Mr Jos Posted July 27, 2023 Posted July 27, 2023 (edited) Hello all, Fixing the "horrible manual operated" machine handler with motors and a tactile remote! [Edited the title "[MOD] RC Material Handler set 42144 remixed" purely as I had proven my statement in the LR13000 topic, people are drawn to drama, no interaction if you keep it calm. Removed some explanation text, added a spoiler to the boring part, the video is now easyer to find.] I recently was invited to become a member of Robotmak3rs RLOC. As a Mindstorms remix project I took the currently available Material Handler, set 42144. For the automation I chose the Mindstorms EV3. Only open the spoiler if you are interested in how it was designed and have the time, it's photo heavy. Else just take a look at the video of it working under this spoiler, it's not a supercar, sorry. Reveal hidden contents The initial thought was to keep the model as close as possible to its designed shape and keep all the pneumatic cylinders to make the movements. This would be a difficult assignment, as most know, operating multiple cylinders/valves with 1 single pump is very difficult due to the pressure drop when opening another valve. In the base model the main arm would drop when lowering the sub arm. Another issue to overcome was by either placing the EV3 brick outside or inside the machine. But this model is not big, in contrast to the giant Mindstorms hub. I started by making a prototype of a pneumatic pump including a valve. One problem I had was that I only had the old 5.5L pumps, so designing a small compressor was not easy. The center distance was not perfect for the 5.5L, but it worked for now. Meanwhile I made an order for 10 of the new reinforced 6L Lego pumps, but it would take at least a month to receive them, so I continued testing with these. Tripled the design into an as compact as possible way. After a while I received the Material Handler and could start building the original model. With this I tried if the claws could pick up Technic Pins from a bin. It was a succes, so I tried the compressors and manual controls on the EV3. Then I started to try get the EV3 brick inside the machine, as this would be better then just laying it outside. Putting it above in the turret was no option, as the motors and pumps had to go in there, and it was to small already. First I put it upside down, with the screen downwards. But this made it to high, and wheels would not touch the floor when the outriggers were pulled up. So turned it around and made it part of the backside of the chassis. Wheels would now be able to touch the floor again, so the model could be driven around. I wanted it to still look realistic, so I had to compromise the ability to program the model during full assembly, to shorten the wheelbase. But I added quick release pins to take the chassis apart. As the new 6L pumps did come in, I had redesigned the motor/pump/valve setup. This did actually fit inside the back of the turret (with some extending at the back). Now I could start with the "operation panel" for this remix. I first tried with some gyro sensors attached to levers, to recreate the controls from inside the cabin. But the gyrodrift is insane and I could not manage to keep it operational for longer than 1 minute. So I quickly connected my PS4 bluetooth controller, that I already used for my 6 degrees of freedom robot arm, to the EV3. The compressor needed a little more reinforcement, so a few liftarms were added. Each time opening the chassis to reprogram the model. I added touch sensors to the base to restrict the turret rotation, so the model can not keep rotating and twist the cables to much. When it started to work as desired I placed it next to my pin sorting machine, to make this "Waste Material Handler" work in the environment it's used to. As everything was as wanted for now, I started designing the model in Studio. Exactly 1 day of 3D work was needed to get it done without hoses, 2nd day was just sculpting the hoses in Studio = Pain to do so.. And letting the laptop run to make some renders, to later be used in videos/photos/GIF's. Meanwhile I also checked which parts exactly are needed to make this Modification. The pressure gauge (manometer) and 2 hoses to connect it, are optional. As they are expensive, hard to get and not needed to make the model work. Some renders from this model; And final photos; I'll keep it short to keep the attention, here's the video when it's in action. Free instructions can be found on my Rebrickable page; LEGO MOC RC Material Handler PS4 controlled by Mr_Jos | Rebrickable - Build with LEGO Robotmak3rs Facebook Community page; Robotmak3rs , RM3 community FB page ROBOTMAK3RS Please leave a comment good/bad/criticism, as anything is better than no feedback on how to improve. Edited July 28, 2023 by Mr Jos No such thing as negative attraction, as the last 36hours.have proven Quote
