Transcript:
By a request, you’re going to adding dual extrusion to an existing 3d printer with all of the problem-solving that that entails [Music] [Music] one of my most frequent video requests over the last few years has been how to add a dual extruder set up to a single extruder 3d printer. And I’ve always replied that while it can be done, It’s not going to be easy a while ago. One of my patrons. Carl told me about an off-the-shelf dual extruder kit that definitely took my interest. It uses a servo to switch between two hot ends with a platform to block goo’s inbuilt. So you don’t need to waste filament on purge blocks or printed boost shields. It looked really interesting, so I ordered one and I have it working quite well right now. But at times, the struggle was truly real after installation was over. I had to replace broken components eventually. Find a bug in Marlin. I had to stop to design an all-in-one enclosure for this printer plus endless calibrations. So this was definitely not for the faint-hearted. Let me take you through the process in fitting this kit and we’re gonna start by looking at the product maker. Tech is the company behind this product. And as you can see, they sell two 3d printers, one of them being the Pro Forge 2s but comes from factory with the dual switching extruder. They also sell the extruder for any 3d printer as a kit, which is what I purchased for this video. It’s currently retailing for just under US $80.00 but it does come with a lot to hot end kits with fans and extruder mechanism, a stepper motor filament sensor, Additional Stepper motor driver, A BL probe, a bunch of hardware wiring and even a small spool of filament The other thing that I purchased and I would recommend if you’re going to follow. This Video is one of these tool carriages and built kits. There’s quite thorough documentation for each step of the process overall. I’d give the instructions about a nine out of ten, what’s. There is pretty thorough, but there were a couple of things. I found were missing. So how do you know if your printer is suitable for this kit? The first thing you need is a 2020 extruded gantry to suit the new carriage. You’ll want a fairly big printer as you can see. It sticks, outwards and down and I lost 50 millimeters from y&z. The ABL probe comes with this kit since this metal, only so glass won’t be any good, however. I’m a Don kit like the Wham-bam I have here will be sufficient. The biggest one is probably the mainboard we need a board like an SK, our version 1.3 or what I used an MKS Gen. L because it has five outfits for Stepper Motor Drivers. If you’ve got the standard end a3 or CI 10 board or you’ve upgraded to an SK our mini or th 3d easy board light. Your board won’t be compatible meeting. All of these conditions for me was a Tivo Tornado, which shares a lot of components with the CI. 10 and end of three. Assuming you have the right type of mainboard, you’d be able to follow on to the build, either of those printers. The kit assumes that you already have one extruder and therefore comes with an additional extruder for you to fit. I had already upgraded to an E 3d tried an arrow, which I disassembled to retrieve the Stepper Motor and found that it was almost exactly the same as the new one in the kit. I decided to scavenge from my Gtech 8 NM. These easiest Reuters. I’ve had a lot of success with them in the past and they’re really easy to mount on the top gantry of the printer now to assemble the hot ends and they come with III DV 6 clones. This means they’re not all metal, and the PTFE tube has to travel down inside all the way up to the top of the nozzle, I also noticed the included nozzles with 0.3 millimeter in orifice, So I changed mine for some spare 0.4 S I had lying around each hot end assembles exactly the same as an E3D v6 and it wasn’t long until I had both of them together and ready for the printer now for the main feature, which is the mechanism for the switching extruders. We start with this servo that needs to feed in through the top being careful not to damage the wiring loom. This is an extremely tight fit and I found it necessary to give it some very little love taps to help push it through. All of the mounting hardware is nicely labeled with correctly scaled images on the back for identification, m4 nuts and bolts are used in each corner to hold the servo to the bracket with these. The servo feels very secure next. We take the smaller bracket that holds the front cooling fan and we use more informants and bolts to attach it to the main plate. The two-part cooling blower fans can now be bolted to this assembly. Lock nuts are used throughout to stop the putt from vibrating free. After this, the auto bed leveling sensor goes into the included hole. The initial aim is to have it protruding around 20 millimeters. Next we prepare the bracket that holds the earth shields and some. Ptfe tape is included to wrap around and coat the surface. I’m not sure how well this will work long term, but at least this tape is cheap and easy to buy from hardware stores next. I decided to open up and assemble the new carriage. If we were building this for a pro forge printer, it would go together as you’re seeing here, However, on the tornado end of three and CI 10 the bolts on the back stick through too far and collide with the metal frame. Therefore, you need to flip the direction of the lower nuts and bolts moving the nylock nuts to the front of the carriage. This will allow the courage to reach the far right as well as the file f4 homing! It does cause another problem. However, in the next step, we bolt the courage to the frame and one of the lock nuts fails on the ABL probe. My solution was pretty simple. Use slightly longer mounting hardware than comes with the kit and spaced out the mounting frame from the gantry by a thickness of two washers. This made it very fiddly to put together, and it was hard to film anything besides the back of my hand, but once. I eventually did get