I'm not inclined to be so harsh on the Prusa guys since most new companies have issues with being overly aggressive and over ambitious. In a way, I think that's what people who buy these things like. However, I'm wondering if they just went a step past credibility with this MMU. I get the feeling they felt the pressure to get it out because there is no way that they did not experience the same issues all of us are experiencing. A very apt comment made by another post was that there are no Prusa videos or pictures of the MMU working successfully, fixes or work arounds - that's probably because they couldn't get it working either. Perhaps if they owned up and sent out some kind of communication to all the buyers it would help - rather than leading us all to keep trying to work through the impossible?