Mk2.5 without the MK3 Problems?  

Re: Mk2.5 without the MK3 Problems?

Or downgrade your mk3... 😛

Just a thought.

Re: Mk2.5 without the MK3 Problems?

Depends on what the issue is with the Mk3. If the problem is Einsy related (such as the TMC2130's) or the stepper motors (if the Mk3's are different, I heard they were chosen for TMC2130 compatibility) then the Mk2.5 probably won't have the same problems as the Mk3. If the problem is firmware related the Mk2.5 may have the same problems, but that should get fixed eventually anyway. Otherwise, they're basically the same printer (except the Mk3 has a more rigid frame and 24v).

Personally, I've had great results with the Mk2.5, but my standards could be lower. I print faster on this than I did with the Mk2 and my quality is better. That could just be because I've got Slic3r settings dialed in more though.

Re: Mk2.5 without the MK3 Problems?

Or downgrade your mk3... 😛

Just a thought.

It’s coming to that for me!! Might not be a bad idea...

Re: Mk2.5 without the MK3 Problems?

While I don't have a Mk2s to compare my Mk3 against. I do have several clones i have built. On my clones I am using Octoprint and I have TMC2130 drivers running in SPI mode with Marlin 1.1.8 allowing me to take advantage of sensor-less homing. I was already utilizing, and continue to utilize, the linear advance capabilities within Marlin. Based on my experience I believe the issues being reported with the Mk3 are primarily software/firmware related.

Re: Mk2.5 without the MK3 Problems?

Anyone upgrade to MK2.5 that also has an MK3???

Essentially the jist of my original question was to see if the MK2.5 folks are experiencing the same issues as I have with my MK3.

Basically it boils down to figuring out what makes the MK3 less stellar than the MK2... Is it the new Einsy/Firmware, or the 24 volt motors, or the bondtech extruders (and firmware)? I just don't want to take a great MK2s and turn it into an okay MK2.5. I am mostly interested in the removable heatbed. The rest I could live without.

Anyone care to speculate if you have no definitive answers?

Re: Mk2.5 without the MK3 Problems?

There has been reports of a change from mk2s to 2.5 having similar issues but there are also mk3s that print fine.... it could be extruder/bondtech gear related of the sort.... no clue.

see:
https://github.com/prusa3d/Prusa-Firmware/issues/602

one guy has 4 machines with an upgraded mk2s to 2.5 and an mk3

Re: Mk2.5 without the MK3 Problems?

Ok folks - it's time for my story! I used my MK2 for a year before deciding that I seriously liked the idea of the spring steel sheet and decided to order the upgrade to MK2.5 and also the MK2S upgrade. I carried out both upgrades in the same session and finished it about 6 weeks ago.

I wish I had never bothered! I have reached the end trying to find out what is going on with this printer and am seriously considering downgrading or trying to sell it as a partially-working printer for someone else to battle with and buy a DLP printer instead.

Firstly the bed adhesion is dreadful compared to the MK2. IPA was all I needed before and only occasionally was acetone used if a large printer area was required to avoid peeling of the first layers before the infill appeared now it's as if the print bed has been greased! I have messed soooo much with live z I'm totally sick of it. The V2 calibration turns a corner and the filament folds over rather than sticking. Sometimes it sticks and sometimes it doesn't. I can peel the layer off by using my artists palette tool which only needs to sit in my hand and not be gripped to push the filament off - it's that bad. I also have this problem in spades...

https://shop.prusa3d.com/forum/prusa-i3-kit-archive--f84/wave-pattern-in-first-layer-t2148.html

...with large amounts of waving in the first layer the second layer ends up with the nozzle catching the waves of the first layer. With small print area jobs it breaks the foundation off easily and also pulls off support layer foundations too. I have to now use rafts which I'd never even used before with the MK2 though of course they are vulnerable to braking off on layer 1 and/or 2. I have poor side wall quality too with variable layer widths causing protruding layers which catch your nail when dragging it down the printed model. Not a cool look for an expensive printer!

Luckily(?) no filament feed issues at all but the sensor is a solution looking for a problem. I was perfectly happy with the old extruder, selecting from menu to load filament and inserting it in. It's simple and it works so why fart around with it? Filament sensor not necessary for me. The S upgrade with new rods/bearings is great - very quiet as is the MK2.5 Noctua fan.

So there you go. I'm a very careful person with good attention to detail and when I originally built the MK2 it calibrated first time with no axes out. I put the same amount of effort into the upgrades and don't know where to go next. I've done z-adjust, belt tensioning, new ptfe, heatbreak, hotend and nozzle, also thoroughly cleaning the steel sheet, retarding the extrusion speed and I can sometimes hear skipping in the gears usually when the initial purge / pressure equalisation is carried out before printing. I just dunno - fed up with it - I am still designing models but wasting so much time messing with the printer. My old MK2 was like an appliance - just load the models in and print - no messing. I have printed some stuff including variable layer height models which worked great once they stopped breaking off on the first layers so it all comes down to that dreaded first layer and side wall quality doesn't always show up either. Oh also getting a lot more stringing that I got before.

Next step to downgrade the extruder which presumably means downgrading the firmware then the new printbed won't be supported and will need to come off - bummer.

Firmware v3.2.1 Slic3rPE v1.41.1-beta

Regards to all, Rob

Re: Mk2.5 without the MK3 Problems?

You can use the MK2.5 bed as long your printer is XYZ calibrated. I am currently running an MK2S with the MK2.5 bed without issues. So if you do decide to "downgrade" back to MK2, get the printer re-calibrated using the original bed then you can put the MK2.5 bed back on. After that, do a Z-calibration and adjust live Z and you should be good.

If for some reason you need to redo the XYZ calibration, you will need to put the old bed back on. Changing out the bed only is pretty easy.

Re: Mk2.5 without the MK3 Problems?

I feel your frustration! For ages my 2S was reliable. Despite a careful attention to detail in the upgrade to 2.5, I've not had a single successful print with any filament in several months of trying. After another 4 hours of experiments and tests today I'm ready to toss the while thing into the river.

I get pretty good adhesion on a first print, but almost invariably a wave develops, catches the nozzle, and there goes the print. Or, the filament won't feed enough. Or the printer arbitrarily demands a filament change. All these from both SD and USB prints. USB prints have the extra bonus of "line not last line +1" errors.

If the system had been hot for a while, the first layer fails either through not extruding at all, or just not sticking. When that happens, a full cool down is needed them it makes a good start until the problems above kick in.

Strongly leaning to rolling back to 2S.

Re: Mk2.5 without the MK3 Problems?

I'm keeping on keeping on with my 2.5 warts and all. Let me break down the upgrade parts and show how I stabilized my 2.5 to working order.

Spring steel sticker PEI and MK54 12V heatbed.
MK2S 25°C to 100°C take 5 minutes 44 seconds and weighs 465g.
MK2.5 25°C to 100°C take 6 minutes 48 seconds and weighs 805g.
2.5 takes a minute longer to 100°C and weighs nearly twice 2S. That is the cost of convenience.
The PEI on the "hold-over" spring steel is unusually slippery and needs treatment to work as well (or near as well).
Also, because you are hampered with the increased near as half kilo heavier bed, think of your printer as having a massive half kilo print already on your bed prior to starting a print. You can see you should adjust your jerk or decrease travel and print speeds if you see too much ringing on your Y-axis. Perhaps Slic3r PE already has these tuned, haven't compared.

On the literal double edged sword of Bondtech extruder.
Simply put, Bondtech extruder gears crushes given filaments into a ribbon like shape. This is bad because FDM relies on feed stock being a cylindrical shape in two ways. One, the guide PTFE tubes keep the filament flowing as well keep it from bunching up too much. If the filament is crushed on both sides essentially creating a ribbon rather than a cylinder, there is far greater chance of the filament collapsing into an s-shape creating delays in pressure application, but it gets better. With it now being collapsed, when a retraction is attempted, there is a significant delay due to the need to straighten then pull that ribbon before physically retracting anything, all the while the G-code/firmware is oblivious and or near impossible to tune out with wholesale retraction settings alone. Two, most hotend's have a round aperture (lower half of the heat break). A round shape matching the round aperture minimizes gap loss in pressure control. Now find an ovoid shape that has large gaps, this will cause even more pressure hysteresis small as it may be making pressure control more sloppy. Add to all that everyone migrating to PETG for it's superior strength, lack of fumes, better environmental impact and you have a filament whose own relatively large thermal/pressure hysteresis to contend with and you have a clusterfork that only serves to cause less inquiring minds to doubt faith as a worthwhile philosophic exercise.

On filament sensor and clumsy firmware handling.
Filament sensor, first of all, is a bit finicky, not what you want for a hobbyist grade fabrication tool. It can be affected by any light bleed whether your Filament Sensor Cover part was printed with too large a hole that the transparent PTFE tube causes like bleed. It can be affected by the feeding of the filament pushing the Extruder Idler to push the Extruder Idler Plug enough to let light in. A very transparent filament with a lot of overhead lighting, this one is a guess. The other two are not. And the most obvious sign of things not being robust is the need to print the Extruder Body and Filament Sensor Cover in a perfectly opaque black filament, which by the way most PETG blacks are not particularly in sharp corners which there are a lot of.
Unfortunately I have a lot of issues with the way auto-loading is handled but safe to say it is clumsy, inconsistent and unintuitive. Github is a better forum so that is where my recommendations will go.

None of these are unsolvable.
1st layer adhesion solutions.

Bondtech woes

Filament sensor hyperactivity

Re: Mk2.5 without the MK3 Problems? Success! -ish

Well this is slightly annoying/embarrassing but I have got some success. I felt the printer was over extruding so had previously used the tune menu to retard the filament flow down to as much as 50% with little difference in sorting the problems out.

I had a brainwave to check the settings->move axis->extruder and recheck the filament calibration. Well it was set at +32 or thereabouts! I put it to zero then did a test print. It took about 1 minute after starting the print before filament started coming out when I could see that the nozzle was way too high. I got it down to around -0.640 then filament started laying down properly. Another test and it is printing how it should.

I lost count of how many z adjusts I had done that had looked fine but with too much filament coming out it ploughs furrows into the filament and gives the impression that you have reached the lowest point on z and need to go higher. So I have printed an object with very large flat area covering most of the bed with absolutely no lifting whatsoever! It is still showing waving interference like patterns so that prob remains but I'm working again!

Re: Mk2.5 without the MK3 Problems?

.....
Simply put, Bondtech extruder gears crushes given filaments into a ribbon like shape. ....

it does? I'm not so sure and I was curious so I just looked at a stub I got when changing filament colors and the bit is round but with a bit of texture to it where the extruder grabbed. The gears each have a semicircle groove in them that should make a round shape between them when they are aligned.

It's pretty curious what's going on because my MK3 has been working very well with PLA and PETG (all I've tried). The bed bonds really well and the overall print quality is excellent. One curious quality issue is the 'basket weave' or 'inconsistent extrusion' and I can find this on prints when I look for it. It does not affect the utility of my prints.

I think that if you have a printer that is working for you its not a great idea to change it except in small ways, slowly over time.

There must be something in the MK3 system that is sensitive to build variations and I do think there is something in the firmware and driver system that needs to be addressed.

Re: Mk2.5 without the MK3 Problems?

@Themzlab
I posted my resolution to extrusion idler for MK2.5/MK3 in another post so might have neglected to post here.
Issue with MK2.5/MK3 R2 Extruder Body was the port hole for springs have a clearance issue if printing in PETG or PC because the holes don't have a top-bridge notch. See image, no notch.

Poor bridging materials will sag causing a D shape. In my case and likely others, the spring would catch not allowing full extension of spring causing far too much tension in order to clear the X-End Motor part. Unlike MK3, MK2.5 needs to be printed by customer so quality control is up to PR writing explicit instructions online etc. in assembly manual and diligence of the customer to read it.

Here is an image of Prusa Orange PETG after correcting spring port sag obstructions, resulting hobb scoring should not cause bow-in-heatbrake issues as discussed in prior post.

I also made a Cold Pull helper tool. It was quick and dirty but I'll post if you guys want. I'm sure someone else could do a better job but it gets it done for me.

P.S. I found out there was a problem after trying my tool for the first time. Happy accidents.

Re: Mk2.5 without the MK3 Problems?

Here is my 2 cents. Two changes can fix your problems. 1 Buy the steel flex bed plate for the MK 3 from Buildtek and put their PEI sheet on one side and Buildtek on the other. I always use the Buildtek side but the pei protects the bottom from scratching. 2 You HAVE to slow prints down maybe 50% of the stock settings. With the added bed weight the default speeds are just too fast for the MK2.5. Slow the speeds and it works great. So Buildtek and slower speeds. It really works. If printing something big and heavy, you must slow it down even more. Just because it can print fast doesn’t mean you should.

