Air compressor

[CP5670]

I recently bought a Lego pressure gauge off Bricklink and wanted to do something with it, so I built this last night:

This is an updated version of my old compressor, which I kept around my Lego collection and ran most pneumatic models from. It's built around the much more powerful 5292 RC motor and pumps faster and longer than the old one did. The two pumps cover their widest possible strokes and alternate through the rotation to reduce stress on the motor.

Here is the manometer itself. It has no studs and only Technic hole connections, but is a lot of fun to play around with. However, at least on mine, the actual gauge is not aligned perfectly with the exterior casing. The white background paper is bent a little and the metal nozzle on the bottom sticks out at a slight angle. It may be possible to disassemble it and fix this, but it works fine in any case.

Interestingly, I wasn't able to get the gauge to reach the full 60 psi, even if I just used a simple hand pump with it. The tubing always pops off around 45-50 psi (it's designed to do this according to the Technic manuals). The compressor will also reach about 50 psi before the tubing comes off somewhere, typically at one of the airtanks.

Edited August 31, 2010 by CP5670

[Blakbird]

The wedge belt wheels you are using for cranks appear in the photographs to be chromed. How did you manage this? Or am I just seeing it wrong?

[KimT]

They're trans-clear.

These are the ones you're talking about?

[CP5670]

Yes, they are trans-clear. I have had two of those lying around for years and they matched up with the manometer's color, so I used them here.

[Front]

wow, I did not even know that there was an air pressure gauge, I got so much to learn after returning to the bricks a couple of years ago. One of my colleages at work says you need six tanks in serial connection, to do serious pneumatics

Front

[Blakbird]

wow, I did not even know that there was an air pressure gauge, I got so much to learn after returning to the bricks a couple of years ago. One of my colleages at work says you need six tanks in serial connection, to do serious pneumatics

Hmm, what is this colleague's area of expertise? From my own experience, I would tend not to agree. I've recently built Jennifer Clark's JCB JS220 excavator. In this model, the compressor is in the remote controller. The compressor is powered by 2-9V motors (geared, lightweight type) and uses 4 pneumatic compressor cylinders at full stroke. This should be plenty of pressure and flow to do any LEGO job quickly. However, I am using about an 8 foot umbilical as well as quite a few splice connections using rigid tubing. I find that the head loss in the tubing is a bigger factor than the actual air supply. This does not make a significant difference in total delivered pressure, but it does make a difference in the rate at which it can be transferred.

Therefore, I think that if 6 air tanks were placed in series, you'd still have a problem with losses in the tiny lines in-between. I'd suggest a solution more like a RAID disc system and put the 6 tanks in parallel. Same storage capacity, but more capability to deliver the pressure.

Of course, you still need to get the air to your model somehow....

[Front]

Hmm, what is this colleague's area of expertise? From my own experience, I would tend not to agree. I've recently built Jennifer Clark's JCB JS220 excavator. In this model, the compressor is in the remote controller. The compressor is powered by 2-9V motors (geared, lightweight type) and uses 4 pneumatic compressor cylinders at full stroke. This should be plenty of pressure and flow to do any LEGO job quickly. However, I am using about an 8 foot umbilical as well as quite a few splice connections using rigid tubing. I find that the head loss in the tubing is a bigger factor than the actual air supply. This does not make a significant difference in total delivered pressure, but it does make a difference in the rate at which it can be transferred.

Therefore, I think that if 6 air tanks were placed in series, you'd still have a problem with losses in the tiny lines in-between. I'd suggest a solution more like a RAID disc system and put the 6 tanks in parallel. Same storage capacity, but more capability to deliver the pressure.

Of course, you still need to get the air to your model somehow....

I don't know his level of expertise in building technic, but he does CAE.

I actually thought about serial vs. parallel connection of tanks before joining this thread, and don't really see a difference in the amount of loss that will be had in the tubings.

Yes, a serial connection would be a problem for a high flow rate system, but I can't imagine a build where the amount of flow would favor a parallel over a serial connection.

But then again, I have no pratical experience in moc'ing with phneumatics.

Front

[furious]

a parallel system would deliver the air more smoothly. In a serial system the air pressure behind the outlet will be greater than a parallel system. There is more resistance in a parallel system that in theory should deliver a smother air supply. It would not have the same power or pressure.

[CP5670]

wow, I did not even know that there was an air pressure gauge, I got so much to learn after returning to the bricks a couple of years ago.

It came out in a fairly obscure Lego Education set last year. The set costs $55 and is not worth buying just for the gauge, but it includes a bunch of other pneumatic stuff. I got it from Bricklink for $19.

As for how many air tanks you need, it really depends on how you're operating the pneumatic switches and how fast air pressure is generated. The "classical" way to use pneumatics (on sets without airtanks) is to first set a switch and then use the pump to move a piston. If you have an air tank in the setup, you do it the other way instead, first pumping several times to generate pressure in the tank and then flipping the switches to move the pistons a couple of times, until the pressure runs out.

