Our team is using the versadrop octocanum drivetrain. There are four pancake pneumatic cylinders. I am thinking about using a double solenoid and an adapter to connect all 4 pistons to one solenoid because solenoids take up space. Is that going to work? Do I have to have one solenoid for each piston?
You can use two or more pistons on a singular solenoid.
This will work, but there are somethings to consider.
- The more pistons connected to once cylinder, the slower air will enter each piston (the limiting factor in this case being how fast air goes through the solenoid). Try it; it may work, or you may need to add a second solenoid if it’s not fast enough for you.
- You need to make sure all of your tubing is the same length. If the tubing to one side is shorter than the other, the air will get to those cylinders first and the robot will tilt up on that side. Something you don’t want to happen.
You shouldn’t run into any issues running like that. I would recommend using the 24V feature on the PCM, and utilizing 24VDC solenoids, as they seem to act much faster than the 12VDC solenoids. We use the 12VDC solenoids on our machine, but we’re not trying to shift 4 pancakes at once.
Also, I would recommend a pair of tanks for reserve. The 60PSI is more than enough pressure, but you’ll appreciate the extra volume of air that a pair of tanks will afford, provided you have room.
And like Ari423 said, if it doesn’t seem to be reacting the way you like, add a second solenoid to the manifold, and see if your performance improves. Those SMC solenoids (if that’s what you’re using) don’t add much weight and are super smooth.
Good luck!
Friendly reminder - only one solenoid may be connected to any given actuator. If you choose to use more than one valve do not connect them together down-stream. [R95]
Why might they care about a few hundred milliseconds of tilt?
In my experience, it can be a lot more than a few hundred milliseconds of tilt. In 2015, we used one solenoid to control two cylinders for the two sides of our tote elevator. The solenoid was on the same side of the robot as one cylinder and the opposite side from the other so the tubing to one cylinder was shorter than the other. When we tested our elevator, there was almost a full second (definitely more than 1/2 second) of difference in the start times of the cylinders. That much time running one side of the robot in “high gear” and the other in “low gear” will likely cause the robot to veer off-course. For such an easy problem to correct, there’s no point in making an already driver practice-heavy drivetrain even harder to drive.
Discussing your concerns in order:
-
Lightly-loaded large-bore air cylinders with long strokes (I’m assuming are what you used) are very sensitive to flow rate discrepancies, but I don’t think that your experience applies to OPs situation.
-
There is no shifting from high gear to low gear here… the versadrop octanum drive uses wheels of the same diameter and OP never mentioned shifting.
It uses 4x 32mmx15mm air cylinders which should fill quite rapidly, even if fed through different lengths of tubing. Even if they do fill at different rates the robot’s weight (significantly more than a stack of totes) will help ‘even out’ the deployment rate.
- Until strafing is engaged there should be almost no perceptible difference to the drive. After all, mecanum wheels will act like tank drive if controlled like tank drive.
Absolute worst-case: flow control valves can be employed to ensure uniform deployment of all air cylinders, and flow control valves are cheaper and smaller than adding more solenoid valves.
I assumed that the VersaDrop modules’ mecanum wheels were geared faster than the traction wheels like most custom articulating wheel designs. If they are geared for the same speed, it shouldn’t be a problem.
The reason it’s more than a few milliseconds of tilt is not because of a delay in the arrival of a pressure wave, it’s because the longer a tube is, the more drag there is for air moving inside it.
Not only the drag, but if your modules are not aligned correctly, the delay in receiving the proper pressure for all modules at the same time could put undue strain on the long-end module causing it to bind.
I am not an expert on the versadrops, but on our modules we are running fairly tight tolerances between our frame(our frame - or rails - house the modules) and the modules. Also, if there is a significant discrepancy (or choking of pressure), the module under undue strain could bind the apparatus. Or if the weight of your machine is not balanced and by chance or design the last module to gain pressure is the same that carries the most weight - you could see significant delay in functionality.
The cost of more solenoids is substantial… but I would err on the side of caution when it comes to pneumatics. In other words, in an octocanum system 4 modules = 4 cylinders = 4 solenoids = 4 equal length plumbing.