Category: Uncategorized

If I make the switch from an Arduino to an SBC, should I go with the Beaglenone or Pi?

The Pi 3 has built in Wi-Fi. The Beaglenone has its dope PRU, though is have to look up how to even touch it.  The Pi has a camera port,  though they both have plenty of USB once they go headless, and I have a couple hubs if not.

I should count up the number of usable GPIO on the Pi and go from there. Chances are the Beaglenone would be better for things with most moving parts due to the GPIO  count and the PRU. So I guess as long as I have  enough  pins, I should call it good with a Pi.

I still need to chop back the basil enough to move it to the Den.

Derp. They didn’t turn off. I had written some ghetto code to turn on the lights (and off again) on a standard schedule. 

The on worked,  but add I’m sure we can all guess from the title, the off again failed. Since it was the world’s largest latest method, I can must assume that it was a basic error that’s achievable in C. What I did originally was a database that had 24 values, one for every hour. A one turned it on, and a zero caused it to do a half second pulse. Since its been on for quite a bit now (four hours longer than it should have), I must assume that is using the Boolean definitionof not zero being false. Tonight, I’ll chuckle the lights by hand. Tomorrow, I’m thinking I might bite the bullet and start moving everything over to running on a Pi. It seems like a bit of overkill.

Then again, for future purposes, it might be the cure to the future ills that I’m already seeing. At a minimum, I need a boatload of analog read and write pins. If I go with a Pi, I’d have more pins, Wi-Fi, and a camera port.

Or I could stay with it being a stupid system.

Rather than working on the things I’ve been working on, I played Stellaris for a while. Then, as coherent thought started to flee the coming of the mad king sleep, I stayed researching things that weren’t on task.

Basically, after bouncing through a dozen or so websites, I found something I liked the idea of, just not the implementation.

So, a vase with a screw driven linear actuator.  The arm of the actuator is a “flower stem”. As it raises out of the vase, the flower opens. Inside the opening flower petals are reflectors and LEDs. Now to draw this out. Prolly do it as an alarm clock.

 .that was my goal tonight, and I slew. Well, I got it done. 

Comedy value, today at work I was joking with someone about having one of those metal brick Netgear hubs lying around, and how I should resurrect it, but I had let the magic smoke out. I now question that. I was going to use this power supply for the extractor, wanted to make sure it was asked up right, plugged it in and touched the wires to the fan the way I thought they’d be. Nothing. Switched which wires were where. Nothing. Slashed the wires together. Nothing.

Stupid power supply.

Found another power supply  that I’d already cut the tail of of, it’s the same voltage, and this one was actually labeled Netgear. Did I really only fry the wrong power supply?

So tomorrow I might take a look inside and see what I have.

Solder up the regulator and LEDs and make some light. But in all seriousness, I need to do that, figure out where in mounting them, mount them, and mount something to keep me from being blinded by light leaking out.

It’s not laser bright, but…

So last night I was throwing some power on them trying to get a solid number on their voltage drop.  Before doing this though, I though I’d do a quick contact connection of 5v while holding the LED on my knee. The light was intense enough that I have myself an instantaneous headache and had to walk away. I guess that’s what they mean by ultra-bright.

I couldn’t get motivated Saturday, so I didn’t. I played some Stellaris. I like me some incredibly slow building games.

Sunday…

• Cleaned enough to find my soldering iron, solder, flux, picks, and stuff.
• Combined all the caches of used beer caps for future mosaic. I’m thinking of how to make an algorithm for it.
• Soldered a janky connector to the range finder, and verified function. It works, but it’s a bit bulky. I should get and use the right tools for the job.
• Soldered leads to two nails for the moisture sensors. Oh. No I didn’t. Those nails don’t more holding solder. I did get them working eventually, but then I realized the hookup wire I used is way too big for a breadboard (and stranded to make out really not work).
• Wrapped two more nails with different hookup wire, though I haven’t soldered yet. I brief I had a reading, I then wasted the fern they were shoved in, and then checked the raging again. These work. I still need to set them up to alternate the current back and forth to reduce corrosion.

I’ve got the in-laws coming in less than a month, so it’s inmate proofing the house time. While I’m sure this sounds terrible, it’s just the easiest way I have to visualize what I need to do, and it gives me a timeframe. A shot clock.

• Clear space for indoor gardening
• Light rig therefore
• Decommision the upstairs cat box
• Mount the cat tree to the wall
• Organize my junk piles downstairs
• Hang my daily wearing clothes downstairs
• Put all my books away
• Vacuum vacuum vacuum
• Hot water service
• Verify operation of ac, and build a better drip pan
• Putty the mid bathroom screw holes
• Clean all drains

So this seems reasonable. I just need to start today, and work an hour a day. Plus or minus.

I hunted down my ir rangefinder. I quickly realized that I had writes coming off, rather than angering handy for plugging into stuff.  Now, going into the Redbot driver board, I’d need a three wire female connector. After more hinting I found some tall headers, so I chopped the positions.  Then, I yanked the pins,  shoved my freshly stripped wires in, and gave it a shot. I got dice, so it’s time to fix this in place.

Header, check. Rangefinder, check. Shrink tube (or whatever it’s called), check. Zippo, check. Light fluid. Crickets.

So tomorrow I’m doing assembly. The cost value is that as I typed this, I realized I should just hot glue it any way. 

Still going to do shrink tube.

And at first I’m thinking I need to find a lot of cat pictures (easy, internet) and use those for building a whatever it is cascade.The thing is though, I need to identify one if two cats, not a dog, a different cat, whatever. So really what I should do is get the rover roving,  spring me pictures, and I’ll start cataloging them. I’ll then take these pictures to use for building the positive and negative results folders for whatever. Now to think about how to focus on guiding stiff rather than naming what I found.

…a lot of cat pictures.

So the first set of tutorials I was working with have me some annoyance. I don’t pay nice when I follow the tutorial, where guy is trying to sell his books, and then I have to do a search to get back to where I was to find out what was wrong in the original tutorial.

The second tutorial sourceseems to alternate between Python 2 and 3. Also there send to be some disconnects on how someone would choke the tutorials to begin with.

Now I’m on tutorial source numero trés. I’m liking them except I’m at an impasse again.

So I have a couple goals here with Mustachio…

• Sweep the floors
• Locate cat toys
• Corral slippers
• Track the kittehs
• Other

This works out to be sure a bit. One issue is that Reese’s kitty face doesn’t trigger with the built in CV functions. This means I need to build my own trainer. This is also a lot of work for a little Pi, so I should off load the work. So I need to run across on the desktop, that takes pictures from the rover,  processes them, and hands them back to the rover to act upon. I do want the operation decisions done on the rover itself. Passing the pictures back and forth should be easy as N-O-D-E that spells moon. The trick is I need to get the trainer going, and that’s not playing nice. I think tomorrow I’ll do the whole setup a  python virtualenv thing. That kind of did gives me a sad, but hopefully it’ll do the trick.