Ultimaker S5 | impressions & opinions

[kmanstudios 1,120]

38 minutes ago, gr5 said:

Infill would be a good place to pause.  Ideally in the "middle" of infill but most infill lines touch shell.  But better to be touching the inner shell than the outer shell during a pause.

Or anywhere on a tower.

Actually, I noticed last night when I had to pause a print to drop weights in that it does an intelligent 'pausing and resume' in the firmware. At least that is where I am hitting pause....on the printer.

 

40 minutes ago, gr5 said:

I like Bob's other idea better - a cura plugin that when you use up X meters of filament it pauses and waits for you to change to a new spool of filament (for jobs that need more than one spool).

The issue I have with that is how do you keep track when switching out spools> It is not all easy peasy to just use one type of material (Color, brand and specific plastics), unless that is how you are set up.

 

And, almost all threads I have read on this subject on the forum are all about physical sensors. However, this does not mean we should not find a way to put such a plug-in in place for those that could use it.

[uncle_bob 21]

Hi

 

At some point anything you do with "how much filament" is going to be a guess. There are cases where *knowing* when the changeover will be is a good thing. Not just from the standpoint of the print. It's also nice for the person swapping spools. Trading off a bit of material left on the spool against this convenience *is* very much part of the deal. You get something / you loose something.  If meters of filament instead of grams makes things easier to code ... go for it.  Translating from meters <-> grams offline is not that big a deal. Coming up with a post it note for whatever I normally use is pretty easy. 

 

Bob

[kmanstudios 1,120]

And when it grinds and you have to pull off a meter or meter and half? Do you know how much you have used before you had to do that? And farms that have multiple machines going balls to the walls with multiple people would greatly complicate that.

 

A physical detector would notice a halt of filament regardless of the situation. And, again, all other threads about this are people working toward a physical solution.

 

And as I said, if it is easy to create a plugin then by all means, but I think it would be extremely limited.

 

I know I would never use it if the physical solution works because it covers more ground with the same result. Also, as I said about pausing and resuming, it seems to do a pretty good job of that as it is. Sometimes I wait a while before it happens and sometimes not so long.

[Xalara 6]

As a small time maker who owns a UM3, I'm clearly not in the target market for this printer, but even for small business and small enterprise the price seems to be pushing it. My armchair analysis is that while this printer is going to be amazing, at a $6000 price point it creates a big opening for competitors to come in at a lower price.

 

P.S. where's my PVB profile in Cura :[

[Daid 306]

2 hours ago, JohnInOttawa said:

Best of luck for a successful production run.  Don't forget the UMO-UM3 crowd, I understand there are some engineering challenges, but if there are innovations that can trickle down, please trickle away (and for those who at or above are my 'silver' age, I mean trickle technology only.....)

The UM3 and the S5 run the same firmware, with a different machine configuration. So any firmware improvements on processes like changing material, pausing, etc, tickle down to the UM3. Cura Connect is also on both, and thus any improvements there also tickle down to the UM3.

 

UMO and UM2 can only expect bugfixes in firmware. But continued Cura support as always.

 

(Also, you commented on the UMO noise, most of the noise there comes from the material feeder, the back plate is acting like a noise amplifier, you can get some serious noise reduction by isolating the vibrations from the feeder to the backplate)

[kmanstudios 1,120]

Price tis not everything. I almost purchased the Raise N2+ in 2016. VERY happy I did not with all that came out later. And, that was a thousand bucks cheaper. And, there are already larger format printers available since that time, but so many other issues that I just never went past basic investigation.

 

And, a lot of those companies are cribbing off UM's work. For instance one company was touting their OEM of Cura had spiffy new features...that >ahem< they did not develop, but are certainly wanting to broadcast about it in a way that sounds like they actually did something other than create specific profiles.

[uncle_bob 21]

15 minutes ago, kmanstudios said:

And when it grinds and you have to pull off a meter or meter and half? Do you know how much you have used before you had to do that? And farms that have multiple machines going balls to the walls with multiple people would greatly complicate that.

 

A physical detector would notice a halt of filament regardless of the situation. And, again, all other threads about this are people working toward a physical solution.

 

And as I said, if it is easy to create a plugin then by all means, but I think it would be extremely limited.

 

I know I would never use it if the physical solution works because it covers more ground with the same result. Also, as I said about pausing and resuming, it seems to do a pretty good job of that as it is. Sometimes I wait a while before it happens and sometimes not so long.

 

Hi

 

Have you tried the other "physical solutions" on the other printers out there? I have. Despite the glowing PR and terrific demo videos those companies put out, they are *very* limited answers to the problem. Are those guys all idiots? Maybe they are. It's still Marlin and a tiny MCU deep down inside these machines.  They all share some limitations. I'm sure that UM will do a good job on anything they release to the public.  The big question really is - just how good is "really good" in this case? 

 

Only field experience will tell if there is a "physical solution". To date, nobody has an answer that really works in the filed. It is a multi dimensional problem and getting it right in even one of the dimensions has yet to happen. Again, that's not from looking at videos, it is from actually using the printers. People have tried this and tried that. They do this or that in the firmware. What works in one situation does not work the next time around. 

 

Unless it works more than 4 times out of 5, it's simply useless for change over. None of what's out there right now even gets to that level. UM could do a "best in the world" answer and still only be at the 80% level. Throwing away 3 or 4 days of printer time and starting over is way to expensive (from a number of standpoints) to do it on a regular basis. For changeover you will be relying on it for *every* large print. If you are really pushing a big printer, that's a lot of the time. 

 

On a "protect me from failure" basis, a "gets it right 80% of the time" answer is way beyond anything we have today. Things don't jam up on every print. It's a rare occurrence. Knocking that down 5:1 would be very welcome. Provided it doesn't false out ( as do many other filament sensor setups ) it would result in more production out of the printer. Nobody is going to say no to that. On many printers, the "false trigger" issue is bad enough that the common approach is to simply turn off the sensor. They create more problems than they solve. That *is* one challenge UM spotted early on and I'd bet has a better answer to. 

 

So the same engineering with the same net outcome is "great stuff" in one case may not really up to it in the other case. That's not a big surprise to somebody who spent 50 years doing design work. 

 

Bob

 

 

[Daid 306]

On the flow sensor, in theory it is simple, put a rotation sensor on the filament, check how much rotation you are getting compared to how much you are asking, differs by X, flow problem and pause.

 

In practice however, you need to get that sensor always touching the filament. Always. Without deforming the filament. This is actually a bit harder then it sounds. Some designs worked fine for a few spools of material, and then suddenly have a 2 cm extrusion of "no touching" due to some mechanical problem.

Next, how much filament we ask and how much we are getting is not a 1:1 relationship. Even tough our prints are coming out fine, there is a lot more happening then you initially expect. Filament itself compresses in the bowdentube. There is pressure in the hotend that slowly relaxes, and a retract+unretract does not move equal distance. The back-pressure of the build plate on the first few layers causes under-extrusion. Just to name a few things that can cause false triggers. And those are just the things I know, and I wasn't even on that team!

[uncle_bob 21]

Hi

 

Toss in something like flexible filament and I'd bet the issues just pile up ....One very basic advantage in this is the 2.85 mm filament. It will be more rigid and probably helps in this case. 

 

Bob

[kmanstudios 1,120]

@uncle_bob As I said multiple times, I get it. You want what you want. I have even said that if it is not hard to get the plugin written, then do so. Why are you arguing? And, maybe you should write it and not expect it done for free by complaining adnauseum. At this point, I will be happy to see what Team UM has come up with. Being first is not the best thing in most cases and most others are just getting it out there to say they have it.

 

But, I choose not to engage in this anymore. You want what you want and cannot just agree to disagree. So, I am letting it go like the hot air balloon it is. Your's will go your way and mine will go my way.

[foehnsturm 969]

I know this is a bold statement 

But our sensor had zero false alerts so far, by design it will always touch the filament.  Now I want to see if we can achieve an accuracy of +/- 2% for 1mm of filament. The 1 mm is because rotary encoder resolutions are limited, we are in the range of 60 steps/mm now. So far 85-95% real flow seem to be the regime for good prints. Flow below 85% compromises layer adhesion, below 75% you will see the classic under extrusion pattern.

[cjs 34]

6 hours ago, Daid said:

No new motion controller, still the ATMega2560. We had it as an option to upgrade this as well. But we made the choice to keep it close the UM3. (You could see the S5 as an UM3+)

The new controller board doesn't by "accident" use the footprint? 

 

6 hours ago, Daid said:

And the market that we are in right now isn't super fond on upgrade kits, even if you guys are :-)

Generally I would agree, but in this specific case I would disagree as it is such an important thing. I also don't think that professional would wan't to worry about a filament runout when they bought a machine for 3000+ Euros. 

I would have been super happy as I already have the plug for the sensor installed on my UM3 but I can understand it from a business perspective. And yes, I do like upgrades as I don't like to be "forced" to buy new things. 

 

Anyhow! Can you share a bit more about the new sensor and feeder @Daid ? Is it a mechanical solution or do you use an optical sensor ? Someone from Ultimaker also said that the gear is also different? @tomnagel did you maybe work on this team? How does the new mechanism disengage from the filament without voiding a patent by makerbot/stratasys? A picture of the internals would be great  

 

@Daid From the video it also looks like you finished reworking the cable stress relief of the printhead ? 

Edited April 23, 2018 by cjs

[uncle_bob 21]

4 minutes ago, foehnsturm said:

I know this is a bold statement 

But our sensor had zero false alerts so far, by design it will always touch the filament.  Now I want to see if we can achieve an accuracy of +/- 2% for 1mm of filament. The 1 mm is because rotary encoder resolutions are limited, we are in the range of 60 steps/mm now. So far 85-95% real flow seem to be the regime for good prints. Flow below 85% compromises layer adhesion, below 75% you will see the classic under extrusion pattern.

 

Hi

 

I think that "zero false alerts" is pretty much what the design has to do. As a protection device, a "measurable" false alert rate will drive people to turn it all off.  With any testing approach, there is always the question of how well it duplicates the real world. My dog just walked by shedding fur all over the place ..... how well does it deal with random dog fur  

 

(Yes, I know, get out the credit card and May 15th I can see for my self ....) 

 

The same step thing that limits you to 2% on 1 mm is not as big a deal on 10 mm. Yes, that's a bit of filament, but now your accuracy is 0.2%. This is the down side of 2.85 mm vs 1.75 mm. You have moved a bit more volume in 1 mm ( or worse 10 mm). I can't argue with the "90% is ok" logic. Tuning a lot of prints over the years, that seems to make sense. Unfortunately it *does* depend a bit on the type of filament you are using......

 

Bob

[Daid 306]

The sensor is an encoder wheel touching the filament, so mechanical. Optical simply doesn't work. If you want it to work for all filaments you need 2 different wavelengths of light, and it will get dirty. (If I remember right, black nylon almost completely absorbs IR light for example)

 

As far as I know, the feeder wheel is almost the same, except for it no longer wearing down on carbon filled materials. Those materials will still ruin the normal PrintCores.

Oh, but there is a quick-release lever on the feeder now. Instead of the tiny handle that you couldn't hold down for longer then 10 seconds without hurting your fingers.

[uncle_bob 21]

2 minutes ago, Daid said:

........

Oh, but there is a quick-release lever on the feeder now. Instead of the tiny handle that you couldn't hold down for longer then 10 seconds without hurting your fingers.

 

Hi

 

Quick release lever !!!!! Now *that* is a big improvement   My poor aching fingers thank you !!

 

So - the obvious question - is the new lever retrofit-able to the older machines? ( Obviously I hope it is. That would save me about $12,000) ....

 

Bob

[cjs 34]

5 minutes ago, Daid said:

As far as I know, the feeder wheel is almost the same, except for it no longer wearing down on carbon filled materials.

So we may be able to buy print cores for such materials soonish. YESSS

 

6 minutes ago, Daid said:

Oh, but there is a quick-release lever on the feeder now. Instead of the tiny handle that you couldn't hold down for longer then 10 seconds without hurting your fingers.

Yes, I already said on twitter that the days of the wedgebot are probably a thing of the past  Looking forward to see how you did solve this mechanically without getting into trouble with makerbot/stratasys. 

[cjs 34]

4 minutes ago, uncle_bob said:

Quick release lever !!!!! Now *that* is a big improvement   My poor aching fingers thank you !!

 

So - the obvious question - is the new lever retrofit-able to the older machines? ( Obviously I hope it is. That would save me about $12,000) ....

Sadly it's not  

If you don't want to spent 12.000$ you can still print yourself a wedgebot https://www.youmagine.com/designs/wedgebot-for-ultimaker2 or buy a bondtech ddg or do both  

[uncle_bob 21]

24 minutes ago, cjs said:

Sadly it's not  

If you don't want to spent 12.000$ you can still print yourself a wedgebot https://www.youmagine.com/designs/wedgebot-for-ultimaker2 or buy a bondtech ddg or do both  

 

Hi

 

I've tried a couple variations on the wedge gizmo. The world *could* use a better solution .....  I have not tried the Bondtech but I have been tempted. Part of the issue is that the printers are packed pretty tight and getting into the extruder in any fashion is a bit of a pain. Yes, that part is *not* a design issue on the UM side.

 

Bob

[ultimateboss 0]

Ummmm idk but 4500 is allot to spend on a printer(UM3 Ex owner)  and not be able to upgrade for filament detection. Ummmmmmmmm.... I don’t see the value as UM3 ex owner to upgrade to this 

[kmanstudios 1,120]

3 minutes ago, ultimateboss said:

Ummmm idk but 4500 is allot to spend on a printer(UM3 Ex owner)  and not be able to upgrade for filament detection. Ummmmmmmmm.... I don’t see the value as UM3 ex owner to upgrade to this 

For me, to be able to stop breaking everything apart would be a blessing. All that other stuff is gravy to me. The ability to use the full buildplate for both nozzles will also be a greatly welcome addition in the size department.

 

The way I see it is that it is basically $1,400 more for 50% more printer. And it is not for the hobbyist I am sure. While not printing as a money making venture, I am real serious about what I  can print.

[ultimateboss 0]

14 minutes ago, kmanstudios said:

For me, to be able to stop breaking everything apart would be a blessing. All that other stuff is gravy to me. The ability to use the full buildplate for both nozzles will also be a greatly welcome addition in the size department.

 

The way I see it is that it is basically $1,400 more for 50% more printer. And it is not for the hobbyist I am sure. While not printing as a money making venture, I am real serious about what I  can print.

 

Maybe if I had a UM2 but the build isn’t much bigger to justify the upgrade to me. Quieter doesn’t matter to me plus  I made my own enclosure also. It’s like my iPhone .... I kept the 6plus until Apple did something big .... which for me was the X. The value is there for you but not for me. Given I like new toys and I’m itching at the Wallet to buy it , it’s a tough sell for me personally. 

[kmanstudios 1,120]

3 minutes ago, ultimateboss said:

The value is there for you but not for me.

That is the beauty of this...it is not for everybody. And, I can go for 50% more buildplate. That will be a very big thing for me.

[Daid 306]

I always jam the Hex2mm screw driver into the UM3 feeder lever to push it at the office  (At home I have an UMO, that always had a quick release)

 

1 hour ago, cjs said:

So we may be able to buy print cores for such materials soonish. YESSS

No guarantees on that, but it would have been silly not to have the option in the future after we identified this issue. It's a bit like the new extension port, better be prepared.

[Torgeir 242]

Hi There,

 

You all are talking about the new Ultimaker S5 made in U.S.A. ?

 

My telephone is a S5, maybe I could have another S5..  Hmm..

 

Take care.

 

Thanks.

 

Torgeir.

[Xalara 6]

Speaking of print head cables. If the print head cable breaks will it be easier to thread the replacement cable through the case? I know on the UM3 replacing the print head cable is a royal pain in the butt.

 

Well, ideally the print head cable doesn't break in the first place but it seems to happen with heavy use it on the UM3.

Edited April 23, 2018 by Xalara
Missed a word