gr5

No need to pull the log if it clearly detects before touching the build plate. The logs will just be a waste of time. It's unlikely this but it takes 10 seconds to check. Open the front door on the print head - so you can see the cores. Part of what swings open includes the base of the print head and there is a circuit board plate there that is one side of the capacitor to detect the print bed. There are 2 wires going to that PCB. Tug very gently on both of them to make sure they aren't loose. They tend to break and come loose resulting in intermittent issues. Make sure there aren't any computers, cell phones, printers, things that radiate electromagnetic radiation within a half meter of the printer. I suppose you may have a rare conductive filament? Have you measured the resistance with an ohm meter? It just seems so unlikely to me. Even conductive filaments tend to be very high ohmage - too high to act as an antenna. On the order of 1 million ohms per inch. But if you are at say <10 ohms per inch then that could definitely be the issue. Oh wait - maybe it has magnetic properties? See if a magnet sticks to the filament. Maybe that also could change how the capacitive sensor reacts.

Oh that sucks. So there is the old way to upgrade and the new way. It switched over a few years ago. 4.3.96 is an invalid version. I'm not sure where you got it. It was quickly replaced by 4.3.97. I forget why - something to do with not being able to upgrade. 4.3.96 is one of the two "stepping stone" firmware versions (the other being .97) that are not meant to do any printing - just get you to the newer version of 5.X (e.g. 5.2.8). I think you will need to do a "firmware recovery". Ask your reseller for the procedure or you can read about it here:

Sorry. You did it perfectly. It's probably no longer an option. I guess I would put the S3 into "developer mode" which is an option on the front menu. Then you can ssh into the printer. Use username/password ultimaker/ultimaker (you can also use root/ultimaker to do other things). In that first account (ultimaker) it pops you into a nice utility and you can type: sendgcode M142 r0 g0 b0 w0 You can search this forum for "sendgcode" to see other commands you can type (e.g. LIST I think is an option?).

Most/many forms of carbon, including carbon fiber, are not conductive. Did you check it with an ohm-meter? It's true that nanotubes are conductive, but carbon fiber should not be. I recommend you watch the leveling procedure - just the first two tests which are different. There are 3 failure modes: 1) Surface detected too early. As it does the leveling it never even touches the glass. This indicates electrical problem such as nearby interference. 2) Surface detected too late. The test relies on the capacitance changing. The surface isn't detected until the glass moves down without the nozzle moving up. Visually you see it touching the glass and still moving down more on one core than the other and it's usually obvious. It could be the springs are too tight on the print bed or too loose on the core but this is very rare. 3) Cores are truly different heights. This more common than you might think. They can vary. If you look in the log file you can see what was measured. The cores are supposed to be 1.5mm different height nominally with an error of only 0.7mm allowed. It could be a bit of gunk on the nozzle tip that wasn't hot enough to squish flat during the procedure. This is extremely difficult to measure on the core itself with a micrometer as it's the height from the tip of the core to how it seats inside the print head so to the 45 degree angled area of the core just above the heat break. So if it's issue #3 I'd just try a different core on the right core. Or left core. This below shows you how you can quickly find what the printer measured so you can see if you are failing by just 0.1mm (typical for issue #3) or by many mm.

No idea about catalina but you don't need 5.1.1. It is only needed for the newest Ultimaker printers that have newer chips (you know - chip shortarge?) which I assume haven't started shipping yet or only just started shipping.

1) Try PLA - it's an amazing material and won't string so much. It's just as strong as PETG and does well indoors or outdoors but gets soft at a lower temp (around 52C or 130F) which is enough to kill humans but is sometimes experienced in a closed car on a hot day. So that's the one thing where PETG is better - it can withstand a little higher temperatures. 2) I also get much more stringing of my PETG but not that bad. I'm not sure why you get quite that much. Did you play with retraction? You want just a little bit of retraction. You can check if those moves that strung included retraction or not as there are two different shades of blue when you look at it in PREVIEW mode. 3) I have been told that PETG strings more if it gets "wet". You could try drying it on a heated bed with a towel over it with the bed set to 60C for 4 hours. I have never dried my PETG (nor PLA) but you must keep PVA and Nylon very dry. PETG I don't care about and it does string but not as much as you show here. Nylon needs to be dried even if it was at 50% humidity on the back of my printer for 1 day.

Detailed, and simple, instructions: https://ultimakernasupport.zendesk.com/hc/en-us/articles/115004226246-Printhead#“printheadcable"

lol. I do order fulfillment - I ship things around the world. It's cheaper to ship something from china to the house across the street from me than for me to ship something to the house across the street from me. This is definitely true for light weight packages under about 4 ounces (about $4 for me, <$1 for china) and still cheaper I think < 1 pound. Not sure what it costs (for china) when you get above 1 lb but I suspect it's cheaper at all weights. My big advantage is speed. Those <$1 shipping costs are for things that typically take 4-6 weeks to arrive in USA. I can get a package within my zone for next day delivery for same price and across the country is 3 days worst case.