everything together. The ABL probe just cleared the lock nut, but of course I had forgotten to install the ooze shield bracket, So I had to undo two of the bolts mounted in place and then talked everything back up. I was starting to make progress with the assembly. Next we have a pivoting bracket that holds the two hot ends, which swivels on a thrust bearing and bolts to the main bracket. The hole for this seems fairly accessible, But I fumbled for quite a while, trying to get the washer and lock nut through the inside of the whole thing. Eventually I got it there and was rewarded with a satisfying mechanism next. We’re meant to slide the two hot ends into these openings. This was incredibly fiddly and another thing. I’ve with for quite a while. One thing I really don’t like about this design is the way the two bolts are meant to retain the hot end as you can see as you tighten the bolts. They twist against it and try to push it out of position. My solution was to design and print this very simple mounting adapter. It slots on to the top of the groove mount for the hot end, and then the sub-assembly slides down into place. And this means when you do up the bolts, they line up with the holes underneath, and this prevents the hot end from squirming and moving too much. You can see that around this time. I also melted it to the machine and did my cable management for each of the two hot ends. You need to make sure your cables are long enough to reach when it’s facing straight down, but not so long that the cables get caught when it’s wiggled out of the way, you’ll also need to provide a long enough path for all of the cables wrapped in some sort of sheathing to be secured elsewhere on the printer frame With everything together, we can also see how the ooze shields block the path of the nozzles to prevent filament from leaking out when the other hot end is being used around this time. You’re also safe to install the PTFE tube that goes from the extruders on top to the switching hot ends. I found the part cooling fan shrouds that came with the kit. It’s at too low and we’re going to collide on the top of the printed object on the website. There’s these alternate versions that sit about a millimeter. Further up installation is very easy with both of them just clipping straight on, and this means we can turn our attention to something more serious. The wiring. The wiring for this conversion is a big deal. Here are all the factory wires. And then when we examine the new loom, we can see that it’s exactly twice as big. I personally find it very frustrating to work on the electronics for a Tebow tornado and probably Ser 10 is the same. You need to disassemble this box and then feed the old wiring out through the grommet. The old metal connector that used to go into the box for the hot end can no longer be used. There’s not enough wires here for two hot ends, so you’re going to need to come up with another solution? I used red and green heat shrink over each set of wires, so I could tell which one went to the left, all right, Extruder and hotend and here was my final wiring. It’s as per the instructions apart from the servo, which I plugged into the servo one slot instead of four and for the ABL probe, I plug that into the standard factory plug for this data in stock. One of the docks pages is dedicated to firmware and takes you through step by step with most of the changes that you need to adapt your existing firmware. – this kid. I say most because some things like safe homing we’re left out and there’s a link to my configuration files in the description below after Flushing the firmware, we can see that we now have two hot ends shown on the top-left of the LCD display. We can now install the little arm that switches between the two nozzles after we assemble it women to put it on to the end of the servo manually, move it through its range of motion and then connect to the computer, enter a gqo command to move it to 45 degrees before placing it vertically and doing up the three retaining bolts. I found the position for engaging the right Extruder was spot-on, but the left one needed tweaking more on that later on beyond this, there were a series of much simpler final steps. These are the type of things that. I’ve covered before, so I’m not going to go into any detail here. The more difficult and specialized parts are in the next section. The kick did come with a spool holder, but it didn’t really suit the way. I had the extruders so I borrowed the mounts from the Gtech 8nm. So that’s how it’s meant to go, But as I mentioned at the start, I had a lot of problems that needed solving. I’m going to work through them. Step-by-step here showing the solutions in case you’re facing the same thing. First up the separate control box it’s found on the CR 10 and Tivo Tornado after I reviewed this printer. I decided I wanted to make it all in one. But this project was the catalyst for finally getting it done. I took inspiration from Electra. Lenin’s design on Thingiverse by mounting the power supply and all of the mains voltage components vertically, the MKS Gen. L sits underneath the moving bed with two fans to keep things cool. There’s a thin lead to let the mains componets breathe and then a clip in panel to cover all of the mainboard wiring and it from hitting the underside of the carriage with the lid. In place you can see. There’s just enough room for the bed to. Cle and everything works as it should. I also designed some supporting pieces to help manage the cables, including the strain relief piece to hold all of my hot end wiring for the LCD. I stuck with a great design off Thingiverse. I still need to mount it. But this case from Adi Nozzle is a really nice design after I’d finished my wiring. For seemingly no reason the EBL probe decided to let out the magic smoke. Fortunately, I had a spare probe from when I made this truly video. The mini version is the same size, and therefore it went straight in works on glass to my next