Fixing motor shortages! + Whack a mole game

[Mr Jos]

EDIT: Instructions are available now on Rebrickable and it can play whack a mole!

/EDIT

 

For a MOC I'm making I want to open several gates individually, around 4 to 20 of them. I don't want to use 20motors (many hubs and motors needed) and it would become to big near the gates. So I thought about using pneumatic cylinders = small at operation location.

The valves can be located at a further location, but the same issue, I don't want to have an autovalve (=1pump+1motor+1valve) for each cylinder, as in my latest finished MOC the RC Material Handler. To reduce the amount of needed motors I've been making several prototypes in the last half year already, but all failed to be reliable and use only 2motors (needed most times 3motors)

I came up with a setup where 1 motor has on 1side a pump, and the other side operates the valves individually. The second motor makes the movement from 1 valve to another possible.

Correct me if I'm wrong, but I think this is the first pneumatic setup that can operate more cylinders than there are motors.

Air pressure can be done in advance, when waiting for an operation at a complete safe location (0:04).

Valves can be opened (0:14) and do some extra pumping if wanted at a small position in between valves (0:16).

Cylinders can go in and directly extend again (0:27).

Or just make all cylinders move in without extra pumping (0:32).

This prototype also shows why encoders are needed, both need to be at exactly the correct location or it would break. If the lever switching the valves is not 100% flat, it will crash into the valves levers.

Edited July 31, 2023 by Mr Jos

[Lok24]

Really nice work!

For my projects with similar functionality I use clutches or touch-sensors to calibrate at program start.

How did you solve this?

[Davidz90]

Wow, that's ingenious! 

Interesting solution to the problem of motor shortage. I guess another possible way would be a distribution gearbox; one motor operates wave selector, one is the input, several outputs move valves through slip clutches.

[Mr Jos]

23 minutes ago, Lok24 said:

Really nice work!

For my projects with similar functionality I use clutches or touch-sensors to calibrate at program start.

How did you solve this?

I just used PyBricks ".run_until_stalled" to find the home position of the carriage. The valve lever I don't have an automatic homing for yet, as it's just a prototype. For now I just make that it's flat at the start (normally when a program is ended it's still flat as I force it to hold within 2° of the starting angle, this is also a feature in PyBricks ".control.target_tolerance(1000, 2)". This will ensure that the end location is between 358° and 2°, so I can always go through all valves safely.

27 minutes ago, Davidz90 said:

Wow, that's ingenious! 

Interesting solution to the problem of motor shortage. I guess another possible way would be a distribution gearbox; one motor operates wave selector, one is the input, several outputs move valves through slip clutches.

This would work for max 4 valves yes with the orange wave selector. But it also would make pumping air always having to make the clutch slip=more power needed=faster broken motors.

This way I have now, I just need to extend the guidings and add more valves, and I can pump "free" in between the valves without having to power clutches. Another advantage is that I can put all valves back in center position, so they are closed and don't move anymore. Using clutches it's guessing when it will reach center again.

[Lok24]

4 minutes ago, Mr Jos said:

I just used PyBricks ".run_until_stalled" to find the home position of the carriage.

Ah, very nice! Good Idea.

Really great.

[Davidz90]

7 minutes ago, Mr Jos said:

This way I have now, I just need to extend the guidings and add more valves, and I can pump "free" in between the valves without having to power clutches. Another advantage is that I can put all valves back in center position, so they are closed and don't move anymore. Using clutches it's guessing when it will reach center again.

I see. Yes, that is a considerable advantage.

[pleegwat]

46 minutes ago, Mr Jos said:

This would work for max 4 valves yes with the orange wave selector. But it also would make pumping air always having to make the clutch slip=more power needed=faster broken motors.

Or three valves plus a pump section. 7 plus a pump section if you use the new selectors from the yamaha. But your setup is definitely easier to extend - and may be more fun to watch too on the show floor.

[Mr Jos]

1 minute ago, pleegwat said:

Or three valves plus a pump section. 7 plus a pump section if you use the new selectors from the yamaha. But your setup is definitely easier to extend - and may be more fun to watch too on the show floor.

Yeah using the wave selector and dedicated pump position would be better than clutches indeed.

Fun it is certainly, I just programmed a color sensor to operate the valves. If I still have this prototype in 1 piece when the next event is I will certainly take it with me to let children play with it by holding a matching color flag above the sensor to extend a cylinder. Just been doing it myself for a few minutes.

[2GodBDGlory]

Very nice idea! It's certainly large, so probably not something that will likely be seen in scale models, but I assume you're going for more of an industrial machine, so it's not a problem! The indefinite expandability is a nice feature, and it certainly does look cool in operation!

