h1

Still Alive

November 18, 2014

I haven’t been writing here so much. At first, it was because I was frantically busy, trying to impress clients with knowledge by reading ahead in the textbook the night before. Then there was a wedding which I maybe over-planned because that’s what I do when nervous (it was great, we stuck to plan A nearly the whole time; I probably didn’t need to create plan N, let alone Q).

I finished with my clients the week before the wedding. I’d rather do something closer to my skillset. I like learning but I want to build closer to what I already know instead of continuing to cantilever ever more esoteric information.

I’ve been contract-free for a couple weeks now. That used to freak me out. I was ok for one week but not much more than that: OMG, will I ever find work again?!?

As I did the accounting, I found that the business can pay us for the rest of the year with the money we have in the bank (not even counting outstanding invoices). Sweet! I have three potential clients starting in the New Year, probably one will work out.

And, yet, I still didn’t write. There was hackaday judging but that finished a while ago, writing up my thoughts didn’t talk all that long since I had plenty of notes. There was some planning involved with my birthday party (cake, check; guests, check; champagne, check; flowers, check; planning done!).

I did go to a class from ST Microelectronics, acquiring an M0+ discovery board,  M0+ nucleo board with a BTLE daughter card. That was interesting. I wanted more “how to do low energy” and they had “look at our super shiney parts you can use to do low energy”. There were some hints but nothing new to me (sleep early, sleep often, sleep deeply). I also got a potential podcast guest to talk about Cortex processors but he was baffled by the idea people would listen to a whole hour of reasonably technical information. (Yes, this does concern me, we’ll see.)

And now I’m at a class about 3D printing, getting terrible and expensive ideas about things to buy for our business. This afternoon, I plan to attend a class about wearables and electronic textiles (stretchy conductors!). I’m here on a press pass (boggle), mostly signed up to attend because I have free time and I’m procrastinating about other things.

Both SOLID and the embedded systems conference have opened their call for speaking proposals. These are both “in my wheelhouse” (I hate that phrase but I find myself using it far too much, I suppose it allows me to say “in my area of subject matter expertise” without sounding like a total nerd). I don’t have a plan for them. I’m tempted to take a chapter from my book and talk about it. I’m tempted to compile a bunch of low power tips and making a presentation. I thought about updating my network comparison presentation (see Element14 text version). But these all seem like so much work. Can’t we just have a podcast about it?

I do have a little project I’ve been working on involving my MicroView (tiny Arduino based system with a pretty OLED screen). I’ll write about that separately since I’ll put the code up (even ask for input!).

Anyway, I’ve been distant for a month or two but you’ll probably hear more from me in the next month or two. I should be looking for a contract but I’m goofing off instead. For a little while.

2 comments

  1. Except using special peripherals/DMA in clever ways, are there any other tricks than “sleep early, sleep often, sleep deeply”? Having worked as an application engineer for several years at a low power mcu-company, I would say that pretty much covers it. btw, I would love to listen to hours of technical stuff!


    • Yes, definitely the talk would have to emphasize sleeping but there are more things to cover. Turn off IOs (analog in! usually…), clocks, and peripherals that aren’t in use. Use DMA to keep the core asleep. Lower your clock rate when you can (write/depend on timer functions that scale with clock frequency). Optimize sparingly and sensibly (possibly including run from RAM to turn off flash). Maybe some tangents into architecting a system that sleeps a lot (instead of tacking that on at the last minute). Keep score as you make mods so you know what works for your system (take out the ones that don’t have big savings but do cause big headaches). Debug techniques for systems that keep falling asleep. There is definitely enough material for a talk but it all already exists in various places so I don’t have a lot to add other than putting it together.



Comments are closed.