2GodBDGlory Posted July 27, 2023 Posted July 27, 2023 Cool! What advantages did you gain by building it with EV3 as opposed to PF or PU? It's definitely a powerful system, and I'm looking forward to learning to use the EV3 I got recently, but it seems very bulky for this model, and maybe overkill? It looks like you're using it hands-off at one point in the video, so that level of automation would be a real advantage, especially working with other machines. Quote
Mr Jos Posted July 28, 2023 Author Posted July 28, 2023 On 7/27/2023 at 10:08 PM, 2GodBDGlory said: Cool! What advantages did you gain by building it with EV3 as opposed to PF or PU? It's definitely a powerful system, and I'm looking forward to learning to use the EV3 I got recently, but it seems very bulky for this model, and maybe overkill? It looks like you're using it hands-off at one point in the video, so that level of automation would be a real advantage, especially working with other machines. Expand The advantage to PF is that I get a better range, as the IR receiver does not work that great. And as extra problem, when I would bring this to events it might get trouble with other IR remotes around, as it would need 2 of the 4 channels to operate all 4 functions. Other advantage is the adjustable speed that is possible with the PS4 joysticks. When lowering the arms you want to open the valves slowly to prevent a sudden drop, this isn't easy with PF. PU is something I don't have anything for yet. No hubs and 0 motors. I did make the 3D designs for PF and PU motors in the triple compressor, but I decided not to launch them, as the interest is sub zero. I don't know if this is because the set is set as free instructions, but all my paid MOC's have done way better off the line. Overkill, for what I show in the video, yes. If you would completely make it automatic, like you say, make it grab pins and dump them in the hopper itself, and repeat over and over then the EV3 might be perfect for it. I did already program some definitions to do that, but pneumatics are very hard to repeat over and over again. You can't measure the pressure/angle of each boom, with linear actuators that would be easy, two fingers in the nose and done. The final reason I chose the EV3 is, because this MOD was made for RM3 to promote Mindstorms, but Lego knew what was happening I suppose, not enough people interested in the line anymore. It shouldn't be the price scaring people away from this.. I recently bought 4x EV3 education sets complete for 125€ each. That is incl the rechargeable battery, 3motors and a hub. Cheapest I ever bought was 70€, you just need to wait for the right people selling it when they need quick cash.. Hope to see what you have done with your EV3 soon! Quote
2GodBDGlory Posted July 28, 2023 Posted July 28, 2023 (edited) On 7/28/2023 at 7:22 PM, Mr Jos said: The advantage to PF is that I get a better range, as the IR receiver does not work that great. And as extra problem, when I would bring this to events it might get trouble with other IR remotes around, as it would need 2 of the 4 channels to operate all 4 functions. Other advantage is the adjustable speed that is possible with the PS4 joysticks. When lowering the arms you want to open the valves slowly to prevent a sudden drop, this isn't easy with PF. PU is something I don't have anything for yet. No hubs and 0 motors. I did make the 3D designs for PF and PU motors in the triple compressor, but I decided not to launch them, as the interest is sub zero. I don't know if this is because the set is set as free instructions, but all my paid MOC's have done way better off the line. Overkill, for what I show in the video, yes. If you would completely make it automatic, like you say, make it grab pins and dump them in the hopper itself, and repeat over and over then the EV3 might be perfect for it. I did already program some definitions to do that, but pneumatics are very hard to repeat over and over again. You can't measure the pressure/angle of each boom, with linear actuators that would be easy, two fingers in the nose and done. The final reason I chose the EV3 is, because this MOD was made for RM3 to promote Mindstorms, but Lego knew what was happening I suppose, not enough people interested in the line anymore. It shouldn't be the price scaring people away from this.. I recently bought 4x EV3 education sets complete for 125€ each. That is incl the rechargeable battery, 3motors and a hub. Cheapest I ever bought was 70€, you just need to wait for the right people selling