Final Robots 101 Deck

These are the slides I will use in the upcoming meetup;


Hope to see you there!

Building your first Robot

Sunday, Sep 21, 2014, 2:00 PM

Harbor Steps North East Residential Tower, Connect Room
1301 1st Ave Seattle, WA

12 Droid Masters Attending

Across the street from SAM, around the corner from Harley-Davidson.Dial door code #565

Check out this Meetup →

I’m in the cloud now, baby

Seems like lyrics for a new song.

I’m moving, and will be losing the most awesome internet ever from CondoInternet.  I had a nice 1Gig pipe and ran my own web sites.

But atlas, new place will have Comcast (I’m scared), but they are negotiating with CondoInternet. I put in the best recommendation I could.

In the meantime, Hello Azure. I tried Amazon for a bit, too much hassle. I’m hoping the Azure pricing stays reasonable.

The first hit will be slow. It should be ok after that.

RoboNUC Status


Motor speed as reported by Netduino;
Untitled-2Unusable as is. (Timing accuracy in managed code not likely. Open Sore requires $6500 toolchain to compile).

Motor speed as reported by PSoC;
(tl;dr  bugs, doesn’t work)

Motor speed as reported by Arduinos
hahahah – not enough interrupts to even bother.

Lessons learned from weeks investigating;
Arduinos do not implement SPI slave writes.
Netduinos do not implement SPI slave reads.
SPI sucks, invented by terrorist to drive people insane and waste time.

Next attempt ATTiny Atmels on I2C? I have not seen anything that can do what 6+ year old LEGOs can do yet, not impressed.


I did some moving averages on the velocity, played with the PID a little and got this;



I guess I can live with this for a while.  So hopefully I am back to real work for a bit.

Hypocritical censorship

bThey claim they don’t censor, but don’t dare point out something stupid that embarrasses the lemming membership. It will be gone in a flash. The entire previous page has been removed because it wasn’t going well for a few.

update: As has this post.  Like I said, sad bunch of cry babies.


Windows on Devices

So I am giving one of these Windows on Devices things a look.


I would have to say so far, pretty smooth and as expected for a very early device.  I have some ideas for projects.

One thing I wanted to do right off the bat was get the device onto my existing network.  Microsoft was kind enough to ship a USB network adapter and cable with the Galileo, so that it can run on its own little network without disturbing your existing network, but I already have somewhat of a Xduino lab (Arduinos, Netduino, and PSoCs) and an ethernet cable for the purpose, so I really did not want additional network wiring.

The short of the story is; here is the command you are looking for

netsh int set address name=”Local Area Connection” static

If you do not understand this, you probably should not mess with it to change what comes as the default.  Also be aware if and when you give the command, the correct response is for your telnet session to freeze. Again, if you do not know why, probably best to stay away.

I’d love to step by step what and why, but it really is an entire 9 week class on networking :)

Follow up:  Well I loved the speed of debugging. It felt very solid compared to other things I have been debugging lately. The second project I tried, Firmata/Cylon using Node.js failed.  I really have never done any Node.js stuff before so I am guessing the instructions made some assumptions as to the audience, and I was not aware of important steps (like npm does not even run until you manually create a directory).


So, like I always wanted to play with a FPGA.

And after spending a week with what I thought was going to be a breakthrough on the Netduino, trying to make it act like an NXT and do PID regulation of 2 motors with encoders, and incorporate Ultrasonic sensors for collision detection, I was getting nowhere.

Sure, it was probably code mistakes on my part, but I just could not get a smooth functioning PID going, and to me it seemed that is was because the velocity measurements were inconsistent. The faster I made the regulator loop, the better it got, but still not usable.

Most code I looked at used some sort of modified PID to smooth things out. Perhaps this is the fix it needs that I did not try.


But anyhow, I decided to take a break and try out a Cypress PSoC4 dev board I recently received. I went through the online video tutorials, and I am thinking this is kinda cool. I did some small variations of the presentations, and then I decided to go for it.


From scratch, no looking it up on the internet, I came up with my own Ultrasonic measurer, serial broadcasting to the PC.  Woa, very cool, and the numbers make more sense and are less noisy than anything I ever saw on the Arduino or Netduino. First try (as in first project, it took several iterations to make it work), this thing has potential. I like!



Robotics, hahaha