Re: Mk2.5 without the MK3 Problems?

Bought the MK3 for my workplace, and whilst I like some of the new features, I find the bed adhesion terrible. The prints just do not stick without applying prittstick first. I am quite careful about set up and keeping the bed clean, but have had only one print work successfully so far without using glue. Tried IPA, and also a wipe with pure acetone, but it seems to make little difference. I am of the opinion that there may be some issues with some of these spring steel plates, given the number of problems.

I have a mk2S at home, and print regularly with it, and never have the need for glue. Just a wipe with IPA and the prints stick just fine. I bought a MK2.5 upgrade before I read about all the problems with getting prints to stick to the new spring steel plate, so haven't bothered upgrading yet. I would like to know more from Prusa first about why there is a problem with prints sticking on the new plates, and what can be done to cure it.

Re: Mk2.5 without the MK3 Problems?

Just a follow up here to the last post ... The MK3 suddenly developed an issue with the calibration, failing on the front right point. After trying this several times without change, I decided to check and adjust the PINDA height (slightly too high), delete the current calibration set up and start from scratch. The wizard went through without any faults, and I finally set the first layer z height. The prints are even better than before and no issue with bed adhesion. There must have been a flaw with the initial set up, as the prints stick just fine now, and i only use a wipe of IPA between prints. No need for glue or acetone.

Re: Mk2.5 without the MK3 Problems?

2C from someone who upgraded 2S to 2.5:

Inconsistent extrusion: The official resolution is "all printers do this!" and for the most part, I was satisfied with that since I'm generally not the type to go looking for a problem. Calibration cubes on the 2.5 looked different, a little, maybe? Whatever.

Then I printed a big 22h long piece I've done a couple times before on the 2S. The walls looked like garbage in comparison. Like I'd printed the part on a cheap Chinese clone. And this was at .3mm layers.

Bed adhesion: I hadn't noticed any problems here until just today, trying to print a part where first layer detail is extreme. I'll admit, it's a hard part in the first place, but I simply cannot get the small bits to stick right. With my live Z dialed in as perfectly as it gets, it's just still not good enough to handle small details. The 2S could.

First layer repeatability: The 2S was always basically set it and forget it. Calibrate, live adjust Z, leave it. Thermal stability has never been good, but with some G code to keep the PINDA away from the bed during heating, the only rule of thumb was to not start a print with an obviously hot extruder. Now, on the 2.5, I have painstakingly crafted custom G code to avoid heating the PINDA during bed heating, then heat the pinda delibrately to precisely 35C before mesh leveling. Even with that, repeatability is STILL complete crap. My first layers are basically setting +/- 40% - hopefully it's good enough, hopefully your model has no detail down there.

Jamming: I find myself printing enclosed a lot for ABS and for odder-shaped PLA parts that peel and can't, for various reasons, be printed with a skirt. It's become literally impossible as of the 2.5. I'd never had a jam before on this printer while it was a 2S. Not one. Now, any enclosed print over 1h is a guaranteed heat creep jam. I've even had jams WITHOUT printing enclosed. This noctua fan is delightfully silent and also completely inappropriate for this application.

EDIT: I forgot to include this bit entirely -
Filament sensor: This was the single most motivating factor for me to upgrade -automatic runout detection to avoid partial spools. I've got some heavy parts >500g I occasionally need to make 3-4 of. The idea of having 5 open spools of white PLA on a growing not-quite-scrap heap was unappetizing. Unfortunately, it doesn't work, basically at all. It's supposed to support both runout detection and jam detection (enabled in recent firmwares). I've had jam detection fail multiple times, taking the (very long) print with it (bonus: jams that never happened on the 2S). I tested runout a couple times to verify functionality with the white filament I was using, and then when I actually let the print run, that failed too, wasting something like 6 hours (and commensurate filament).

TL;DR: the 2.5 is worse than the 2S in every measurable way. It's quieter, except I can no never print enclosed, so even that's not quite accurate now that the door has to stay open. It's got all the problems of the Mk3, and the one untainted advantage of the Mk3 (silent motors) isn't present on the 2.5.

I need to downgrade, it's just a question of when I'll have the time.