I power on the printer for the first time and set up the Limit Switches, Axis Movement, End Stops, Bed Extents, Bed Levelling, Z Probe/Bed Sensor and do the Extruder Calibration!
I spent quite some time going through the Marlin documentation to see what each possible configuration instruction could do. I had already done a very basic run through of the Configuration.h file in this video/post but now it was time to get it spot on.
One side benefit of the exercise is that I can see I will need to change the way I mount the heated bed. Most importantly the nyloc nuts sitting on top reduce the Y axis extents by nearly 25%!!!!. But secondly the ply wood under the bed allows the nuts to squash the edges down too much, resulting in the opposite bed bend to that I had before! Still, good enough for a first print 😉
As I am not sure yet exactly where the RAMPS will end up, I am just going to wire everything long and hook it all up temporarily to get the printer up and running. I will revisit the wiring once everything is working and tidy it all up. But this is good enough for now!
I cover a few little things that I needed to look up, pin outs etc. The rest was an easy job… so long as you have the right crimps to hand!!!
If you have a Z probe that needs more than 5v to operate, then you can run it directly from the 12v supply and run the signal wire (usually black) through a voltage divider and from there into the RAMP Z Min Endstop connector.
The values for the voltage divider are R1=10kOhm and R2= 6.8kOhm. This will drop the 12v to 4.9 and result in a power loss of only 85mW.
I didn’t end up needing to do this – although the probe was spec’d for 6-36 Volts (if I remember correctly) it worked fine with just 5v.
So before getting all carried away and wiring in the Arduino and RAMPS I thought it might be a good idea to make sure the Arduino actually works!
As I already work with Arduinos, I have the Arduino IDE installed. If you don’t you can go and grab it from here.
I uploaded a simple blinky sketch and all looks good so I also create a short sketch to wipe all data, if any, lurking in the EEPROM. Just in case it later confuses Marlin.
After downloading the latest version of Marlin I spent quite some time going through the configuration.h sketch file making some basic changes for an initial start up. If anyone is interested in a copy of that, you can download it here. (You might want to change it so my name doesn’t show up on your printer’s LCD all the time!!)
There are a lot of resources on the internet for the configuration of Marlin, here are the main ones I used:-
I am starting to shift away from the Prusa build manual now as I move into the parts which are quite different to the original’s. Sadly the heated bed is one of those parts with the MK42 not really available yet, so I have to make do with a 3mm MK3 from Banggood.
Of course it is not flat, about 1mm lower in the middle than the edges, so I make a support to try and pull it level as it is tightened down.
I also put a piece of 3mm ply between the support frame and the bed, mostly for thermal insulation.
Quick solder job and the wires are on and the thermistor is stuck in place with common silicone sealant. I see some people use high temp sealant, not sure why. Your average bathroom sealant (RTV) is good for a constant 200°C which this bed will never get to.
I have also seen people use huge wiring for the heated bed. I used 2mm² wire which is good for 25 Amps. If I run the bed at 24V (which I intend to) it will max out around 8 Amps… with losses of 70mV and 560mW. So I am OK with that.
That is all of the initial mechanical build done. Next up – the electronics 🙂
As I mentioned in the last post, after building up the extruder/hot end assembly I tried to feed some filament into it as you would in normal use. I noticed that this could only be done by opening the tensioner and having to use a screwdriver to push the filament so that it lined up with the teflon tube.
Shaded area represents the straight path
Even once you have done that, because of the forced curve in the filament around the gear, it is very tight to push the filament down through the hot end.
The basic issue is that hole in the extruder body which fits the teflon tube and the hole in the top where you insert filament are offset from the surface of the hobbed gear. I am going to guess the the MK8 hobbed gear is a slightly larger diameter than that the body part was designed for.
I have already filed the left side hole out in these photos
By drilling/filing out both holes I have reduced the problem considerably, moving the filament is a lot easier but it is still not great. I know that the extruder stepper and driver already are a bit limited on power so the additional friction cause by this situation isn’t good.
I will see how it works once I start printing. There are solutions, but nothing that simple. You quickly run out of material around both holes if you drill out anything more than 5mm. Also there is then nothing holding the teflon tube in place.
Aside of being careful to keep everything lined up, parallel & square the build of the Z Axis, mounting the Y Axis and putting all the extruder and hotend parts on was pretty simple and in line with the official Prusa Manual.
