gr5

Please check if you can heat up the print bed. On front panel go to middle menu on left and choose the bed and set the temp to say 40C and see if the temp slowly increases. If you 24V is gone it will not heat up at all.

Okay so the limit switches have nothing to do with it. Well the servos are tough as hell so it's the servo drivers. Is it in a hot environment? Servo drivers get very hot and can die. But it seems unlikely you would lose both X and Y at the same time (separate components). I'm going to look at the circuit diagram but I think you lost your 24V somewhere.

I just read the ER33 notes and it mentions 3 VERY DIFFERENT failure modes when homing the head. Which do you see? 1) The head doesn't move at all 2) The head only moves in one axis 3) The head moves in both axes but it doesn't quite reach the corner 4) The head reaches the corner but makes a loud shaking sound and flops violently like a dying fish for a few seconds.

1) When you press "Home Print Head" does the head move at all? Does it make any noise at all (in addition to just before you press the button)? 2) Is your S5 in a hot environment? Is the airt temp of the room it is in over 80F/26C?

This is underextrusion. It has many many possible causes. The most common is printing too cold and/or too fast. 1) Are you using default profile settings? Have you changed any of the temperatures or speeds? 2) If not I'd do a cold pull of the nozzle of core 2. Especially if this is a BB core. For a BB core do some hot and cold pulls as described in the maintenance menu. It will help you with the procedure. The purpose for all of this is that plastic can caramelize in the thinnest part of the nozzle (the last few mm above the tip) and make the diameter of say a 0.4mm nozzle become a 0.3mm nozzle which increases friction and causes underextrusion. 2b) Another thing to try is to put a different core in slot 2. Your printer should have come with two AA 0.4 and one BB 0.4 cores. It's okay to use BB 0.4 cores with PLA. Just slice for AA 0.4 and when you go to print it will complain that you have a different core and hit "ignore". 3) Feeder. I'd test the feeder. Pull the filament out of the print core so you can see it in the bowden. Then go to "move material" (it's in the menus on the printer front - look around - maybe it's called "move" but make sure for second core). No need to wait for it to heat up. Turn the dial and the feeder moves the filament. Now while doing that, fight the material below the feeder: pull down hard with one hand while spinning the knob to move the material with the other. You shouldn't be able to stop the pull. Unless you are very strong. Typically the feeder pulls with about 7kg or 15 pounds force. Anything over 9 pounds pulling force I consider "working fine". If you have lead weights you could old around 10 pound weight (or 7 or 8 pounds) and sort of grip the weight against the filament and then try not to add any more force up or down. Many issues are on the feeder end. Many issues are on the hot end. It's good to at least winnow the possibilities.

The normals are likely backwards. or maybe you have double walls. Blender lets you create impossible shapes with infinitely thin walls and so on. Anyway start out by reading this to fix your normals in blender: https://www.sculpteo.com/en/tutorial/prepare-your-model-3d-printing-blender/ Rats. The above link has changed a bit. There's lots of great advice but it's harder to find the normals part. Search on the above link for "recalculate". And read all the sections 3.X. Alternatively you can install the mesh tools in cura. Cura has an amazing plugin to test your model to see if something is wrong with it and can repair a very few of the many potential problems: In the upper right corner of Cura click "marketplace" and make sure you are on the "plugins" tab and install "Mesh Tools". Then restart Cura. Now right click on your model, choose "mesh tools" and first choose "check mesh", then "fix model normals" and "fix simple holes" to see if that helps. Cura doesn't fix most issues so... Also look at your model in xray view in cura and good things are colored shades of blue and white. Other shades are for errors in your model. Sorry I can't look at your stl right now. Bedtime for me.

For people who are confused, @Jermanator7777 posted a zip file containing some STL files and pictures of models that can be printed. @Jermanator7777 - maybe you can explain a bit about this. If you are looking for an alternative place to post models you could try thingiverse.com or youmagine.com. If you want to show off a model you printed, why not post the pictures (and not hide them in the zip file). In other words, please elaborate why someone might want to look at the STL files in your post.

Oh - I didn't realize it was polycarbonate. I don't know if that changes anything as I've only used PC once I think? Actually I think it was ABS mixed with PC. Maybe. Okay so I just read this. I recommend you do as well: https://support.makerbot.com/s/article/1667337602519 So they recommend the glue just like I said but even better use magigoo. There is a type specifically for PC. I've never used magigoo but I know many people who have and they all love that stuff. So buy some now. In addition it will help your part from warping off the bed if you use brim (or raft - raft works quite well but you end up with an ugly bottom layer). Also it helps if you can raise the air temp in the printer. Do you have the air manager? If not you can just put a box or something over the printer. It doesn't have to be air tight - just something to raise the air temp to 35C will help a lot. Most of the warping happens between the softening temp of PC and the air temp in the printer. I don't know where PC gets soft but I'll wildly guess 110C. So 110C to 20C is 90C difference. If you can raise the air temp to 40C that's now only 70C difference so significantly less warping forces trying to rip the print up off the bed. Hotter than 40C is I believe out of spec for the printer because the servos might get too hot (it's probably fine up to 50C but I wouldn't risk it).