You can buy a "wifi enabled SD card" which is both an SD card and has wifi at the same time. I put one in my DSLR camera. You can have any computer with wifi read and write files to the SD card.

Sorry - I was on vacation the first half of this week. The door fan should be on 100% if nozzles are above 60C and if that is off it explains your issue and is usually the cable where it connects inside the head. Again, if door fan is off this explains your issue and perfectly matches your symptoms. The side fans are more complicated. Typically they are off for the first layer (for example). In Cura put "fan" in the search box (just above the detailed settings) and there are 8 settings in the cooling section that directly affect fan speed and possible other's that are more complicated but certainly those 8 settings come into play and you can hover over each one to see what they do. HOWEVER, while it's printing you can just go into the TUNE menu and set the fan speed to 100% and you should hear them.

I don't have the pro bundle so I usually just lift the lever on the feeder and slide in the filament until it is in the print core. Since you have the pro bundle, I would go to the middle menu and then click at the top left and right corners - on the PVA filament button and then in the corner click the "..." and choose "MOVE" and move the filament until it is deep into the print core. Repeat for the top right button (PVA) and choose MOVE and move the filament into the print core. If you want to see if things are clogged you can wait for it to heat up and move some more to make sure it comes out the nozzles. Don't leave the PVA at printing temperature for more than the minimum to test that it extrudes at baking PVA for many minutes will caramelize it into a minor clog. Once you get this working and have had a successful print with no aborts you shouldn't have to worry about any of this. Potentially you shouldn't have to worry about these things for days/weeks/months/years. But sometimes as you saw and especially after unusual circumstances (like an aborted print) the filaments end up in the wrong place. Also make sure you have the latest firmware. Older versions of the firmware had a bug that is similar to what you describe but I think that was all fixed around a year ago. But again - once you see the filament well above the cores, maybe do MOVE. I mean the firmware is supposed to keep track of where left and right filaments left off but it seems sometimes it loses track.

Oh it looks like you have an S3 which I think means when you save it won't be ".gcode" but will be ".ufp". Just rename the file extention to ".zip" and you can see there is a simple directory tree and it's easy to find the gcode file. You can edit it and save it or just look at it and then rename the zip file back to ufp.

So try this: Add that line of code in the "start g-code" box. Then slice and save the file. Then open the file in a text editor and see if it added your gcode. If not then you will have to edit each gcode file that you create and add the line. Or do what I suggested and turn the frame light off manually.

Unfortunately, I don't think those are used if g-cdoe flavor is "griffin". You could try it but I think you have to edit the gcode with any text editor. Or you can just turn off the lights in the menu. There is an option to set the light's in the menu somewhere I think. "frame light" maybe?

If that includes the front fan then this would explain everything and your symptoms perfectly make sense. I'd check the cable going into the top of the print head. The rear two screws on top of the head come out easily and that allows you to remove the cover and you can reseat that cable. Careful - it doesn't just pull out so I'd leave it and just try pushing it down some more. If you only mean the side fans well this sounds like a problem but a different problem. ABS doesn't really need much side fan but a tiny bit will improve the quality of ABS prints. Just a tiny bit of fan.

118 hours is not a lot. I suspect there is no PTFE aging but that doesn't mean there wasn't a "bad" lot of PTFE that had a slightly thinner inside diameter. But I'm thinking this is not the issue. Your UM3 should have come with 2 AA 0.4 and 1 BB 0.4 cores. So you should already have another AA 0.4. Why not try that one? Print speed? Well if the filament is jamming due to being too hot (common with ABS) then faster is better. For all other reasons slower is better (e.g. feeder can't keep up or too much friction somewhere or quality of print is also better when printing slower). So actually I'd try both - you can do that in the TUNE menu on the printer while printing - speed up to 150% to see if you get a jam sooner. Or if jam issues go away.

Test the front fan in the door (NOT SIDE FANS) - stuff can get in there after years and ABS won't care but PLA or PETG will and will fail just as you say like 20 minutes into the print typically with the filament kind of stuck in the print core. Fan should start spinning around when printcore reaches if temp rising 40C and stay on until below 60C if temp falling. Seriously consider using a new core that you havn't used much. The higher the temp, the faster the PTFE inside the core ages. Again - doesn't matter as much for ABS if the PTFE is getting soft but a problem for PLA and PETG. So maybe go into the maintenance menu and look how many printing hours you have on each printcore. keep temps low. I use 220-230C for PETG (on both my S5 and my UM2go and my UM3). And use a little more fan to get good overhangs and good looking prints (ABS barely needs any fan at all - PETG needs about half the fan of PLA which doesn't mean necessarily at 50% as my UM3 fan at 50% is the same as at 100% - on my UM3 I do 9% fan for PETG but every printer is different). Some filaments - particularly ABS - will clog if at too high a temp for too long. Normal printing temps like 240C for ABS left at 10 minutes (without printing - just sitting in the nozzle) will turn the ABS into a gum-like mess that is hard to get out of the core.

Looks good. Interesting that you have to subtract 1 TWICE from layer_number. I'm guessing that the last layer or two are not actually layer information but are instead like the finishing up stuff and final gcodes and comments at the bottom.

