Well, well, well...I guess I forgot to mention that the stork visited me yesterday and brought me a new addition to my family...of test equipment.


Yep...you guessed it. Bought me a used network analyzer. It's an HP 8753D 3kHz-3GHz vector network analyzer with a 2-port S-parameter test set.

I decided to buy it three weeks ago on eBay because the price was really good ($6k). Normally, 8753C's and 8753D's go for about $8-10k and I figured I would need one since I'm going to be designing a bunch of RF hardware soon. I didn't really mention it though because it was kind of a risk. Spending $6k on eBay and wiring money to a guy in the countryside of France isn't exactly a run-of-the-mill transaction for me. If I got burned on this deal, it would have been pretty embarrassing. Lucky for me, the guy was really cool and walked me through the transaction. The packing was good, and the analyzer is up and running and passed all the self tests. 

Did I mention I also bought a 50 ohm calibration kit? *cackle*

I also got word from Chris at Accurate CNC that my PCB mill will be shipped this week. There was a delay because its custom built and one of the parts got delayed. 

Yep, looks like I won't be getting a car anytime soon...but along with my scope, logic analyzer, and other equipment, I'm starting to feel like my home lab is kicking some serious ass!

Contiki OS made it on to the main page of Slashdot . For those that don't know about it, getting on to the front page will usually bring your site down due to the volume of traffic that flows into it. Looks like Contiki hit the big time. Congrats to Adam Dunkels, Fred Osterlind, and the rest of the Contiki team.

Yes, that title is correct. I'm finally going to take my rightful place among the extreme electronics geeks of the world. No, it's not because I'm writing an open source stack based on a bloated spec. It's because I finally got up the balls to purchase a CNC PCB Milling machine. Sure I told myself that the cost will justify itself by the time saved in spinning boards, or that it'll be useful for consulting, or helping out the companies I'm doing part time work for, but the reality is that I just wanted one really, really, really bad. 

It basically means that I'm one step closer to my dream of coming up with an idea and having a working board within a day. Sure I tried the UV PCB kits, the iron-on toner transfer PCBs, but they just weren't repeatable enough. They also lacked the resolution that you need for real electronics these days. That means about 6 mil traces with 6 mil spacing at a minimum. Otherwise, you won't even be able to touch the QFN radios out there. 

I purchased the PCB mill from Accurate CNC which got good reviews on the web and is less than half the cost of the similar machine from LPKF . It sports a resolution of 0.1 mils, low runout, and did I mention less than half the cost of the LPKF (BTW, the LPKF  is close to $20,000). It should be able to handle 6 mil/6 mil layouts with ease and also handles drilling the board and matching up the 2 sides of the board so that they align correctly. In other words, I can finally get repeatable prototype PCBs in a single day (or make that about 10 minutes). That kind of capability is almost unheard of at most electronics companies I've been to (and I've been to quite a few), much less for open source projects. Chris at Accurate CNC was also extremely generous and held my hand through the whole process, answering all of my n00b questions with alot of patience. He's even helping me purchase the tools to ship with my order since the shipping to Tokyo from the tooling shop is over $100. I actually felt bad because I thought I was making him work too much. Now that's service...

Well, one of the main reasons I decided to purchase the machine now (other than lusting over it), is that I  just figured that if you're going to make RF PCBs, you gotta be able to spin them fast because you end up throwing a lot of them away. And since the PCB-making time is approaching on this project, might as well be able to crank out a couple of quick protos to see how the layout will perform. Also, I'm hoping this machine will take away that PCB mental barrier where I have a good idea, but its such a pain to spin a PCB that I never implement it. Now I have no excuses not to act on any bizarre electronics whim that catches my fancy.

Anyhow, I'll post videos and let everyone know how things go with it. But for now, I'm going to sit back and bask in my growing geek manliness...

BTW, here are some stock pix of the machine....beautiful...

I'm slowly getting off that mental block and have been writing code again. I'm not letting something like that hold me down. I kind of anticipated something like this happening which is one of the reasons why I shifted to working part time at my regular job. Its similar to burning the boat behind you (although I guess I still have a life preserver), where the only option is to actually finish the project. With a lot more riding on this project, there's more motivation to get past the problems I encounter and get this thing out. I didn't anticipate that it would be such a mental game, though.

Anyways, I'm starting to employ some project management techniques to clarify the remaining items that need to be done, and I'm also keeping track of my time so that I give myself adequate breaks. After listing out the things that need to be done and breaking them down into manageable steps, it's easier for me to see where the end of the road is (for the first release) and how I'm going to get there. That was one of my main issues where the task started to feel overwhelming. Now it's just a bunch of small tasks that I know I can pull off. As Joel Spolsky said (Joel on Software ), software (and life) is a game of inches .