it when they need quick cash.. Hope to see what you have done with your EV3 soon! Expand Thanks for the detailed reply! I'd imagine that PU would have been easier to incorporate and more accessible for builders, but then you wouldn't have had enough ports to add those touch sensors (though I imagine you could do without them?) However, the EV3 makes it a much more unique MOC, and I like that! I wonder if the requirement for four M-motors is turning people off? Putting bigger motors in would be pretty impractical, but I don't imagine a lot of people have that many of that motor lying around, given that a kit only came with one. You're right, you can get a lot of Mindstorms for not much money sometimes! I spent $90 CAD ($68 USD, if that's any more meaningful to you) to get all the electronics from the base kit except for a touch sensor and a cable or two, after my brother found it, with some other parts at a thrift store for $70 CAD! I've now bought a couple cables, an NXT motor, two NXT touch sensors, and a NXT ultrasonic sensor to add capability, and can hopefully get started on it once I finish up some other projects. You mentioned programming it with PyBricks- I had been planning on using Python for programming, but the documentation I was finding online was all for an official solution that involved putting alternate firmware on an SD card and leaving it in the EV3 indefinitely. Are you familiar with this option? If so, do you think it or PyBricks is the better choice? I've used PyBricks just a little for some PU stuff, but hadn't really thought about it being compatible with EV3 Edited July 28, 2023 by 2GodBDGlory Quote
Mr Jos Posted July 28, 2023 Author Posted July 28, 2023 On 7/28/2023 at 7:32 PM, 2GodBDGlory said: Thanks for the detailed reply! ... You mentioned programming it with PyBricks- I had been planning on using Python for programming, but the documentation I was finding online was all for an official solution that involved putting alternate firmware on an SD card and leaving it in the EV3 indefinitely. Are you familiar with this option? If so, do you think it or PyBricks is the better choice? I've used PyBricks just a little for some PU stuff, but hadn't really thought about it being compatible with EV3 Expand If you want to program with MicroPython on the EV3 (= with PyBricks), you will indeed need a micro SD card. It will not work without it. You just flash the SD with the file you can download on the Lego education website with a simple software like BalenaEtcher. Then you put this SD inside the EV3 (Tip!: Put a little bit of tape on the end of the SD, so you can pull it out easily later). If you pull the SD card out, start up the brick it just works the same as before, nothing gets changed to the original firmware. If the SD is inside and you start it up, the PyBricks firmware will launch. By putting the USB programming cable in the top of the EV3 you can download a program you make with Visual Studio Code (VSC) to the EV3. When the cable stays connected you can debug your machine by printing some lines on screen with data you want. If you remove the cable, you can still launch your downloaded programs without a connection to a computer needed. PyBricks is just, very powerfull, together with the EV3 you can make very fast responding machines and do complex calculations, multitasking etcetc. Quote
2GodBDGlory Posted July 28, 2023 Posted July 28, 2023 On 7/28/2023 at 7:41 PM, Mr Jos said: If you want to program with MicroPython on the EV3 (= with PyBricks), you will indeed need a micro SD card. It will not work without it. You just flash the SD with the file you can download on the Lego education website with a simple software like BalenaEtcher. Then you put this SD inside the EV3 (Tip!: Put a little bit of tape on the end of the SD, so you can pull it out easily later). If you pull the SD card out, start up the brick it just works the same as before, nothing gets changed to the original firmware. If the SD is inside and you start it up, the PyBricks firmware will launch. By putting the USB programming cable in the top of the EV3 you can download a program you make with Visual Studio Code (VSC) to the EV3. When the cable stays connected you can debug your machine by printing some lines on screen with data you want. If you remove the cable, you can still launch your downloaded programs without a connection to a computer needed. PyBricks is just, very powerfull, together with the EV3 you can make very fast responding machines and do complex calculations, multitasking etcetc. Expand Sounds good! I was expecting PyBricks to be Bluetooth based for it like with PU, but cable-based is probably quicker anyways. It sounds like maybe that's what I was looking at before; I just didn't realize it was PyBricks at the time. Thanks! Quote
