BlueBasic: BASIC for Bluetooth

One of my goals for this project (of which I have many) was to provide a simple, easy way to upgrade the firmware on the devices. I had three requirements: 1) that you could upgrade from anywhere, 2) that you didn't have to scramble around obscure website looking for the correct image, and 3) that your BASIC program would survive the process.

The original console code could upgrade the firmware manually (with appropriate scrambling) and maintain your code. The recent iOS version didn't do upgrades at all.

But ... finally ... the code has dropped to support auto-upgrading. It's not so "auto" that it will upgrade a device without telling you (like Nest does), but it will discover if new firmware is available and prompt you to upgrade. And it works across all the console clients (OSX, iPhone and iPad as of today)

One extra; I added a recovery mode in case the upgrade fails bricking the device. This will be detected and the correct upgrade can be attempted again.

Almost like a real computer ...

Just uploaded an iOS (iPhone and iPad) version of the Console program to Github. They're part of the same xcode project, but a different target. Of course, unlike the OSX version, you cannot side load this one onto your device so I will have to get a version submitted to the AppStore soon (later this week probably).

I realize that a WIndows version would be more welcome, and that remains on the list, but adapting the OSX version for iPad/iPhone was a quick way to reach a few more devices.

Now ... if only my icons didn't suck.

In the next few days I'll be putting in an order for about 100 CC254X modules to build a batch of prototype boards. This is most an experiment for me to see how easy, or not, it is to do a "small batch" run. I've been negotiating with various people on Alibaba because that seems the cheapest place to get the core module, but I admit to being somewhat nervous about handing over $400 without understanding the risks involved. Anyone have any insight or experience here?

Had chance to assemble the prototype 2 board this evening. I made my solder stencil using the Silhouette Portrait cutter and mylar sheets (see http://hackaday.com/2013/03/11/pcb-stencils-for-200/) which worked amazingly well. Hand assembling the components is definitely fiddly, and i think for the next batch I'm going to buy slightly bigger capacitors because these ones are way too small for me (0402s). But, once done I popped the board into the T962 oven, baked at gas mark 10, and voila! Popped it into the programming rig, flashed the latest firmware, and everything works!

Just added the battery to the back, and the power management and boost converter appear to be ticking along correctly

So - pretty pleased and, as ever, slightly shocked that it all works.

The latest board from OSHPark arrived today. Can you spot the mistake? That's right you at the back; while I very carefully position the battery holder, I forgot that it would block various signal holes. Fortunately they're still accessible from one side of the board so fine for prototype-2, but not what I need in the final boards.

It's been a busy week, most devoted to rewriting the console program in Swift (Apple's new development language). As a poor Objective-C programmer and a much better Java one, there's much to like about Swift. The new console is now on GitHub (the old one is still there too). Unlike the old one, this is much better structured to allow me to do interesting things with it in the future (including an iPhone/iPad port).

And yes, I will re-learn C# at some point soon to build a Window console.

Finally wrote up a simple "Getting started" guide. You can find it here: https://github.com/aanon4/BlueBasic/wiki/Getting-started:-Hello-World

Just uploaded the latest board to OshPark - https://oshpark.com/shared_projects/puBoiqUm and sent out an order. As ever, a few guesses here and I'm not at all happy with the power and ground (but this is largely because I'm still learning the tools). This board includes the boost circuit and provision for a CR2032 battery.

The second power testing board turned up today, the TPS61220. Running this board with the CR2032 battery in the same configuration as the previous test, the board consumed 49uA. This is a little better than the LTC3525. This board also has an easy enable/disable switch. With the module disabled, the power consumption is 34uA; 6uA more than bare configuration. So, given testing, I think I'll go with this module. It's also half the price which is an added bonus.

So the LTC3525 board turned up today and I've been doing some power testing. Running the basic board from a CR2032 battery doing only limited advertising (every ~5 seconds), the board consumes 28uA on average.  If we run the battery through the LTC3525 so we have 3.3v for any potential peripheral devices, the average is 57uA; twice as much (see the photo below). Not great. I probably need to use the shutdown feature of the chip to switch over to pure battery power when sleeping (perhaps make this configurable on my boards) to reduce power some more. Not clear the test LTC3525 board makes this easy :-(