If you're doing it the first way or if your compressor is weak (like the one in 8868), more than one tank can be a bad thing since it will just take longer for pressure to build up. Multiple tanks are good for a fast and powerful compressor though, since you can keep the compressor running while playing with the switches and don't have to worry about the tanks filling up too fast. Even in that case though, I think two or three tanks should easily be enough.

As for how to arrange them, putting them in parallel makes more sense, but I don't think it would make much of a difference in practice unless you have 4+ tanks. 8462, which was the only set to use multiple tanks, had them in series.

[Parax]

Mine came with this...

a parallel system would deliver the air more smoothly. In a serial system the air pressure behind the outlet will be greater than a parallel system. There is more resistance in a parallel system that in theory should deliver a smother air supply. It would not have the same power or pressure.

A parallel system has more cross sectional area hence delivers greater volume at greater speed and is therefore less smooth (delivery surges hit all tanks, passes 1) . a series system has very small crosssection and delivers the same initial pressure but slower so has greater pressure drop off, but is more consistant and smoother (delivery surges are smoothed through each tank, passes multiple). both systems have the same initial pressure. Dont forget the bottleneck is in the single switch! so the pressure storage is kind of irrelevant... I haven't noticed any significant difference. Maybe ported switches and cylinders would??

P.

[SlyOwl]

If you're looking to keep the tubes on the nozzles at higher pressures, you can put one of these over the tubing on the nozzle, to tighten the connection.

It works best with the older, thicker tubing, but I'm not sure if it makes a significant impact at higher pressures.

I did a simple experiment by connecting a hand pump to a T-piece with tubing, with both joints secured using the wrench. Then I connected the two remaining nozzles of the T-piece to each other with a bit of tubing, but with only one of the joints with a wrench on it. 4 out of 5 times it was the wrench-free joint that burst off.

[CP5670]

That might be worth trying out. Thanks for the tip. Although the 45psi is probably more than enough for most situations, as the pistons already move very quickly at that pressure.

By "older, thicker tubing" do you mean the harder tubing from the 1980s? I don't use that stuff at all and have replaced it with newer tubing on any models using it, since it damages the nozzles over time.

[SlyOwl]

By "older, thicker tubing" do you mean the harder tubing from the 1980s? I don't use that stuff at all and have replaced it with newer tubing on any models using it, since it damages the nozzles over time.

Yup. I didn't realise it damaged the nozzles, though...

[CP5670]

It seems to grip the nozzle so firmly that it eats away at the plastic when you remove it after a long time. I've had a few bad experiences with it, ranging from slight damage and discoloring to the otherwise perfect all-yellow pistons on my 8854, to actually having the nozzles torn off altogether on some pistons many years ago. TLG replaced those for free though.

[Brickthus]

I wouldn't recommend operation at more than 25psi for long periods. The LEGO pneumatic components are not built for high pressures. You might get an instantaneous pressure peak of 40psi but that's a transient. If your model requires steady pressure above 25psi then it needs more cylinders in parallel.

I use a car tyre air compressor that has its own pressure limiter, which I set to 25psi. That's plenty for my pick and place robot.

Theoretically, a meter should read 2/3 scale when all systems are working normally. That would be 40psi. I suggest that the manometer uses a standard off-the-shelf meter in a LEGO box, so this principle may not follow.

Another thing is to drill out the nozzles of the cylinders, especially the 48mm ones. If you do it to the top nozzle, be very careful not to damage the top seal of the cylinder - air will leak in easily if you do. Drilling out the nozzles will let air get into the cylinders quicker, for burst performance. I have found that to be a limit of the system. That would have a greater effect than the single valve or the arrangement of air tanks. Check the performance by contracting the cylinder fast. If it springs back then not all the air has escaped so there's a flow limit in the bottom nozzle. Check top seal performance by contracting the cylinder whilst covering the top nozzle. If it doesn't spring back then air is leaking in.

The air tank is just an accumulator. Its two end nozzles are connected together so the only issue for air mass flow rate is the diameter of that solid tube. The diameter is the same as that of a T-junction, valve or cylinder nozzle. You need to parallel-up the whole system if you want really fast air flow, but load cylinder nozzle performance would have to be tip-top for it to make any difference.

It will also make more difference to use as much 3.2mm solid tube as possible, keeping the flexible hose for just the corners, minimising the balloon effect.

Mark

[CP5670]

The LEGO pneumatic components are not built for high pressures. You might get an instantaneous pressure peak of 40psi but that's a transient. If your model requires steady pressure above 25psi then it needs more cylinders in parallel.

That was my impression too. In any case, it would have to be a fairly heavy load that could actually benefit from such high pressures. I hooked up this compressor to various official Technic models and the pistons on all of them move very fast beyond 30 psi (making movements jerky), even the ones on the 8421 crane boom or 8868 turntable.

It's interesting that the nozzles are actually a significant bottleneck in practice. I probably wouldn't mess with mine though for fear of damaging them.

[darksheep]

It's interesting that the nozzles are actually a significant bottleneck in practice. I probably wouldn't mess with mine though for fear of damaging them.

It is very true the guys at http://www.nicjasno.com drill the nozzles to make them wider when they make their lego engines to get better performance out of the cylinders