rotorhead

Hey @Porkpie I am starting a similar project and looking at using either the SKR 1.4, the Mini E3 V3 or maybe the Duet 2 Wifi board. I was just wondering why you decided you didn't like the SKR 1.4 in the end?

I've had an Ultimaker 2 since 2015 and it's been generally great apart from a recurring and intermittent under extrusion problem. To try and solve this I have ended up doing lots of modifications and upgrades but nothing has really done the job. The problem is that it's all very sporadic and actually for the majority of my prints it hasn't been too much of an issue. However, I'm at the point where all the evidence is suggesting to it being a problem with the main board. If that's the case then I might have lived with a faulty board since it was new (which would be rather frustrating!). Anyway, the only way to know is to replace it but an original UM2 board is not only expensive but also now very old tech and so I was thinking about something else. Has anyone tried this? Having looked around the net, the BigTreeTech SKR Mini E3 v3.0 seems like a nice little unit but I've no idea how easy a job it would be having never attempted this before.

That's a very good point although I think that adding temperature probes to monitor the power brick as well as the heating elements, print head and enclosure might be beyond the scope of this project - for now! Maybe my project does need to be more complicated to be worth the effort but I was just trying to increase the protection to a reasonable level from a reasonable amount of work. I presume that the firmware has some heating protection built-in and I also assume that the main board has been designed to minimise the extremity of a failure but I don't know enough about it all and can't find the info online. My thought was to just remove an additional potential problem. If the main switch is switching a mosfet and this fails at the same time as something in the heated bed then potentially my kill relay would not stop the thing indefinitely heating itself up and catching fire? Or maybe this isn't possible?

I'm trying to work out how the power switch actually works. I've had a look at the circuit diagrams but I don't know enough about what I'm looking at to answer the question. Does it control a relay/transistor which then switches the power to everything else? I'm asking because I'm trying to work out what components could potentially fail and therefore where the emergency power-off relay should go.

Earlier this year I setup OctoPrint on a Pi 3b+ for the purpose of additional piece of mind during prints. I don't make a habit of leaving the printer unattended but occasionally it's necessary to be in another room. I also like the idea of many of the other features that OctoPrint adds to printers. I need to work out where to wire in a relay kill-switch - parallel with the power switch or on the main 24v supply line. My suspicion is that the power switch on the rear of the printer is actually operating a relay/transistor. My plan was to connect a Pi controlled relay in parallel to this switch however I wonder if this might not actually provide the reliable protection I am looking for. The relay is a last resort kill switch in case the printer is about to set on fire but if it's being used to control another part of the printer then could I have a double failure that would mean the relay doesn't actually turn the printer off? Could I have a failure of the temp probe in the hot end and also a failure of something else on the main board that would mean that my relay would not actually turn it all off and the printer would heat on full power indefinitely? If this is the case then I need to think about splicing in to the main 24v power cable and switching that with the relay instead but obviously this is much more complicated and also means that the full 220W will be running through the relay

So in which case and as I understand it, I need 175 150 50 and it's absolutely fine that the sum of the parts do not add up to the total? Presumably the 220-175 difference is so that there is some allocation for the steppers and main board? I have also really struggled to find any decent information on this for the UM2 but you are correct about that article being very good. I have just followed through the process as described there with the cooling fans on full, nozzle almost touching the build plate and in the XY centre and with the bed set to 50 degrees. There was no material in the head however and I was only heating to 100 degrees (for speed.) After about an hour of tinkering, I ended up at a set of PID values that were disappointingly close to the default values for the 30W heater! My trial and error got me to P:13 I:0.4 D:50 and 35W default is P:12 I:0.75 D:55 and actually this looks slightly better! The autotune suggested 23 1.2 55 I will start some prints and continue to fine tune for real world use and then update this thread for anyone else needing this. I did some screen shots and copied the PID values on to some of them for reference.

