Smoothing pneumatic cylinder operations

[Omikron]

Hello there! I ran into a small problem recently. I'm trying to synchronize  a couple of pneumatic cylinders which are not connected directly to the same frame and they keep extend and retract at different speeds.

The hoses are exactly the same length and the cylinders are from the same batch - they are not old+new. But even if I get the synchronization correctly then one of the cylinders does a harsh or abrupt action.

For example when the pressure in the system is rising one cylinder starts to operate right away and the second cylinder is "stuck" until the pressure is high enough and it extends abruptly. I've spent couple of hours manually extending and retracting cylinders in order to make them work smoothly without success. Did you have such problem and if so - what was your solution?

[Maaboo the Witch]

Could be that the two cylinders were made slightly differently. If you have two PF motors working in tandem, one acts faster than the other. Alternately, the two parts of this assembly you mention could have differing amounts of friction, so the system needs a certain amount of air to overcome it.

Edited May 2, 2018 by Maaboo35

[Omikron]

So you suggest to get more cylinders and match them according to their performance? This is the most simple but the most expensive way but I get it.

Any alternative suggestions are always welcome :)

[Captainowie]

I don't think you can expect even identical cylinders to be synchronised unless you connect them together mechanically. Even allowing that they're identical in every respect, _one_ of them is going to start moving first. The air pressure in that cylinder only has to overcome dynamic friction within the cylinder (in addition to whatever load is on the cylinder itself) while the same pressure would have to overcome static friction (which is always greater) in the other cylinder. The result is that the cylinder that starts moving first is going to keep moving.

You can, however, use multiple switches to pneumatically synchronise the cylinders. There was a good resource about this on the Internet many years ago, but all I can find now is a page on the Wayback Machine, which has the text but no pictures. http://web.archive.org/web/20100124210053/http://www.kclague.net/Sequencer/index.htm

If anyone knows of a better location of this information, I'd be very interested.

Owen.

[MAB]

Some of it is in the book "Classic Lego Mindstorms Projects and Software Tools: Award-Winning Designs from Master Builders".

You can see a limited preview of it at google books:

https://books.google.co.uk/books?id=_l7rUCDmMiMC&pg=PA280&lpg=PA280#v=onepage&q&f=false

 

[TeamThrifty]

problem is, air (or hydraulic fluid) follows the path of least resistance, so even a small variance is resistance means they won't synchronise.. However, if you have more pressure/flow than each cylinder can handle, then they'll both move at 100% so will be virtually synchronised.

So if you restricted the flow to each to be less than 50% or your supply... then the path of least resistance is equal on both side. In real life, big pipe feeding small pipes.

 

...but i doubt they'd be sync'd even then. Just better than they are.

 

[allanp]

In real life hydraulics this is solved by either mechanically connecting them (so they are either connected to the same frame or the separate frames connected with a linkage so they move together) or via a kind of check valve (with one input and two outputs) which only allows oil to exit one output if there is oil exiting the other, keeping flow out of both outputs equal. Being as Lego is pneumatic, the pressure is equalised automatically so you are only left with the option of mechanically linking them. Having said that, have you checked to see if the cylinders are still lubricated? You could try some good silicone based lubricant as it will make them work smoother, with less friction/sticktion and prolong the life of the cylinders.

[Omikron]

My initial thought was to use some-kind of buffer to equalize the pressure before it hits the cylinders like this white or blue air tank. In big things like excavators in real life they use tiny pressure sensors to get readings and to equalize the pressure from the buffer cylinder which is either built in the whole assembly or is added in the chain before the actual cylinder. Electronics provide feedback to the system so it operates smoothly. There is nothing like this in Lego which has it's own benefits like pricing.

I will try the idea with tank a little bit later since the mechanical link is not an option. Also may try getting more cylinders.

By the way - what's your impressions on the pneumatic 2.0 manual blue pump. In my experience it outputs less air volume with less pressure compared to older yellow one. But it has somewhat better mounting options.

UPD. I highly overestimated availability and underestimated pricing on the air tank part =(

Edited May 2, 2018 by Omikron

[allanp]

The new blue manual pump doesn't seem to have quite the same performance as the older one but like you say the mounting options are better. The newer motor pumps seem good as ever though. 

[Lipko]

Pneumatics will have a sticking effect after some standing time, especially if they are not high quality (and Lego pneumatics is not the best quality). No matter how do you run them in. We call this "Monday Morning effect" in the pneumatics company I work at. In case of Lego cylinders (plastic and rubber (?) in contact at the piston), this strong firs-operation-sticking effect can arise even after half an hour or so in steady state. So I don't think you can do much about this.

[1963maniac]

What about friction in the mechanism itself. Make sure everything is able to move freely without pneumatics attached.

[mocbuild101]

Other than mechanically connecting them, my only thought would be to split the system in half - have a separate mini-pump for each cylinder, and synchronize those.

[Omikron]

Found the culprit of the madness. One of the cylinders is from Unimog set (an older version of the cylinder) and it was causing inconsistency with other cylinders. It has about 2 seconds delay until it fires. Lubed it with silicone  lubricant and it didn't help. Maybe there is a bit of junk inside but from my tactile feedback it feels like there is a notch inside the cylinder which causes problems. I will use it in less critical application or somewhere where the abrupt action is a benefit.

 

Thanks everybody for the help! Much appreciated!

[Maaboo the Witch]

No problem!