Purge is indeed done quite high. That sounds normal. Sometimes it doesn't stick to the glass and is carried over to the start of the skirt/brim.

How many hours are on the core? I'd try the other AA core that came with your printer. Or get a new core. They are considered to be "consumables" by Ultimaker. Similar to filament. And Ultimaker was nice enough to keep the price pretty close to cost for cores. The PTFE can age from the heat but I still have all my original cores. Basically there are several components at and above the heat break that set the internal diameter. There is the steel heat break (inside the heatsink) and the PTFE tube (about 2cm long inside the heatsink) then there is to top of the aluminum heatsink for maybe 5mm and then the top tube. Steel. Ideally everything should be exactly 3mm in diameter but sometimes the PTFE might be say 3.1mm in diameter (or the heat break). When you do retractions and also just heat creep can cause the filament to expand into that 3.1mm space and if it hardens then you have a wider section of filament that has trouble getting past the heat break or trouble exiting the core (upwards) or has trouble when re-inserting. Or maybe it's nothing to do with any of this and the filament sensor in the feeder needs cleaning or replacing or maybe you need to angle the filament better. Or disable the filament sensor (I think mine is disabled? I forget but my S5 hasn't complained about filament issues in years). Seriously consider taking apart the feeder - if you only remove the 2 screws it's easy to get back together as the rest is still well attached to the printer but be very careful with that strong spring and how it goes back - most people put the spring/screw back in incorrectly and then their feeder is at the minimum tension after that so look at a video or something. And see if there is a mess to be cleaned particularly around the sensor. Or if you don't have the Material Station, look at how the filament sensor works and consider putting the filament on the floor. The angle that the filament enters the feeder matters to the filament sensor.

And it *only* does this at the end? So I know a guy who has several S5 pro bundles and he has been on the latest firmware since it came out and his printers work nonstop almost 24/7 and he said he just did lots of prints yesterday with only core 2 (right core) and there were no problems. Of course he didn't watch what happened in the last few seconds. Maybe he just didn't notice? It's really hard to tell what's going on in your video. Having the cores touch the top of your print like that is pretty common - overhangs can lift up on their leading edge. Think the "bow of a boat" (benchy) - in that case, when printing Benchy, the tip of the bow sticks up above the print level and the nozzle will hit that part of your print repeatedly. This happens on many many of my prints and is pretty normal. It looks awful. It sounds awful. But 99% of the time the print comes out fine. You can reduce it with more cooling which is why you always want 100% fan with PLA (but not with every material). So I really can't tell if what you show in the video is normal or not. Now in your photo you show the silicone rubber cover - your printer should have come with a spare. You can get these cheap - I forget the price but around 5 covers for 5 or 10 dollars/euros. So 1 or 2 euros/dollars each. So in the video is it only the silicone cover that hits the print?

@CarloK - new firmware bug on latest firmware (see ultiflow posts above). If you only use the right core, when the print is complete, it raises the right core and then proceeds to move to the front right corner hitting the part(s) on the print bed because the left core is now 0.7mm lower than the top of the print. This can pull the glass bed off the bed TO THE LEFT. @ultiflow - firmware bugs are harder to report. In case Carlo doesn't report this somehow, please create a ticket. So in the upper right corner of this page there are 9 dots. Click that, then "support". Then on the new page, just to the left of the 9 dots there is a button: "submit a request". Follow the instructions. Meanwhile I suggest you downgrade your firmware - or don't do "right core only prints" until this issue is fixed. I saw the video, thanks. If CarloK asks please give him access to the video also. Why are you trying to keep the video private? I didn't see any big secrets or anything.

Oh - I see you sent me a direct message. You could post that here as people will still need to ask permission to see it.

I really really really need to see that video otherwise I'm talking out of my butt. However I'll continue talking out of my butt: The way the right core goes up and down is using something called the "lift switch". If not calibrated the right core can be down when it is supposed to be up and vice versa. If it's down when you are supposed to be printing with the left core then the right core will hit all kinds of objects which can push the glass right out of the printer (pushing open those double doors) and onto the floor in front of the printer or it can push the glass left or right. I'm not convinced this is the problem as without seeing the video I'm not sure what you have been saying at all really. If you have gmail then google will host your video no problem. They call it "google drive". Or you can sign up for a youtube account and there is an "upload video" button and they really want to make it easy for you. Or there are hundreds of other hosting services but unfortunately not Ultimaker.

Yeah. I don't know either but seeing that code I'm 90% sure "layer" holds all the gcodes for a given layer. "split" is the same in javascript such that lines contains all the lines of gcode separated out. Layer is almost certainly a string type. Just a simple string. I'm sure that python (like javascript) has easy ways to insert something into a huge string like "layer". I suspect though that if you edit a single "lines" object it won't modify the "layer" string. Did later code put the lines back into a layer? Also in python I'm sure there is a way to just look at the final layer in the data object and not have to iterate through them all. In fact you could probably just add a string at the start of the final layer with no regex needed. No substitution needed. No splitting up of the individual gcodes needed.