Recommended Posts

My TC-10 entry will be a Useless Box or Useless Machine.

The sole purpose of such a contraption is to expose an ON/OFF switch, and whenever it is turned on, it mechanically pushes the switch to turn itself back off.

There are various LEGO incantations, most notably an

, but no pneumatic versions I could find.

I'm thinking a simple mechanical design with two pneumatic cylinders and a switch. The bottom cylinder slides a compartment up through the roof of the box. The compartment contains the second cylinder, which can then push the switch back to its OFF position. Something like this:

draft-idea.jpg

The mechanics are very simple but not so with controlling the cylinders. I have yet to come up with a working design, but here the things I've worked on so far.

Basically the controller is a state machine. With a switch and two cylinders that makes 8 possible states, where 4-6 of them are relevant, depending of the implementation (like whether the switch is a pneumatic value or some mechanical input)

Functionally, the states are:

1. Switch ON

2. Bottom cylinder extracts

3. Top cylinder extracts

4. Switch OFF

5. Top cylinder retracts

6. Bottom cylinder retracts

States 3 and 4 may coincide as the cylinder pushes the switch and 1 and 6 may also coincide.

My next few posts will discuss the designs I've tried (and failed) so far. Seems this project will be a risky - and learning one :sweet:

Share this post


Link to post
Share on other sites

That was my idea, I just didn't reserved it, which is against rules.

Well, I didn't just reserve the idea, I started to discuss the design which for this kind of thing certainly is "showing progress" if that's what you mean. By all means, go ahead and make your version as well.

You might have the bottom cylinder activate a second internal valve that powers the top cylinder.

Yes, the "controller" will definitely contain a number of valves, and possibly internal cylinders. The question I'm getting at, is whether to build a generic state machine or a minimal state machine optimized for the problem at hand. This requires a lot of tinkering to answer, it seems.

Share this post


Link to post
Share on other sites

Well, I didn't just reserve the idea, I started to discuss the design which for this kind of thing certainly is "showing progress" if that's what you mean. By all means, go ahead and make your version as well.

Nice progress :)

Share this post


Link to post
Share on other sites

Here's my mockup pump and regulator used for testing.

pump-draft.jpgregulator-draft.jpg

That was my idea, I just didn't reserved it, which is against rules.

Nice progress :)

Hmm, I just read the rules again, and it does seem that regarding "photos of build in progress" i misread "e.g." instead of "i.e." So Jurrs may be right. I'm gonna ask Jim.

Share this post


Link to post
Share on other sites

It's indeed against the rules to start a topic without progress, but since you just posted the mockups, it's okay for me.

Jurss, you can build your own version. There's no rule against same type of entries. Or choose to build a totally different model. The choice is yours :wink:

Share this post


Link to post
Share on other sites

It will be interesting to see if this kind of machine works in reality, I would imagine as long as the air pressure it high enough, it would be ok, but it could all be about the speed of the application when that switch is exposed. I look forward to this one.

Share this post


Link to post
Share on other sites

Wow, I thought I had a unique idea. Interesting to see 3 different implementations.

Great! To me it's all about exploring the design space so the more the merrier :classic:

It will be interesting to see if this kind of machine works in reality, I would imagine as long as the air pressure it high enough, it would be ok, but it could all be about the speed of the application when that switch is exposed. I look forward to this one.

Yeah that's definitely a concern. If it's too slow it won't be fun to watch

Share this post


Link to post
Share on other sites

If it is slow, you can make it look like it is intentional slow, like a skeleton finger reaching out of the ground to push the switch

Share this post


Link to post
Share on other sites

OK, I deacided then also to make this thing anyway. Will be small contest inside TC10. Of course, there will be more subcontests - loaders, forklifts, excavators etc. Will try to start this evenenig, because rain is promised.

So, good luck!

Share this post


Link to post
Share on other sites

After some experiments using a feedback loop, I've come up with this that kinda works. There are some timing problems that can make things go out of sequence but with just the right amount of rubber bands it seems to be working consistently. Though I had more "technical" ideas in mind with levers and mechanical switches, I think this is the way to go for me.

feedback-draft.jpg

OK, I deacided then also to make this thing anyway. Will be small contest inside TC10. Of course, there will be more subcontests - loaders, forklifts, excavators etc. Will try to start this evenenig, because rain is promised.

So, good luck!

Good to hear and sorry for my mistake. Thanks - good luck to you too.

Share this post


Link to post
Share on other sites

An overdue update. While my previous feed-back loop idea had promise, it was not consistent enough. This was due to one state change that had to flip two valves as well as a main cylinder. Just too many ways that can fail. So I went back to Mark Bellis' design (1) and tried to extend that instead. I have found it really hard to reason about complex pneumatic designs, but looking over the "truth table" of (1) over and over, I came up with a solution that I'm really happy with, just by adding three secondary valves to it. This means that control pressure is not mixed with "work pressure" so state transitions are always consistent, as in (1). So the only thing that can go wrong is a cylinder not extending/retracting far enough in its "time slot".

I made a small video that demonstrates it with a mock-up box. The box needs a lot of mechanical work - i had to help it as you can see. But the pneumatics work great now :classic:

I'll present the pneumatic design in detail in a later post.