[BusterHaus]

Excellent work Mr Jos, I really like that the solution allows controlling the "on" timing of the valves and the amount of pressure built up before the next valve is open. Would you have a picture of the horizontal travel mechanism? It looks very smooth. 

[vascolp]

This is pretty cool and nice to watch!

I guess you could also replace all the air stuff with linear actuators? Maybe slower to move each gate, but you could do it with only one motor! Also, it’s probably easier to route tubes than axles if the gates are too far away or in strange positions.

[gyenesvi]

That is a pretty ingenious solution for the use cases where you have plenty of space!

[Jurss]

Cool solution! Mechanically simple, but that should be genious solution.

It will have some length limits, and for 20 it will be already pretty long. As I see, chain is used for movement, maybe that would start to be weak point for that length already. 

[Mr Jos]

16 hours ago, 2GodBDGlory said:

Very nice idea! It's certainly large, so probably not something that will likely be seen in scale models, but I assume you're going for more of an industrial machine, so it's not a problem! The indefinite expandability is a nice feature, and it certainly does look cool in operation!

Yes, it will be for one of my industrial machines indeed. First I wanted a multivalve solution for my gantry pinsorter.

Pins would be put in the top bins, and cylinders could open up the trapdoors, 1 by 1. Conveyor underneath would transport the pins to bag them.

But now I'm actually trying to make a ball sorter GBC that stores the balls temporary in my high bay warehouse. For this I need multiple 30ball waiting locations, that are closed by gates. I think this is a good solution to open those gates and close them fast again.

15 hours ago, BusterHaus said:

Excellent work Mr Jos, I really like that the solution allows controlling the "on" timing of the valves and the amount of pressure built up before the next valve is open. Would you have a picture of the horizontal travel mechanism? It looks very smooth. 

Yes, few photos here, and I think I will make instructions for this one. At least there's more interest in this design (or it's just because I posted it separate and not in my WIP topics, or if this was in a directly completely finished MOC topic).

13 hours ago, vascolp said:

This is pretty cool and nice to watch!

I guess you could also replace all the air stuff with linear actuators? Maybe slower to move each gate, but you could do it with only one motor! Also, it’s probably easier to route tubes than axles if the gates are too far away or in strange positions.

Slower certainly as it would need multiple rotations for the actual opening/closing. With pneumatics you can 'preload' the force and quickly apply the power. Routing the tubes is indeed easy to do at far locations instead of axles going everywhere + you would get a lot of gear play. Pneumatics are lovely and great to use, but expensive

11 hours ago, gyenesvi said:

That is a pretty ingenious solution for the use cases where you have plenty of space!

Most of my MOC's are massive indeed, so space enough. Still need to invent the extra GBC layout, but will have to take this in account to put it somewhere nice (it's great to watch on its own already)

3 hours ago, Jurss said:

Cool solution! Mechanically simple, but that should be genious solution.

It will have some length limits, and for 20 it will be already pretty long. As I see, chain is used for movement, maybe that would start to be weak point for that length already. 

Every 12studs long (between the 5x11 frames) I can put a reinforcement on the lower side to keep the side tracks together, so track length/sturdyness is no problem. The small chain is also relatively low weight and the carriage being pulled forward/backwards is always the same weight. I use this type of chain in a lot of my MOC's, they are very usefull.

[Toastie]

This is an incredibly nice mechanism and solution! I love pneumatics, and yes everything needs to line up precisely, otherwise bad things will happen, as you said!

Cool

Back in the days, when encoders were hardly found in LEGO motors (the RCX era and before) such precise alignments were also possible using the light sensors and a light of course (https://www.eurobricks.com/forum/index.php?/forums/topic/154740-moc-lego-switch-drive-controller/)

All the best,
Thorsten

[Mr Jos]

After a long day designing this MOC in Studio, searching currently available pneumatic hoses in correct colors on Rebrickable and molding them with the "great" Studio flex parts.. I have finished instructions, renders etcetc for this model. (Most of my MOC's, including this one, I make for myself so I can remake it when I want it later, for an event, or for repairs/cleaning etc)

LEGO MOC Quadruple individual pneumatic valve control with 2 motors only! by Mr_Jos | Rebrickable - Build with LEGO

I programmed it so when you show the sensor a color, it will extend the corresponding cylinder untill you remove the color from the sensor. My son (1year 4months old) has been playing with it all day whilst I was modelling in Studio, took him 20seconds to know how it worked! So it had a good stress test already.

