Might be interesting to 3D print, or maybe buy, some rack-and-pinion gears to make a true linear output servo. With the right-shaped (triangular?) gear teeth and some side spring preload to keep them in mesh, backlash could be eliminated.
Also interesting to try Hall-effect feedback, for both the linear and rotary output versions. I've got some discreet Hall-effect sensors, that look like small transistors, and give an analogue output according to the position of a small magnet nearby: I've not tested them to check their range/linearity/sensitivity to nearby motors and electronics yet though.
The ProPo project, a DIY proportional servo
-
- Posts: 747
- Joined: 16 Feb 2018, 14:11
- Location: Warwickshire
- PaulJ
- Posts: 604
- Joined: 16 Feb 2018, 19:01
- Location: Ipswich, UK
Re: The ProPo project, a DIY proportional servo
Tobe, I have opened the zip and they are STEP files which my slicer can't handle ........ Can you do them as .stp files?
Paul
Paul
- Mike_K
- Posts: 686
- Joined: 16 Feb 2018, 06:35
- Location: Hertfordshire
Re: The ProPo project, a DIY proportional servo
Easy peasy when you have a slow-moving servo! I've been testing an old Futaba S3001 with a servo amp using an ATMega328P with DRV8838, where have you seen that combination before?! I started with a method similar to yours and I had the most terrible overshoot and then hunting around neutral unless I had a very big dead-band, even when I was using the DRV8838 braking. I then looked at what a normal Futaba servo amp does and as the servo approaches it's intended location, the driving pulses get progressively shorter until they stop. Reproducing that methodology has it working as well as the original amp, even under load. Now it's time to port it to the ATtiny85 without the benefit of a 16-bit timer to measure the ppm pulse, so will have to revert back to micros(). Hopefully, it'll work OK with Tobe's servo just as well.
Cheers
Mike
-
- Posts: 747
- Joined: 16 Feb 2018, 14:11
- Location: Warwickshire
Re: The ProPo project, a DIY proportional servo
I suppose the PID-controller algorithm is the best (relatively simple) way to go if you're using a microprocessor for servo control - pretty easy to adjust the three constant terms by trial-and-error to give good control, or you can even implement a self-tune routine to calculate the terms automatically when trying out different motors, gearboxes, and operating voltages.
-
- Posts: 670
- Joined: 16 Feb 2018, 06:19
- Location: Varberg or Stockholm, Sweden
Re: The ProPo project, a DIY proportional servo
.stp files are not suitable for printing just for CAD work but they can be converted in a CAD program to .stl files which are files your slicer will work with. I have included here the proper files as a curiosity as I can only guaranty they will be right in my printer and with ABS which for this one I use a high end printer. See PM
To Phil:
One of my "easy peasy" are strain wave/harmonic drive gearboxes as I have use them in several Med Tech & URV's project in robotics and I have a few solutions which are unique and proprietary unfortunately but maybe one day...the key is that they will hold position w/o any power and works as good in booth directions. Some of the older Bonner Digimite RS utilized a strain wave gearbox and was one of the reasons Bonner closed down, patent litigation. They were used by the military contractors of the time in "sight system" as they are extremely precise and no play.
To Martin:
This project is based on simplicity and with just 2 moving parts and everything else from over the counter. ..and yes I have a linear one in the pipe also that one extremely simple. Here follows a PID-based servo amp
To Mike:
I found an old Futaba servo and running it's amp and it work fine
- Attachments
-
- Servorutiner.ino
- (1.8 KiB) Downloaded 193 times
-
- Ny mapp.zip
- (98.37 KiB) Downloaded 192 times
Cheers,
Tobe
Tobe
- Wayne_H
- Posts: 811
- Joined: 17 Feb 2018, 05:26
- Location: Temora, NSW. Australia
- Contact:
Re: The ProPo project, a DIY proportional servo
Tobe - I really like the 2 dual-servo configurations you've shared, something very retro but practical/serviceable about them
This is great discussion!
Using the softly softly approach, I was intending to start by using some amps ratted out of old servos. Longer term, I'd thought about an approach very similar to what Mike has done, but that will require a lot more research and much trial & error on my part, hopefully without losing too much blue smoke along the way
But, first I have to find a commercial 3d printing company in Wellington......
This is great discussion!
Using the softly softly approach, I was intending to start by using some amps ratted out of old servos. Longer term, I'd thought about an approach very similar to what Mike has done, but that will require a lot more research and much trial & error on my part, hopefully without losing too much blue smoke along the way
But, first I have to find a commercial 3d printing company in Wellington......
Cheers,
Wayne
Once a Retrobate, always a Retrobate............
Wayne
Once a Retrobate, always a Retrobate............
-
- Posts: 670
- Joined: 16 Feb 2018, 06:19
- Location: Varberg or Stockholm, Sweden
Re: The ProPo project, a DIY proportional servo
...or I put a few in a "Blue Envelope" with all hardware.
Cheers,
Tobe
Tobe
- Wayne_H
- Posts: 811
- Joined: 17 Feb 2018, 05:26
- Location: Temora, NSW. Australia
- Contact:
Re: The ProPo project, a DIY proportional servo
Tobe - PM sent
Cheers,
Wayne
Once a Retrobate, always a Retrobate............
Wayne
Once a Retrobate, always a Retrobate............
-
- Posts: 431
- Joined: 16 Feb 2018, 13:57
- Location: South Devon, UK
-
- Posts: 670
- Joined: 16 Feb 2018, 06:19
- Location: Varberg or Stockholm, Sweden