I ran across this from one of the Dell ads that is periodically emailed to me and thought it was hilarious...

One of the questions that I'm asked most often by people is: "How are you going to make money if you write open source software?"

The questions is actually not that hard to answer, but it takes a bit of time to explain. Most people just want a simple answer like "I'm just doing it as a hobby and will be returning to work soon" or "I'm doing it to get exposure to get a better job". In fact, as the software progresses and I show it to people, many people have told me to rescind the open source license and sell the software. Ha ha ha…no way.

I've been working almost full time on the software and living off of working part time for about four months now. In fact, it's been like paradise for me. The fact that I'm working part time relieves a lot of the work-related stress I had previously because the expectations people have of me at the job decreased along with my hours. I'm amazed how much distraction work-related stress caused me. Since I've been part-time, my thinking has become clearer, my goals are sharper, and my software skill has improved immensely (I'm not trying to be humble right now).

I started noticing the differences about two months into fully devoting myself to the project. That would be about June since my conversion to part-time occurred on April 1st (April Fool's Day…what a coincidence). It took about that amount of time to start shedding the emotional baggage of working for companies for the past 10+ years. As I became more independent, I realized that the success or failure of the project rests solely on me, and that motivated me to work harder on it. It's not like working at the companies I've been at where it requires meetings and management approval to get something done. The fact that I can do something that I enjoy and also believe in is unprecedented for me. The freedom I started to feel from being able to make my own decisions was very addictive.

I'm finally going home today. Although this was a long trip, I got the chance to meet many interesting people in the Zigbee field. I also learned a lot about real implementation, especially regarding Zigbee and power metering. Of course the most important point of this trip was that I got to meet my 6 month old niece and she actually liked me!

Thanks for everyone's patience. I'm hoping to start up the software effort once again in a day or two so that I can plod forward towards a release. 

Well, still in the US and attending the training for my part-time job. Unfortunately, for this week, the part-time job becomes full-time...as in all day and most of the evening. Because of that, it's impossible for me to make any progress on the software until I get back to Tokyo and resume my normal schedule.

It sucks that the evenings are also taken because I'm looking forward to implementing the ZDO and the Zigbee application layer now. I think that will be more fun than doing the transport layers which I've been working on until now. Anyways, I still have time to post updates to the WSN news feed on my site in the mornings and late evenings. 

See you all when I'm back in Tokyo next week. 

Lately, I've been feeling a bit fatigued from working on the stack so much so I decided to try and change up my scenery a bit. Until recently, I've been writing the stack out of a cafe near my apartment. It was quite nice, but they closed for renovation so I've had to move to a nearby Starbucks (ugh!). Location is pretty important to me, and the Starbucks felt like it was just sapping my energy so lately, I decided to make the trek to Akihabara everyday. Now, I'm writing the software at the Linux Cafe in Akihabara . It seems like a fitting place to write open source software, and I felt rejuvenated since the first day I was there. The vibe in Akihabara is great. I can window shop all of the microcontrollers and sensors , check out the used test equipment store (I'm still eyeing the 3 GHz network analyzer but I can't afford it), buy surplus resistor and capacitor reels for $5 a piece, and be surrounded by anime action (and erotic) figures. And that's just when I'm taking a break from writing the software. Ahhh...Akihabara...the jewel of Tokyo...

I feel bad that I've only been doing DevJournals lately. I wanted to do more tutorials and comparisons, but the stack is eating up all of my free time. It's like a jealous girlfriend, only I don't get any sex. Hopefully I'll get some breathing space soon to put up some more flash tutorials and update my chip comparison. I'm also starting to do research for a module comparison since I think module's are a better way to start out using Zigbee than chips are.

Okay, now it's anime time. G'night.

[Question to myself] Hey Chris (aka Akiba), how come you're posting so much news? I thought you didn't want to be a news aggregator!

[Answer to myself] I dunno. I tend to sway with the wind. More precisely, I'm basically indecisive and spontaneous. Two excellent traits for an engineer. 

 [Now here's my post..]

Lately, I've been posting a lot of links to news articles and blog posts. One of the reasons is that now that I can safely cordon off the main part of my blog from the News area, I want to show everyone how much info there is out there. There's a lot of talk going on regarding Zigbee and WSNs in general. Also, I don't think that just posting press releases or articles based on those releases is sufficient to get a good picture of what's going on. Often, standard news outlets are the last to pick up on trends. To really get a feel for what's going on, you have to balance the news with grassroots bloggers who are usually the first ones on the scene when something interesting happens or are actually trying to implement the things that the press releases just talk about. The trendy term in the Silicon Valley set is "Noise", which is probably best explained in this article but it does have a ring of truth to it. As I mentioned before, it's up to the individual to choose what's relevant and what's not, but the information is definitely out there.