Wireless success… I guess

Well, I took some time this weekend to prototype a wireless grant circuit using the 315Mhz RF modules I bought from Sparkfun.com. On the left of the screenshot you can see the grant and its float switch wires, as well as the prototype grant wireless circuit. It reads the float switch positions and grant exit temperature and sends them wirelessly via a 315Mhz channel over to the other prototype receiver circuit, located on the right. That simply grabs the transmission out of the air and sends it to the laptop via the serial download cable.

As you can see by this screenshot, the data is being transmitted over to the 28×1 circuit. The first number is the float switch state… 0 = low up, high down, 1 = low down, high down, 2 = low down, high up, and 3= low up, high up. The second number is a raw 10-bit reading from the LM34 temperature sensor that I intend on measuring grant exit temperatures with. I held on to the sensor for a little bit before starting the test, so it gradually cooled down over time (as is reflected by the numbers slowly dropping).

Now there are some things that the test showed which you can see, and others you can’t – first, error transmissions do happen. You’ll notice that the temperature reading goes wacky in the middle (162 -> 199 -> 34 -> 157), which was caused by some problem with the wireless serial transmission. The second problem is a little more troubling – the transmission rate and range. I receive a new reading on the 28×1 circuit about once every 1-5 seconds, sometimes longer than that, even though there are 5-10 are sent per second – this means it is only picking up <10% of the data being sent. Now I don’t know if that is because my testing are is right next to the circuit breaker, my circuit sucks, or what – but it is not reliable enough to realistically control the grant pump as is. A long 10-20 second delay could result in the grant overflowing before the pumps kick on, and that danger is enough to disuade me from pursuing it further with these el-cheapo units.

I may have to order a pair of XBee units to play around with and see if I get any better luck from a 2.4GHz, $50-75 solution. I sure hope so… the XBee chips have error checking and many other functions built into the hardware, so if not… then I just suck at wireless. Its not like the “wired” version won’t work – in fact it will be alot faster, but… not quite as chic.