Share this post


Link to post
Share on other sites

Interesting! Is there a perspective of it working faster? Because I think the big part of the charm of these useless boxes is in the speed of the counter action. Delay between swithing and reaction is good, but the action itself should be fast :)

Share this post


Link to post
Share on other sites

Interesting! Is there a perspective of it working faster? Because I think the big part of the charm of these useless boxes is in the speed of the counter action. Delay between swithing and reaction is good, but the action itself should be fast :)

Yes, there are a couple things. In the video I'm using a hand pump and the PF pump provides a little more air so that will be slightly little faster. Then there's my latest regulator that provides more pressure but I'm not sure how far that will last using a single air tank before the pump kicks in. Finally I might try a 1x11 cylinder for the switch. This should be faster but I'm not sure whether its up for the job. Oh, and that would in fact free up another medium cylinder so I could try using two of those for the feed-back loop, which would also make the loop faster. So many things to try...

Share this post


Link to post
Share on other sites

You might try to extend the lever on the switch - if you make it about 3L, thin cylinder would have easier job on shifting it :)

Share this post


Link to post
Share on other sites

You might try to extend the lever on the switch - if you make it about 3L, thin cylinder would have easier job on shifting it :)

Yes, good suggestion. The 1x11 has a larger extension so that's certainly possible. I also wonder whether it could shift two valves in one go as that might enable a certain "stretch goal" feature ;-)

Share this post


Link to post
Share on other sites

Everything is working now and I'm getting really pleased with it. I just need to steal some gray panels off of my 42043 to complete the box. Here it is in action:

There are a few things I'd like to improve though:

  1. The first two times the speed is ok and I guess I could live with that. But the third time I leave it off long enough for the regulator to turn off the motor. And then when the machine is turned on, it quickly runs out of pressure and is really slow until the motor starts again. There are a few options but they all take up more space and I'll need to go shopping for more pumps and/or air tanks.
  2. The battery box doesn't currently fit inside the box. If I'm really lucky there may be space for a small battery box but I don't think so. Any ideas besides making a bigger box?

Share this post


Link to post
Share on other sites

It's time for another update. I managed to just squeeze a small battery box in there so was able to complete my entry :classic:

A note of warning to future sequence builders: Be careful when lubricating cylinders. While the dynamic friction was greatly reduced, increased static friction caused all kinds of problems in a sequencer. I ended up using non-lubricated cylinders, except for the 1x11.

Here's a couple of shots of the final box:

dsc_2448.jpg

dsc_2462.jpg

dsc_2474.jpg

dsc_2485.jpg

A manual pump option is hidden behind a door.

 

And a few inner details:

dsc_2415.jpg

dsc_2508.jpg

dsc_2551.jpg

Share this post


Link to post
Share on other sites

Very complex build. How much cylinders do You hav eat the end.

I have working concept with one cylinder now, but i'm afraid, that i will not manage to make it nice looking etc.

Share this post


Link to post
Share on other sites

It looks very nice!

I'm curious to know how you implemented the sequence "up, push, retract, down", i.e. "extendA, extendB, retractB,retractA". The normal sequence (as in 8868 B-model) is "extendA, extendB, retractA, retractB". I found out that you can't just swap some hoses to reverse an order, since you end up having air pushing into both ports of a cylinder (if that description makes any sense).

Share this post


Link to post
Share on other sites
2 hours ago, Ludo Visser said:

It looks very nice!

I'm curious to know how you implemented the sequence "up, push, retract, down", i.e. "extendA, extendB, retractB,retractA". The normal sequence (as in 8868 B-model) is "extendA, extendB, retractA, retractB". I found out that you can't just swap some hoses to reverse an order, since you end up having air pushing into both ports of a cylinder (if that description makes any sense).

Thanks!

The question makes perfect sense. As I'm working with the video, I've been wondering how to explain/show this in a simple way. I haven't found such a way just yet so here goes ...

As you notice, the problem with "just swapping hoses" is that you likely get pressure on both ports of a cylinder. The simplest way to avoid that is to have a dedicated control valve X for the up/down cylinder and another valve Y for the push/retract cylinder. This turns out to be a good start.

The sequencer itself has two cylinders and two valves V1 and V2. The sequence then has the following steps or "truth table":

Step 1:  V1 off  V2 off

Step 2:  V1 on  V2 off

Step 3:  V1 on V2 on

Step 4:  V1 off V2 on

What we want to happen is:

Step 1: Up

Step 2: Push

Step 3: Retract

Step 4: Down

 

The thing to notice is that Up/Down should only change when V1 is off and Push/Retract should only change when V1 is on.

Also, Up/Down only changes when V2 changes and Push/Retract only changes when V2 changes (but in opposite positions of V1).

So, by adding an extra control valve Z that is controlled by the sequencer to have the same position as V1 we're almost there. The next thing is to attach X and Y to be controlled by the sequencer to have the same position as V2 and using the outputs of Z as inputs of X and Y, we have all the motions covered.

The only thing remaining is to set up the main on/off switch to block the air flow to one port of V1 so that it can only move from step 4 back to step 1 when the switch is on.

As I said, I hope I can come up with a visual way to show this :grin:

4 hours ago, Jurss said:

How much cylinders do You hav eat the end.

Sequencer and control: 2 x 2x11 cylinders and 5 valves

Lift: 1 x 2x6 cylinder

Push/pull: 1 x 1x11 cylinder

Regulator: 1 x 2x6 cylinder

Motor pump: 2 x 1x6 small pump and 1 PF L motor

Main ON/OFF switch: 1 valve and 1 light sabre

Manual pump input: 1 valve

 

So, a total of 7 valves, 5 cylinders, 2 pumps and 1 motor. And I'm not going to count the hoses :wacko:

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.