Saturday, 31 December 2011

Using a DS1307 clock with a Teensy

This is a fuller description of the clock I put on Google+ yesterday.

The purpose of this is not so much to build a clock but to verify that I can make a ds1307 clock work with with a Teensy++ 2.0 board. Since I had an LCD lying around I hooked that up as well to display the time and date.

The ds1307 is a clock on a chip. This variant comes as a neat little package with a watch battery on the back. This is important because once power is disconnected from the Teensy it 'forgets' the current time. The ds1307 clock remembers it and keeps on ticking so when I power up the Teensy it gets the current time from the ds1307 and uses that. The ds1307 also has a few bytes of spare memory which, like the time, is also battery backed. So stuff I store there also survives loss of power.

The Teensy is a nice little package with a AT90USB1286 processor, a USB port and a load of pins I haven't figured out properly yet. Unlike the Arduino the Teensy's USB is supported directly by the processor. That means we can fully mimic USB functions, which I'm not interested in yet, but I also gather it means loading programs into the board is faster (though I haven't used the Arduino). This processor has loads more memory than the Arduino UNO, though there are other variants of Arduino which may be more comparable. I program the Teensy using the Arduno IDE, however. That means the Arduino libraries for device support are available, as well as all their samples. There is a custom boot loader in the Teensy which some people might not like, ie not completely open source. It's freely downloadable though.

The LCD is a KS0108-compatible 128x64 and I adapted the demo program for that to simply add the time pulled from the clock. All wired up it looks like the image on the right.

The ds1307 is the red board just left of the LCD. Almost all the wires are for the LCD because the clock only needs 4 wires

The code is adapted from a demo for the LCD so there is a bit of rubbish in there that I could trim out if I were less lazy. The vital bits are the references to RTC and RTCE. RTC is a reference to the DS1307TRC class which manages fetching the time from the ds1307. The whole trick, really is this line:


which tells the time library to pull the current time from the ds1307 so when we request the time using:

time_t currentTime = now();

we get the time from the ds1307. Actually there is some buffering going on in the library which I don't have to care about.

But that doesn't quite get everything I want because the time in the ds1307 doesn't know what time zone it is in. The simplest way to handle this is to just set the time to the local time and forget this issue, but it will not look so simple when we change from/to daylight savings, or when I travel to a different time zone.

Besides, that extra non-volatile memory is quite attractive too, so I added a subclass of the DS1307TRC class called DS1307TRCE, which has a couple of extra methods to save and fetch stuff from that memory. In this case I pass a struct called config using this:

RTCE.read1((char *)&config,sizeof(config));

this reads from the extra memory and copies it into the config struct. There is a similar write1() method which does the opposite. What do I use this for here? I use it to hold the time zone as an offset from GMT. Having fetched the config struct I do this:

time_t currentTime = now()+ (*SECS_PER_HOUR);

That corrects the time. The current value for is +13, when we change to winter time it will be +12. I also store a string in the config struct but I don't use it yet. I anticipate storing other stuff in there though, depending on how I want the eventual device to behave. This is nicer than hard coding things.

To change it I have another program which gets data from the serial interface. The development environment allows me to just type in stuff and it ends up in the config struct which I then write1() to the ds1307. But I might eventually build a UI on the eventual device so set what I want (when I have decided what I want it to do...)

Friday, 9 December 2011

Competing on Content

This is probably not huge news everywhere but it is stirring the waters a bit in the publishing community.
Amazon have announced a deal for writers who want to directly publish through them. I won't bore you with the details here because you can easily find out more elsewhere.
But the thought I had about this is that Amazon is trying to compete with other stores on content. The deal includes an exclusivity clause so if you have a book you want to publish under this deal you cannot sell it elsewhere eg Apple, Barns & Noble, Sony etc. You can only sell it through Amazon. This makes a lot of sense for Amazon and I see why they want it.
There is also a more vague clause which prevents you from selling something that 'competes' with the book you put on Amazon. many people are taking a benign definition of 'competes', ie it means you can't just change the title of the work and submit it elsewhere. Again, I see their point. But it is always dangerous to assume the benign definition in an enforceable legal document. It might, for example, be interpreted to mean other works of the same genre by the same author. And Amazon may have no intention right now of it meaning anything sinister, but over time the people you deal with change around and you can find someone less benign on the other side of the transaction clutching the document you signed up to, so it is best it means what you want.
My point here is that competing on content serves no one but Amazon. If only Amazon has the book you want to buy then they can set the price and level of service they want. If it is only worth publishing your book on Amazon then the same applies. We all become deal takers.
From everyone else's point of view (both readers and writers) we want the likes of Amazon to compete with each other on price and service, but never on content.
We want them to pay authors the best royalties and charge readers the lowest price.
We want them to combine that with awesome searchability and bang on recommendations (and I have no problem with Amazon in this area).
But we want them all to be able to supply any book ever published, ideally, with those same levels of price and service.
Sometimes Amazon will be a little ahead of the others, sometimes it will be Apple, sometimes it will be B&N etc. And we will reward the ones who keep up in this game, of course. But we will only maintain high standards if there is more than one, and we must be able to make our decisions regardless of content.