Really the only thing I had to find a solution for was mounting my lead screw nuts. Toms version uses 5mm threaded rod and he provides STLs for parts to hold a M5 nut for that. The original MK2 looks like they use their own lead screw nuts. The ones that came with my lead screws had their mounting holes too close together and there was no way to centralise the nut in the Z carriers. I customised Toms Z nut holdahs to do the job – which it seems to do fine.
I will say this though, after building it all I examined the extruder to “sanity check” it and it turns out that it was very tricky to feed filament through the extruder and into the telfon tube running to the hotend. I will do an extra post discussing that as I think it is a mis-match between the extruder body design and the commonly available MK8 hobbed gear. I think if this was left unchanged, apart from having to use tools to get it to feed in, it would put a lot of load on the extruder stepper just to overcome this tightness.
I would also recommend that if you can squeeze an extra £20/$30 to get a Z axis frame laser cut – do that. It will save you a lot of time and insecurities about the Z axis being parallel and square. My wood will probably work fine, but it’s just a lot harder to make sure it is correct.
I thought this part was going to be simple – all pretty much in line with the Original Prusa Build Manual. There were a couple of complications though, one minor and one slightly major 😉
The minor one was that the left side X Axis ABS part was modified by Tom in his build to take the limit switches included in his, and my, parts list. The only snag is that the ABS is quite thin and has to act as a threaded part as there is not enough room on the back for a nut. I expected that the screw itself would cut the thread OK, but it started de-laminating so I had to drill and tap it. Not a major issue.
The slightly bigger job was I had to cut slots in the front (from front of printer) side of the extruder carrier to be able to thread through the tie-wraps which secure the bearings in. The idea is that they pass through hollow channels designed into the part but mine were bunged up with infill.
That last one is definitely worth checking before you start – its quite a lengthy job to do it “carefully” if you cant push them through.
Other than that – this section is simple and quick.
In the last video I noted I could’t work out how the Y limit switches fitted. In the course of investigating I noticed a bigger issue – the 3D Printed parts on all four corners of the Y Axis should have cutouts in the top to let the linear bearings partly slide “into” the corners….. mine didn’t.
Turns out that there is extra material in the prints that needs to be removed – revealing these cutouts!!! If you know you know – and I didn’t 🙂
I’d like to think this is an easy mistake to make – so worth mentioning here and in the video.
Sadly it means that the Y corners were also the wrong way around, so will have to take the Y Axis apart and make the corrections. Doh!
The good news is that it should fix the issue with the limit switch. Every cloud and all.
I decided to get on and cut the frame out first – made from two identical pieces of 12mm plywood glued and screwed together to maximise rigidity. ( I used 2x12mm as I had a sheet of 12mm lying around)
As I discuss in this video – I found the PDF template found on Thomas Sanladerer’s site a little bit tricky. It might be that its just lost quality due to scanning or something but the exact cut lines are hard to figure out as there is a double line around the edge and the hole marks are also slightly vague. On top of that I found it quite awkward to line up when printing on multiple sheets of A4.
So I made my own! Hopefully it helps out someone in the future looking to cut their own Prusa i3 frame.
As you can see, it is designed to be printed onto 6 A4 sheets – then each sheet should be cut along the green dotted lines. The edges these leave should be butted up against the red lines (not to be cut!!). Tape it all together and there you have it.
You will notice that this is slightly different to Tom’s – not in dimensions but in corner treatment. I have changed all internal corners into 5mm radius arcs. The idea being that you drill all these points with a 10mm drill before cutting out. This is much easier to do that trying to make accurate tight angles like the original.
One thing to note; I have only included the centre marks for the Z Axis top and bottom mounts. The rest are either for tie wrapping wires to or mounting the power supply/control boards. As none of that is critical, I left them out.
The other thing I might suggest is to only drill the two Z axis mount holes at the very bottom, then drill the rest using the actual 3D printed parts as a template. The reason being is that, with all the best will in the world, it is unlikely you will drill them all accurately from the template. As Tom found out.
My method was to drill the two bottom holes, use those to bolt in place the bottom Z axis mounts, then attach the smooth rods and top mounts and align everything up with a square. Mark the position of the rest of the holes through the 3D printed parts and then drill them.