Je ne vois aucun problème évident. En vue radiographique, les problèmes sont affichés en rouge et dans d'autres couleurs. Les nuances de bleu et de blanc vont bien. Il est possible que les normales soient inversées. Je vous recommande de le découper en tranches, puis de faire défiler les calques pour vous assurer que tout semble correct. Cherchez des trous remplis qui ne devraient pas l'être. Par exemple les orbites. Ou toute la bouche. Et recherchez les parties manquantes comme les dents manquantes ou d'autres parties du crâne. Si vous ne voyez rien d'anormal dans la vue PREPARE et que vous passez au moins une minute ou deux à la vérifier, c'est probablement correct. Si vous voyez des problèmes majeurs, recherchez "réparer les normales" et je pense que vous pouvez le trouver quelque part dans les paramètres pour les résoudre automatiquement et essayer de trancher à nouveau.

Oui.

Veuillez montrer une capture d'écran cura de ce à quoi ressemble votre modèle en vue radiographique.

Nothing strikes me as obvious. Was there no retraction for the PVA before? I'm looking at my default PVA settings and I see that the PVA only has 4.5mm retraction (default value) but the other material (nylon in my case right now) has 8mm retraction. I didn't touch those values. Maybe it's because nylon is so flexible and squishes a bit while printing so you need to retract more and PVA is stiffer? Anyway I didn't get much stringing at all on my last print. Maybe the gizmodorks PVA is much more flexible? I really don't know what it is that you changed that might have helped things. My print from yesterday where the PVA was nice and dry. If I don't dry the PVA I get lots of stringing just like you. :

I would order a new set. You might be able to get them for free since this problem is a known problem for certain printers with certain serial numbers. The set is not expensive. If you want to fix it yourself, I recommend you get some 2 part "5 minutes epoxy" and some string. put the string on tight around and around the block holding it closed tightly. Each wrap adds more tension (and more strength). No need for any knots. Then coat the string with epoxy. This will be strong as hell. I'd probably do this with the rod popped out. But over the years, eventually, all 4 blocks will fail.

I think they are pretty common but not 100% sure. That looks like good quality PVA which is rare on this forum. Your material handler is helping you probably more than you know. For me, my PVA gets "ruined" if I leave it on the back of the printer for a few days and has to be dried again which is not a trivial process but basically you put it on a heated bed at a certain temp for like 24 hours under a towel. But your PVA looks quite good. Almost clear. When it gets wet it boils and looks more white. More like snow.

Let's talk about #2. This is a critical issue and the most important one and the easiest to fix. There are many ways to prepare the glass. Since you mentioned the glue stick let's use that. spread it around sparingly in the area where the print and the tower will be. Then take a napkin or tissue, wet it, and spread around the glue. This will also remove 90% of the glue which is great. You want a very thin layer. Thicker layers of glue don't hold the part as well as thinner layers. If there is no glue at all then it won't stick so well either (and the glass would need to be cleaned after every 10 or so prints to keep dust and oils off). I usually only clean the glass about once per year as I have many thin layers of glue. Oh - and while doing this set the bed temp to 60C because you now probably need to wait for the wet glue to dry. At 60C it takes maybe 2 minutes. If you start at 20C then by the time the bed reaches 60C it may be dry. Leave the bed temp setting alone. I believe 60C for PLA is perfect. 1) This is normal. This is basically the purpose of the prime tower. 4) This is normal. PVA burns (caramelizes) very easily. Caramelized PVA is the leading cause of BB core clogs and results in the need for many hot and cold pulls on the BB core if it gets clogged or has a partial clog causing underextrusion (not a problem for you I think. Yet.). 5) You should be using engineering mode if you care that much about part size. And I also recommend 0.1mm layer height (not 0.06 which can cause quality to be worse - sometimes). In the profiles there are some that are "engineering mode". These give you MUCH better dimensional accuracy but also lower quality visually (more ringing noise around letters and corners). 3) "top" surface? Do you mean outer surface? maybe take a picture and circle what you mean?

I can't make any sens of that, sorry. I'd have to read the python code. @CarloK? Any ideas?

1) There are 4 pages of posts in this topic. Did you realize that? Did you go back to the first posts? In particular read the first post by tinkerGnome. And the posts right after. It gives one hints about how it all works. 2) I know almost nothing about the NFC tags but I know that some types of IDs have an uneditable serial number (that is globally unique) and that 100% can not be edited and if that is the case for NFC tags then that part can be ignored because it's useless for determining filament type. 3) There's nothing wrong with code from 2019. That might even be what I have on my UM3 right now 4) Finding the folder with all the code is a search through windy passages but once you find it it's all there. I was hoping you knew linux very well and could find it easily. Try these folders: /usr/share/griffin /var/lib/griffin I think the first one has the python files maybe (and in subfolders) And the second one has some configuration json files maybe? These are old notes and I don't remember details and I'm too lazy to turn on my UM3 and connect it to the network right now. Once you find all the python files you can do: grep -iR nfc * I think that will search the current folder and all subfolders (the "R" part) and the "i" will do a case insensitive search. If you change it to "*.py" it will only search python files but it will not go into any of the subfolders. For that you want something like "find". Just search all files for the initial search (starting in /usr/share/griffin or a subfolder of that) until you know what folder has all the nfc stuff.

