Replacement Robot Chassis and hardware for ArloBot

27 views
Skip to first unread message

thomasco...@gmail.com

unread,
Feb 16, 2022, 9:25:58 AM2/16/22
to ROS for Arlobot
Hi Chris,

Based on your posts in other threads it looks like you are planning on building a robot chassis using the Roboclaw Motor Controller in place of the Parallax DHB-10 Motor Controller.

I assume that you will start with a fresh slate chassis wise and use only readily available, off the shelf components?

If so, I plan to purchase the following hardware to be able to participate in your new chassis design:
Motor                    # High torque and includes high resolution quadrature encoder and                                           front ball bearing
Motor Mount        # 6 mm thick for strength
Wheel                     # 5.7 inches in diameter
Wheel Adapter      # Adapts wheel to 6 mm motor shaft
Chassis                  #Almost the same diameter as the Arlo chassis

Comments?

Regards,
TCIII

thomasco...@gmail.com

unread,
Feb 16, 2022, 10:43:38 AM2/16/22
to ROS for Arlobot
Hi Chris,

Since SLA batteries are so heavy versus energy storage, i wonder if it might be smart to look at Li-Ion batteries as a power source?

Regards,
TCIII

Christen Lofland

unread,
Feb 16, 2022, 11:18:02 AM2/16/22
to ROS for Arlobot
My goal for the next version is to include a list of parts and building instructions that anyone can use to build a robot with "off the shelf" parts and possibly 3D printed parts that they can print or order to have printed, but ALSO to make it such that it should work with any framework. The only requirement really should be:
1. A Raspberry Pi 4 - This is also open for people to experiment with, anything running Ubuntu 20.04 will work, but it is easier for me to target one platform for the "I tested it and it works" version.
2. A differential drive robot (not ackerman steering like a car, although in theory if ROS supports it, then a config change could accommodate it)
3. Use the Roboclaw motor controller, because there is a pretty tight link between the motor controller and whatever ROS code drives it.

After that, I feel like the rest is entirely a matter of config parameters in a file. Even the current code would work on any size or shape robot if it worked with Roboclaw instead of the Parallax Propeller boards.

I use the SLA batteries for three reasons:
1. They are cheap
2. They are very safe. It is hard to make one explode or even get very hot. Since they are gel, even puncturing one should result in very little acid release, while puncturing a LiPo battery can cause an amazing fire. (Remember those Samsung phones?)
3. They are EASY to charge, hard to damage, and don't catch fire when you do abuse them. Just slap a $10 charger on it and leave it forever. You should put a discharge disconnect on them, as it damages them to discharge them below 11.6 volts, but by "damage" I mean, shortens their lifespan, they won't catch fire like other chemistries can do.

Again, all that, and being very cheap. I have 8 of them now so I can swap them in and out of my other projects. It is very convenient.

They are heavy though, I grant you that. And I'm shy about amperage too, so I tend to limit my motors to 3 amps each. I noticed that while many motors have twice the torque of the Pololu motors, they are also only getting that at 9 amps! I have  10 am fuse on my robots, 10 amps for both motors and the Pi and everything else. Imagine dedicating 18 amps to just the motors! I would probably want to improve my wiring. Still, at 3 amps each, I don't have much trouble getting more speed than I need indoors. For outdoors, I could see it being nice to go faster, and lighter batteries would reduce the need to run high amperage to the motors.
That is all stuff that shouldn't really affect the code though.

That said, again, the battery shouldn't matter. The Roboclaw has a very wide voltage range available to it, so it is a matter of matching your motor to your voltage input.
There are some nice drop in replacement LiPo batteries for the little SLA batteries I use, but they are about 4x as expensive, and require a very expensive charger, and again, I don't want them piled up in my basement and/or left charging or discharging inside my house untended.
Probably my long term plan would be to build my own robots with the ability to swap out SLA for LiPo batteries at will, so I can keep the SLA batteries inside and have the LiPo in a safe place and only get them out for special occasions like operating outdoors or at a venue where I need long run time and lighter weight. That gets back to cost though. My experience with LiPo is that they don't have a long shelf life when ignored, like the SLA do.

Basically, my hope is to make it very easy to configure and just use the same code. Lashing myself to the Propeller hardware was something I did before I even had a 3D printer and before I knew anything about electronics. I regret now that it is so  tied together, when they are discontinuing their products.

I think the Roboclaw will be my only part to be tied to next, and they are a pretty reliable company just making a product rather than a platform. Also, once I am down to ONE motor controller, and really even supporting multiple motor controllers shouldn't be terribly hard if we want to once I isolate those parts of the system to use the Roboclaw.

In short, yes, as long as you are using the Roboclaw motor controller, and probably a Raspberry Pi 4, your plan sounds good. I personally plan to have at least two entirely different robots running on it, and hope to build more. My only impetus to even make a "standard chassis" is that I know many people find my code and really want to "just build it" but have no idea how to just start designing a robot. If I can suggest a 3D printable setup using easy to order parts that are usually available along with cheap pieces available at hardware stores, then people can get things going easily without relying on an expensive discontinued platform from a company. It also makes it easier to present a video to people of a working system that they can actually replicate.