Mr Jos Posted August 2, 2023 Author Posted August 2, 2023 @eric trax well, here it's on topic to try change your mind about pneumatics. As it uses only pneumatics for the 2x boom + end effector. I don't think it lacks precision? Both up/down and with enough power are possible. About taking up space, ok the compressor part a bit yes, but you can put it wherever you want, you don't need a complex geartrain that goes through the exact middle of your model where it hinges. Just run 2 hose lines for each cylinder, this will free up some space in the model. I did make a dual compressor unit (pretty compact and strong, 2motors for 2valves+cylinders) for every motor possible (PF/PU/EV3) If you could put that in your model, and in the middle you have space to put your drivetrain motor/axle/gears etc, it's not that big? And you have 2 independent movements with pneumatics, that are precise. Quote
gyenesvi Posted August 2, 2023 Posted August 2, 2023 This is pretty cool, I never thought pneumatics could be controlled with such precision. What is the key to that, for a single valve? Is it that it's opening the valve slowly? It's opening them fully though, right? Can it make small adjustments as well? And I still don't get what prevents a valve from dropping the arm when it gets open in the downward direction? I mean I am not even sure why it would drop in a manual system. Is it because there is pressure in the tubes and that is put to the valve too suddenly (in this case I understand how a slow opening can prevent the drop), or is it because opening the valve lets the air flow out and the weight of the arm just pushes it out immediately? I hope it's clear what I am trying to ask.. Quote
Mr Jos Posted August 2, 2023 Author Posted August 2, 2023 On 8/2/2023 at 8:46 AM, gyenesvi said: This is pretty cool, I never thought pneumatics could be controlled with such precision. What is the key to that, for a single valve? Is it that it's opening the valve slowly? It's opening them fully though, right? Can it make small adjustments as well? And I still don't get what prevents a valve from dropping the arm when it gets open in the downward direction? I mean I am not even sure why it would drop in a manual system. Is it because there is pressure in the tubes and that is put to the valve too suddenly (in this case I understand how a slow opening can prevent the drop), or is it because opening the valve lets the air flow out and the weight of the arm just pushes it out immediately? I hope it's clear what I am trying to ask.. Expand Normally if you have a heavy weight pushing on a cylinder (like an arm here); Opening the valve completely fast will let out all the pressure build in the bottom of the cylinder (the pressure was needed to push the weight up) through the valve This results in no power left to keep the arm from dropping down on its own, but.. at the top side of the cylinder there was no pressure, that side is now closed by the valve in connection with the pump, so you will create a vacuüm. This vacuüm results in a slow lowering of the arm, by slowly pumping extra air in the top of the cylinder you can control the lowering speed. Going back up is easy, switch the valve and start pumping to build up pressure. If the weight of the arm is lower (like the 2nd arm in this model), the vacuüm will be enough to keep it in place very accurate. Slowly opening valves only make it even better to control heavy arms. Quote
gyenesvi Posted August 2, 2023 Posted August 2, 2023 On 8/2/2023 at 9:50 AM, Mr Jos said: Normally if you have a heavy weight pushing on a cylinder (like an arm here); Opening the valve completely fast will let out all the pressure build in the bottom of the cylinder (the pressure was needed to push the weight up) through the valve This results in no power left to keep the arm from dropping down on its own, but.. at the top side of the cylinder there was no pressure, that side is now closed by the valve in connection with the pump, so you will create a vacuüm. This vacuüm results in a slow lowering of the arm, by slowly pumping extra air in the top of the cylinder you can control the lowering speed. Going back up is easy, switch the valve and start pumping to build up pressure. If the weight of the arm is lower (like the 2nd arm in this model), the vacuüm will be enough to keep it in place very accurate. Slowly opening valves only make it even better to control heavy arms. Expand Interesting info, thanks! Quote
Mr Jos Posted August 12, 2023 Author Posted August 12, 2023 @eric trax has your mind changed yet about pneumatics? Quote
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.