I have an ultimaker 2 With Tinker firmware 17.10.1 and have just fitted a 50Watt heating element. I moved the head to the middle of the bed, almost touching it and with the fans on and then ran a PID auto-tune using the firmware. This gave me p23.4 i1.2 d30.4 but I am still getting a 10 degree overshoot, 8 degree undershoot and therefore a very oscillating temperature for the first few minutes. Searching the internet and people are suggesting figures ranging quite significantly from 5.6 0.35 23 to 12 0.75 55. Is there another way of working this out or has anyone else found a solution that works? I feel like it should perform better but maybe not? I've also looked at the power budget. Default was 175W total 150W bed and 25W heater. Presumably I need to increase the heater to 50W but do I also need to increase the total or is this a fixed value based upon the 220W power brick?

This worked perfectly using a £1 switch from CPC. I have actually used a DPST switch and so will double it up to also switch the power to my OctoPi (taking the supply from the main 24v input). It's also now in series with a relay controlled by the Pi so I can have some temperature driven failsafes. I'll post a pic when it's done

Perfect, they were exactly my suspicions but the confirmation is ideal. Thanks for getting that information for me. Thank-you both

I'd like to move the UM2 power switch from the rear to the front. I already have a Pi running on the printer which has 2 status LEDs and a reset momentary switch and so it would be really nice to have all the controls in the same place. From looking at the original switch on the mainboard, it does not 'seem' to be very significant which makes me wonder if it is actually switching the full 24V 9A power supply or if it runs through a transistor or something? Has anyone done this before? Does anyone know what specs I need for an additional switch on the front. Does it really have to be able to handle 24V 9.2A? I guess I could just move the existing switch but getting all those tabs off the board without doing any damage does not fill me with excitement! Also, it wouldn't look that great on the front. Thanks

It's a UM2 and running your firmware! Yes you are probably correct in that just adding the end code as required might be the only solution - I just need to remember to remove it for large prints. I've tried to make octolapse do this but I could not see a way - hence thinking that Cura would be better. The problem really is that the bed drops all the way to the bottom and all the timelapse programs run on either time or z-movement as the trigger. It therefore needs to pause at the appropriate height in order for a photo to be generated to then make use of it in the post process. At the moment, my choices to end the timelapse are either the final layer (usually with the head in the way) or with the build plate home.

ah ok that's good. So I just need a command to move down by 50mm and then if it can it will and if it can't it will just hit the end stops and stop there?

I've just been searching and am amazed that this has not been covered before.......or I have missed it! I was hoping to be able to set the end code in Cura to lower the buildplate by a smaller amount rather than all the way to the bottom. This is purely for making nice time-lapse end shots in Octoprint so there might be a better way of doing this but I have not come across that either! On the face of it this seems simple enough by just using a relative command and then lowering by 50mm but this might cause a problem if I ever print anything fairly tall and the instruction will therefore tell the buildplate to try to go below the end stop. Would this be a problem on an UM2 or would it just stop when it hits the switch? If this does cause a problem then my thought was therefore some kind of query to lower by 50mm unless the Z axis is already within 50mm of the bottom. Is there any way of achieving this? I found this: prusa forum thanks

Excellent. another upgrade to get on with :-) although the grub screw holding the stock one onto the motor must be the cheapest that UM could find, none of my hex keys will grab it - and no they are not knackered. An upgrade for a later time then.

I've done an awful lot of fault checking, and upgrading on my UM2 recently to try and get it to print just right (the low friction spool holder, the Olsson block, door on the front for ABS, dust filter and a new filament feeder) and so have inevitably spent quite a lot of time getting to know my printer and have noticed that the feeder motor shaft is slightly eccentric! Initially I thought it looked like that simply due to the flat on the shaft but when changing the feeder I had chance to check it and it's definitely not rotating only on its own axis. Over 0.5mm ellipse, maybe up to 1mm, it's difficult to measure exactly. I have had some problems in the past with feeder motor skipping on PLA and grinding on ABS and was just wondering if this change in pressure from the motor could be the problem, or will the spring in the feeder adsorb this change in pressure from the eccentric motor? Thank all again in advance!