Results 1 to 6 of 6

Thread: Rip Van Winkle

  1. #1

    Rip Van Winkle

    It's been a while since I was actively involved in Robotics. My passion hasn't changed but sometimes life's priorities do.
    I feel like Rip Van Winkle, just awakening from a decade long slumber.
    On one hand I am a little disappointed, 10 years ago I hoped that everyone would have personal robotic "servants" by now. Mundane chores like grass mowing, laundry and vacuuming would only be done by bots and the die hard purest. I really thought that cyborg type implants would be here, I could really use an email, spreadsheet, and word processor implant myself and will gladly wait in line when they become available.
    The other area that I expected to progress faster than it has are the sensors and motor drivers. Oddly enough the MaxBotics sonars, the Sabertooth motor drivers, the Sharp IRs, just to name a few of the parts that have been waiting patiently in the parts bin for ten years for me to get back to them are the same parts that seem to be preferred now.

    So what has changed?
    Obviously the Raspberry Pi has been a game changer. I remember being excited about the Pico ITX, they were "tiny" at the time, little did we know. The Arduino had just started catching on then, but "serious" robot enthusiast were using the Axon, which by the way has been from what I can tell mostly unchanged in that time frame. I am pretty sure I can be quoted as saying "why limit yourself to 17 I/Os when you can use 55". Obviously, I would never expected the Arduino to catch on like t has.
    Which brings me to the biggest difference between now and 10 years ago, it is the "open source community". By "open source community" I am referring to things like GitHub and YouTube, which I attribute to the Arduino popularity but forums like this one also played a part in making robots, and electronics accessible to anyone.
    Having looked back over this forum for the last 10 years to see what I missed I am excited and inspired to be a part of it again.

    For the record, I not only still have my Trossen Robotics "I build Robots" shirt from way back then, but I still wear it proudly.
    I reject your reality and inject my own.

  2. #2

    Re: Rip Van Winkle

    In my opinion, compared to 10 years ago, "motor control" is now a solved problem.
    Back then, some people were still using BJTs and large heat sinks / cooling fans.
    Robust, affordable MOSFETs have changed all that (as has robust, affordable LiPo batteries -- back then, NiMH was still a thing.)
    These days, a brushless motor and an ODrive is the go-to solution for hobby motive control. Such a nice solution!

  3. #3

    Re: Rip Van Winkle

    jwatte, Great point - the LiPo's and the brushless motors! How did I miss that?

    I hadn't found the ODrives until your post - that is exciting - I instantly have a project in mind for the big one. Thank you.

    Also, cameras have come a long way since then as well.

    I see there are some inexpensive LIDAR products, I haven't read up on them yet, but that was a multi-thousand dollar proposition back then.

    Makes me wonder what else I haven't found yet?
    I reject your reality and inject my own.

  4. #4

    Re: Rip Van Winkle

    If you think the Raspberry Pi 4 is too weak: Neural networks on drones / robots:

    As far as LIDARs go, I'm not a fan for mass market outdoors applications (like self-driving cars,) but I think they're pretty great for indoors navigation, industrial applications, and just basic data gathering. I'm excited by the RPLidar-S1 which is a time-of-flight device they recently came out with for < $1000:

    And then there's robust stereo vision ready to buy:

    (There was the whole "structured light" fad starting with the Kinect and using a bunch of RealSense sensors, but I'm even less a fan of those than of LIDAR -- I think stereo is a better solution for most mass market applications there.)

  5. #5
    Join Date
    Sep 2010
    Rep Power

    Re: Rip Van Winkle

    I almost posted before about this in one of the Malum IK threads, but that ODrive board brings the ick back to the surface...
    if you are not using BGA ICs with sensitive power requirements and/or dealing with extremely space constrained applications, then double-sided SMD components on a board is simply a failure on the part of the lazy designer. Sure, many fab houses are capable of it, but it adds extra time and equipment costs to manufacturing and creates new modes of failure - especially when dealing with high-power components that already have the potential to get hot enough to desolder themselves and nearby components.
    Please pardon the pedantry... and the profanity... and the convoluted speech pattern...
    "You have failed me, Brain!"
    gives free advice only on public threads

  6. #6

    Re: Rip Van Winkle

    double-sided SMD components on a board is simply a failure on the part of the lazy designer
    Double-sided is useful for two reasons:
    1) You can make things smaller!
    2) You can put heat spreading parts on different copper planes!

    Routing traces out from components requires a certain amount of space, and even without BGA, most modern components are QFN with 0.4mm pitch or less. Often, you need the other side just to fit the de-coupling capacitors ... Many parts want 1uF or even 10uF de-coupling, often on multiple rails, and that shit is BIG! (0402 and 0603 size)

    From my point of view, if the components get hot enough to melt lead free solder, I already lost. And if MacroFab charges nothing extra for the double-sided, and does the QC for it, then that's not a problem for me. (Also, they charge nothing extra for 4 layers over 2, so you could assume their default is "4 layers, 2 sides" and just run with it.)

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts