I recently came across your random positioning machine (RPM) design and found it particularly interesting due to its simplicity. I aim to scale up this project to accommodate at least four 12x12 cm square Petri dishes. The RPM would be used to study the metabolic responses of plant roots to simulated microgravity, with a particular focus on root exudates. Do you think that would be feasible?
I am relatively new to electronics, 3D printing, and the technical aspects of designing such a device. I have already built a few project, including the smaller version of the RPM and have access to all the necessary equipment. However, I have less experience in development and design, so I wanted to reach out to see if anyone might be interested in offering guidance or collaborating in some capacity. Any form of assistance—whether in the form of general insights, design suggestions, or technical advice—would be greatly appreciated.
Sounds like fun! The small-scale RPM works fairly well for this sort of thing, and we’ve seen it pop up fairly regularly from users performing similar experiments. Scaling it up is an interesting challenge to tackle.
I am sure you will find no shortage of eager makers and very helpful advice here! I would recommend going for the “collaborative” approach rather than “guidance”, as I feel you’ll come out of it feeling much more confident in your abilities.
. Now to start some discussion about the project:
The first issue regarding the size that comes to mind is that 3D printing will be less applicable, assuming you have a hobbyist 3D printer. Most smaller 3D printers would likely struggle to produce usable pieces at that size. My thoughts are that it may be relatively easy to whip together the frame out of structural pine or steel/aluminium from a hardware store, and only 3D print the smaller stuff, like gears.
The other thing to consider is that your servos will need to be considerably higher rating than those used in the original, as the scaled up variant will bring with it a higher torque requirement for the servos. More torque also requires higher current, so you will likely need a more sophisticated power delivery system. I would aim to overspec the servos by quite a bit, as the lifetime of the servos will increase if they aren’t working as hard.
What are your thoughts on the shape/attachment of the petri dishes? I would think you could reduce the size of the machine by stacking the dishes on top of one another, see image below.
(P.S. Ignore the circle, I hadn’t realised you said “square” petri dishes, and was imagining the classic circular ones)
This would compact the usable space, seeing as the machine is already quite square. Will this type of thing work with the dishes you plan to use? It would also be good to brainstorm how you will secure the petri dishes. Images of the dishes would be useful, if you have access to them.
Looking forward to your response! Scaling up a concept is always a fun challenge.
Thanks for the quick answer. I have access to a few 3d Printers, so anything except the outer frame should be fine for 3d printing. (The outer frame of the original design should be similar in size to the inner frame of this one). My initial plan is to print the outer frame in 4 pieces, I’ll see if it’s solid enough, but of course I could also make it out of aluminium.
For the petri dishes, I was planning to stack them, similar to your picture. Perhaps a platform at the center with 2 (or 3) petris on each side secured together would be a simple design.
Here is a picture of the petris I will use:
(They are 12 x 12 x 1,5 cm) and only the bottom half is filled with agar media
My biggest issue for now is to find a servo that will fit on the rpm, but is also strong (and reliable) enough to last roughly 10 days without stopping.
Thanks for the recommendation! With this motor upgrade, do you think the original design (cables, breadboard) could cause some issues? Can they handle the higher voltage/amps?
Also, I don’t think there will be any problems with the original code for the Pico working with the larger servos, but I still wanted to check—do you see any potential compatibility issues?
The Piicodev driver can deliver up to 5V via the USB-C ports, which would be on the lower end for that servo but should still work. It would be limited to a max current of 3A , which is also the max current draw of those servos while stalled, unless it gets stuck or something prevents it from moving it should be fine. We do also have FS5109R - Standard Continuous Rotation Servo (10kg/cm) | Buy in Australia | CE09388 | Core Electronics which has a lower max stall current of 1.7A, but a lower torque rating.
With these servos as well, they are physically larger than the ones intended for the project but the STL files should be able to be modified to suit.
I’m experiencing issues with the servos overheating and shutting down after running for a few hours. From what I understand, servos are generally not designed for continuous operation. It seems the motors are stopping even when they’re not connected to the machine (I’m trying to replicate the original design), which makes me think the problem could be related to the extended 24/7 use.
Have you encountered this issue in the initial design? If so, are there any solutions or workarounds to prevent the servos from overheating or to improve their endurance under continuous operation? I posted about this a few months back, but I never really succeeded in solving the issue.
Additionally, is there any other type of motor that I could use without having to redo the code?
One way to combat this would be to lower the speed that the servos travel at, in the code it would be a matter of changing the speed_inner and speed_outer from the default 0.5 to a lower value.
Is there much friction inside each joint? The increased load will make the motors inside the servo run hotter and burn them out faster.
I’d also consider swapping out the servo’s for some low RPM DC motors - it would require mods to the 3D prints and light changes to the code but should make for a much more robust system.
I’ve tried to loser speed_inner/outer, but it doesn’t seem to change the outcome. Actually, the servo fail when they are not even connected to the machine, so I don’t think friction or load could be the issue. I will try with the bigger servos, to see if the problem solves itself . Thank you for the DC motor recommandation, I’ll look into it.