Now this is finished, I can break it down to try to get 5 or 6 valves inside, so I can use it in my GBC.

 

For anyone interested, Rebrickable/Bricklink has these cylinders/pumps/ some hoses very expensive. But they are for sale at Lego Pick a Brick! Around 5€ each only, but comes with 6weeks shipping time at least. But the total price will go much lower.

Edited July 30, 2023 by Mr Jos

[Mr Jos]

And I made a little menu in the program, now it has 3 modes (2game modes [Whack a mole and just playing with the sensor for the little ones] +1 preset routine)

Using many of the EV3 capabilities. Speech, accurate positioning, random generator, onscreen menus, etcetc

WHACK A MOLE

Now the breakdown will really start.

Edited July 31, 2023 by Mr Jos

[gyenesvi]

Really cool! I did not know that pneumatic hoses are available in Studio, how did you do that? Are they available like other parts, or is there a teick for using them? (I cannot check Studio now)

[Mr Jos]

2 hours ago, gyenesvi said:

Really cool! I did not know that pneumatic hoses are available in Studio, how did you do that? Are they available like other parts, or is there a teick for using them? (I cannot check Studio now)

Thanks, yes they are available in Studio, but not all current ones, and by the same numbers. You can find them by "75c##" , where ## is the stud length of the hose (1L = 8mm, so a 10L hose = 80mm , will be 75c10 . Then later in the instructions + on Rebrickable I have overwritten them by currently available hoses+colors from sets sold starting 2015-2023, so a lot of work one might not notice, but it was there. And starting to model them in Studio might be a pain, but the more you do it, the better you get at this.

Warning before you try it (don't do it if you have a swear jar at your home):

• Hoses don't snap to cylinders.
• Hoses don't snap to valves.
• Hoses don't like to do what you want at all.
• You need to learn that they go far in the depth, and make very small movements, as it crawls like thick tar.
• First position one end at the perfect spot, then start taking the other end in bending mode, and try to get it where it should end up, this is one of the hardest parts, because they don't snap.
• Now the fun part starts, start pulling the middle part in bending mode, ready to press ESC key to stop it crawling. Go back in bend mode to pull more/let it crawl. If you need to go through small holes is the hardest.
• They don't stay in place, so if you need to go through more than 1 hole, good luck! And have enough €'s for the swear jar.

Example from my inventory page, where I added the correct number.

And how a step looks like with these hoses, length etc is shown enough I guess.

Edited July 31, 2023 by Mr Jos

[gyenesvi]

5 hours ago, Mr Jos said:

Thanks, yes they are available in Studio, but not all current ones, and by the same numbers. You can find them by "75c##" , where ## is the stud length of the hose (1L = 8mm, so a 10L hose = 80mm , will be 75c10 . Then later in the instructions + on Rebrickable I have overwritten them by currently available hoses+colors from sets sold starting 2015-2023, so a lot of work one might not notice, but it was there. And starting to model them in Studio might be a pain, but the more you do it, the better you get at this.

Hmm, thanks, I suspected it would be PITA (same as other flexible items), not saying I want to try it immediately..

[ludov]

75c## are rigid hoses, not pneumatic...?

Really cool solution for switching. I'm wondering if something like a Oldham or Schmidt coupling can help with misalignment?

 

 

Edited August 2, 2023 by lcvisser

[Mr Jos]

12 hours ago, lcvisser said:

75c## are rigid hoses, not pneumatic...?

Really cool solution for switching. I'm wondering if something like a Oldham or Schmidt coupling can help with misalignment?

Indeed rigid hoses, that's why I change them in the inventory of Rebrickable, I delete the rigid hose and add the correct pneumatic one. But in the instruction they look nice like pneumatic ones, as the 5202c## which are pneumatic hoses are not available in Studio.

I used the Schmidt couplings at my work before, but don't see what they could do in this case? I did use a Lego version before to try and make a paper feeder, with both rolls driven (top&bottom) and the top roll could open up. Everything in this machine is aligned already.

[ludov]

11 hours ago, Mr Jos said:

I used the Schmidt couplings at my work before, but don't see what they could do in this case?

I was thinking about a setup where one end of the coupling would be "free" and somehow lock into the part connected to the pneumatic switches. If the motor assembly were to be a bit off alignment (e.g. on a long belt drive), the coupling would be able to absorb that. But admittedly, this was more of a brain fart than a well thought out idea ;-). Thinking more, I think even that even there was some misalignment in the current setup, the slack in the belt may be enough to absorb that: the tilting would force the motor to be aligned if it wasn't.