To be clear, I mean 3D printed parts combined with other materials for structural integrity, but those parts will be things any hardware store has, not just an entire 3D printed chassis.

Ultimately though, the chassis I design is intended to demonstrate that you can make pretty much any block of matter into a robot with a Raspberry Pi, a Roboclaw motor controller, and your choice of motors and batteries.

Sorry, I wrote a lot of words. Writing words is easier than actually doing stuff. :)

How does that sound?

thomasco...@gmail.com

unread,
Feb 16, 2022, 12:28:14 PM2/16/22
to ROS for Arlobot
Hi Chris,

Sounds like a plan to me.

You are right about the SLA batteries as far as charging and safety are concerned, I just wish that they weren't so heavy as they eat up a lot of the robot payload capacity.

Unfortunately RoboClaw Motor Controllers are out of stock everywhere until about June of this year.

Also Rpi 4B/4GBs are out of stock everywhere unless you want to pay exorbitant prices. 

Regards,
TCIII

thomasco...@gmail.com

unread,
Feb 16, 2022, 1:14:47 PM2/16/22
to ROS for Arlobot
Hi Chris,

I just checked BGMicro, the manufacturer of the RoboClaw Motor Driver Modules, and they are saying that they are in stock for $99 plus shipping for the 2 X15 Amp RoboClaw Motor Driver Module.

This is very strange as the RobotShop, Pololu, ServoCity and a number of other resellers are claiming that they are on backorder.

Regards,
TCIII

danial dunson

unread,
Feb 16, 2022, 3:09:02 PM2/16/22
to ROS for Arlobot
Hey cris, 

I race offroad motorcycles, and very good battery Li-ion packs are available on the market today. I have used these Li-ion Packs with no issues and a standard 12V battery charger that has the LI-ion option. zero explosions and much more reliable. I have found firepower to be the best value for performance. around $100-$200. 

Christen Lofland

unread,
Feb 16, 2022, 3:13:51 PM2/16/22
to ROS for Arlobot
I had a Roboclaw controller on back order with Pololu for months. Then I went and ordered 2 or 3 of them direct from Roboclaw's website and got them within a week.

The supply chain is weird right now is all I can say. I found the price direct from Roboclaw to be within a few dollars of Pololu, so I plan to just order them direct from now on.

Also, the 2x7 should be fine for most uses cases: https://www.basicmicro.com/Roboclaw-2x7A-Motor-Controller_p_55.html
It is a little cheaper and smaller/lighter. I believe the 7A model can go to higher peaks, and since the Roboclaw can limit the amps, you can guarantee it won't be overloaded. As i said, I have mine limited to 3 amps per motor and they move a lot of weight.

Yeah, the page says:
  • 7.5 Amps Continuous Per Channel
  • 15 Amps Peak Per Channel

That is a LOT of amps!

To be fair, I do have one of the 2x15 controllers, because I bought that first as my "test" so I could experiment, and if you only plan to buy one, it makes sense to be free to try everything.
If you do end up wanting to build multiple robots though, you can save a bit with the 2x7 controllers and they are a lot smaller too.

Christen Lofland

unread,
Feb 16, 2022, 3:18:18 PM2/16/22
to ROS for Arlobot
Thank you for the batter recommendation! It does prove my point, at 5 to 10 times the price of an SLA battery! On the other hand, adding a $100 battery to a working robot doesn't sound so bad.

I also do love the idea of have a good reliable steady source of other battery options to recommend, so I will add this to my list.

Christen Lofland

unread,
Feb 16, 2022, 3:20:35 PM2/16/22
to ROS for Arlobot
Yeah, the chip shortage is eating our lunch on the Raspberry Pi front. Basically my recommendation is if you need another one, order it at the Canakit web site. They have some that they will put on back order. I'm not sure if they have the 4GB models though. Maybe only the 2GB? Looks like March 31 is the ship date.

It will clear up eventually. In the long run, these things will sort out and get back to normal. Probably faster than I can design, print, build, code in my hobby time.

Christen Lofland

unread,
Feb 16, 2022, 3:21:38 PM2/16/22
to ROS for Arlobot
One nice thing about the heavy batteries, is that if you build a taller robot, they keep the center of mass really low, not to mention giving you plenty of traction! :)

thomasco...@gmail.com

unread,
Feb 20, 2022, 1:01:53 PM2/20/22
to ROS for Arlobot
Hi Chris,

Looks like Clearpath has decided to go back to the iRobot sized chassis.

This is why I think that a robot chassis of around 14 inches in diameter is the way to go compared to the 18 inch diameter of the Arlo.

Keep the weight down with a high capacity NiMH battery in place of SLA batteries, Pololu 37D Motors with encoders, Scooter 5.7 inch diameter wheels, Rpi SBC, and a low-end RoboClaw Motor Controller.

Might even be able to use two NiMH batteries: One for the motors and one for the electronics?

Comments?

Regards,
TCIII

Reply all
Reply to author
Forward
0 new messages