problem related to homing and that’s at the corner of the bed was nowhere near the nozzle at coordinates 0 0 Fortunately, this is easy to fix, use the LCD to move the nozzle into the position where you want it and then via terminal, you enter M 206 with your coordinates in my case, 5 and 50 which is enter this M2O 6x minus 5y minus 50 The next problem was another component failure and this time the servo stopped working and I needed to replace it. Judging by the maker tech forums, this seems to be a fairly common occurrence. A new one cost me 35 dollars. But by far the most annoying part was fitting it after I fitted it. I stumbled across another problem that after it moved into position after 10 seconds, it would go limp and this was disastrous as it meant that the nozzle was dangling out of position with results that you might expect from such a problem. There’s a line in the firmware, where we specify the angles we want for selecting the two nozzles, the angles of 0 and 100 Ament to correspond to where the arms should be facing. I later worked out that these angles were incorrect being too far apart from each other and jamming the survey. If, for instance, are shown in orange. You asked for an angle that the server has no chance of ever reaching, it’s going to run at really high amperage, straining to reach that target until it overheats and then turns itself off to relax, the trouble is. I don’t remember the original servo ever doing that, which might explain why mine stopped working and a bunch of others have over and failed as well, so I needed to set a better angle, But the trouble was whatever Angles are input in. Marlon would not actually change the angle selected by the printer. I wasted an entire weekend chasing this. But the solution in the end was to simply switch to the bug fix branch. I finally had full control over the angle of the servo. This means I can enter em to 80 P. Zero and then a servo angle until I found the ideal ones for each side. Hopefully you can see. I’m creeping up one degree at a time trying to find the exact engagement point after this will repeat for the second side again, inching one degree at a time, and when we found our two values, we put them back into the firmware and recompile, having the larger number as the main value selects the right extruder as the main one, the next problem was to calibrate the correct offset between the left and right hand nozzles. You can see here that the white is overlapping the red instead of coming up against it cleanly if you home the Machine and then put down some tape on the bed and position the nozzle above it, you can then change your nozzle on the LCD and see if it moves to the correct position. We then enter m-21 8 T 1x and then a minus value that we’ve determined after this trial and error Enter Em 500 to save once. You finally have the right number. Now that X is dialed in. We can run a dual extrusion calibration model, like the one available on their website to download and see how close the fit is and tweak our numbers further from there. It took me a few guys to get it spot-on. But my final values were minus 20 point to 0.4 and minus 0.2 my final problem, What blobs being deposited on the model despite the US. Yield in the worst cases. These would knock the model clear when the next nozzle slammed into them. The instructions state. You can bend the shields into place to make sure they’re sealing firmly, but this had no effect. I then noticed during printing that the blob was coming before the tool change, so it had to be slicer related maker tech. Have a custom version of cura. If you are using their printer which I was not and I have made a guide in the past for setting up a range of slices for jewel extrusion. In this case, I was using simplified 3d and I started with an inbuilt Gtech, a 20 M profile. I did things like to remove the mixing extruder and then come to the g-code tab and input the correct build dimensions, but the real problem for me was in the scripts for the tool change. You can see here It doesn’t retract. But then it tries to move to a new position and Prime before it actually gives the T 1 or T 0 command to perform the tool change. I ended up stripping out most of that, so all it did was the tool change retract, and then the T command to switch to the other tool and this completely fix a problem with the blobs gone and the models no longer being dislodged. The surface artifacts are from the ancient a 4 9 8 8 stepper motor drivers, So I might modify those next. But in terms of the dual extrusion, this model is very clean and I didn’t even need a purge block. It’s satisfying to finally get this working, but I’d be lying if I didn’t admit it was a nightmare at times apart from the broken servo and a BL sensor. I’m still a fan of this kit. Much of the problem solving. I had to do was because this extra set wasn’t really intended for this printer. But instead for a specific model, were just getting started on this because this setup has a lot of possibilities. You could have dual color like we’ve seen here. You could have a big nozzle and a small nozzle. You could have multi material. Maybe your PLA with the Flex you could print PLA on one extruder and dissolvable support material on another or you could even try the same material from two different nozzles at different temperatures like the foam filament recently tested on CNC kitchen. If you’re after these parts for the tornado all in one conversion, I’ll release those in the near future. If you’ve got any thoughts questions comments, please leave them down below. Thank you so much for watching and until next time, Happy, 3d printing. Gday, It’s Michael again. If you liked the video, then please click like if you want to see more content like this in future click. Subscribe and make sure you click on the bell to receive every notification. If you really want to support the channel and see exclusive content, become a patron, visit my patreon page. See you next time you.