Meeting Chat
00:38:50 Alice Crowe: It might be a few years out of date, but my impression has been that eGPU ports can be somewhat flakey and poorly supported. Entirely possible I’m wrong, as I only looked into it a few years ago when I was working on a homelab server.
00:41:41 Melissa Chhaunkar: I missed last month’s meeting. What are these models being used for? Do you put in your hardware specs and try to tune the software? Or is this just generating the general software stack? When I was going through university, we used a lightweight simulator to iterate over a very specific course and tune for that specific scenario.
00:48:37 Scott Smallbeck: UNO Q
00:49:48 Randy C.: There is a set of motion control chips made by Performance Motion Devices that does the acceleration with jerk. This is the motion chip in my 4 foot tall Roger robot.
00:50:27 Connor: yes
00:50:55 Randy C.: All you have to do is load the registers with the parameters then start the motion.
00:51:04 Chris Powell: Can you post the link for later review?
00:51:25 James Newton: This one right Randy?
https://www.pmdcorp.com/products/multi-axis-motion-control-ics
00:52:32 Randy C.: Yes that is it. My motion chipset is over 10 years old.
00:52:50 James Newton: I believe that does point to point, but I don’t think it manages trajectories with pass through way points which is where it gets super complex.
00:54:04 James Newton: I believe this is the video we are watching? In case Lloyd’s connection drops out:
https://otter.ai/pricing?utm_source=oa-chat-basic
00:54:26 Randy C.: You can set it to start a motion with no end point then change the parameters on the fly.
00:56:04 Randy C.: The real benefit with the multichannel versions is you can setup a virtual gear ratio between 2 of the channels.
01:08:49 Melissa Chhaunkar: Is this a modification of a commercially available robotic mower?
01:10:57 Blake Raymond: I’m also curious as well, because I delivered one of these mowers to somebody’s house a day or so ago as well. LOL!
01:16:10 Craig Austin: Yes, modification to add autonomous capability to an off the shelf RC slope mower
01:18:29 Melissa Chhaunkar: I guess my mic isn’t working. My friend has an autonomous mower off the shelf and the battery is so bad it has to charge before finishing the lawn.
01:20:32 Craig Austin: Mine has a gas engine which runs the cutting blade plus an alternator which keeps the batteries charged for running the electric motors that drive the tracks. Battery is a small 12V lead acid that will last about 5 minutes running the tracks without the engine driving the alternator
01:22:45 Blake Raymond: With that gas engine, that’s gotta be super heavy?
01:25:19 James Newton: Heavy, solid steel, spinning cutting blade… everything you want in a robot!
01:26:45 Craig Austin: Yeah, it’s well over 100 pounds…
01:37:17 James Newton: DO-178C and 332 (for OOP)?
01:41:55 Scott Smallbeck: James, I have a RoArm-M3 arriving Monday. First time doing anything with an arm (really first time doing any robot project), may reach out if I end up having a breakdown after failing to get it to cooperate.
01:51:31 James Newton: Sounds good Scott,
01:53:11 Melissa Chhaunkar: Static analyzers are great at catching errors in AI-generated code if your compilation is expensive.
01:53:28 James Newton: I’m going to be VERY impacted started Monday with the new job, but I’ll point to things if I can. The new discord server is a great place to connect.
01:54:59 Scott Smallbeck: Reacted to “I’m going to be VERY impacted started Monday with the new job, but I’ll point to things if I can. The new discord server is a great place to connect.
https://discord.gg/SdMmfC8f” with 👍
![srs-logo-2[1]](https://seattlerobotics.org/wp-content/uploads/2021/02/srs-logo-21.gif)