You can ssh into the UM3 if you are good with linux and you can see all the code (it's all in python). So if you are a programmer and know python and with a little "grep"ing you can find the code and you can debugging to see what is going on. it may already be logging this information. To ssh to the printer you have to put it into "developer mode" in the menu system on the printer (probably under maintenance? I forget). Then it shows an ip address and you can ssh to that username/password root/ultimaker. Search for files that end in ".py". I forget where they are - it's a bit tricky but once you find the right folder it's a treasure trove. You can search for NFC perhaps. You can dump the Ultimaker python log files with journalctl. For example, to see all python errors since most recent boot do:journalctl -b google journalctl for more options. In there there *might* be some info already logged about the NFC tags. If not, and if you know python you can add your own logging to help you debug the issue. It sounds like you are putting in the right serial number but maybe something else is missing such as a checksum.

Another trick is to use G92 gcode. Here is a typical layer change - note the "Z 1.2" which means move the Z axis to 1.2mm such that the nozzle is 1.2mm from the bed. ;LAYER:6 G0 X119.348 Y137.166 Z1.2 What you can do is fool the printer telling it "hey you are already there" with the G92 command so: ;LAYER:6 G92 Z1.2 G0 X119.348 Y137.166 Z1.2 The G92 says: Hey I know you thought the Z was at 1.0mm but pretend it's actually at 1.2mm. Already there. So then when the printer sees the command to go to 1.2mm it will just ignore it. Just add the G92 for the layers where you want Z not to move. The bonus of this trick is it will work even if you have z hop enabled (but in that case don't search for Z as it will take forever, instead search for "LAYER"). Once you stop adding G92 gcodes the printer will continue moving up but only in small increments (it won't suddenly jump up for every layer you "locked").

You want to repeat a layer or two so it prints the exact same thing 2 or 3 things without the Z moving? Oh wait! You want it to print exactly normal except the Z axis doesn't move for 2 layers. Well that's pretty easy to just do it by hand in the gcode. Search the gcode file for the letter Z. Once you get past the first few there is only one Z in the file per layer and it's the height. Just take a look. Unless you have zhop enabled. If you do then that's unfortunate as there could be tens of thousands of Z's in there. But most of the time there is only one per layer so if you have 50 layers you only have to edit at most 50 numbers/heights. Take a look at the gcode! It's not that complicated.

1) Open the front door on the print head and look carefully inside. In the base of the head that is now swung down there is a flat plate (a simple PCB). There are 2 wires going to that. tug gently on both wires to make sure they are still connected. This was the problem for me. I just resoldered the loose wire and no problems since. 2) In the maintenance menu on the S5 there is a way to test the levelling sensor. Run that. 3) Keep RF noise sources away from the printer during active leveling. As a test maybe put everything that can generate radio noise at least one meter away from the printer. That includes: any kind of computer like a cell phone, computer, monitor, thermostate, smoke detector, almost anything electronic, even computer mice, keyboards, pretty much anything that has electrical power these days has a computer in it. Even toasters. Hopefully one of those three things will help. If not it's time to start replacing components and there are a lot of components to replace! So if 1-3 above don't do anything then contact your reseller. Alternatively you can just turn it off and do a good accurate manual leveling. I don't recommend this as the glass tends not to be perfectly flat: https://community.ultimaker.com/topic/39188-ultituner-a-tool-to-tweak-your-printer

Great news. Since you already installed tinkermarlin you can do a quick test. These power bricks tend to fail where they can't put out as much power as they used to. So you can lower your power usage by say 50 watts. In tinkerMarlin there is a feature "power budget". By default it should be: budget: 175W bed: 150W nozzle: 25W With the above values, the power budget does nothing. Try using the same values as above but lower the budget to 120W. This difference (175 - 120) should be more than enough for all the servos. If this works then you can increase the budget by 10W every minute until you find the point where the power brick (and printer) reboots. This will prove it's the power brick (in my opinion). Then you can choose to run the printer on a lower power budget (you might have trouble getting the bed hotter than 80C) or you can choose to get a new power brick.

The main heater is about 150 watts (maximum). The nozzle is about 30 watts. The 4 steppers total about 40 watts (if I remember right). That last bit is what puts things over the edge. The wattage needed is similar to what the power brick can put out. Just a little bit less. Some power bricks can't quite do everything at once (bed heater, nozzle heater, steppers). Does that make more sense to you? One way to lower the power is backwards from what you might think: Set the bed temp to a higher value. At 100C, the bed resistance goes up so it draws less (maximum) power and sometimes that will be just low enough that the power supply is happy. However if you are printing almost anything but nylon or ABS, then 100C is much too hot for a successful print. Alternatively you can set the bed temp to 0C and use blue painters tape (washed in isoprpoly or ethyl alcohol first to remove the wax). This can help provide evidence if the problem is the power brick or not.

What kind of printer do you have?

So did you also setup the power budget? If you don't mess with that then it was just chance it didn't power off in those 7 hours.