I’m tinkering with a bot based on the OOPic-R. Eventually, I intend to use it as the target of a learning engine, as my background is much more in evolutionary learning engines rather than robotics. For this, I will be mostly using the robot as a mobile sensor base, with all of the intelligence moved onto a network of computers.
My question has to do with communications. I’ve been looking at various options for getting a wireless connection to the computer, so that the bot can roam my house while connected to my home LAN. It seems like Bluetooth would be the way to go, but I’m looking for any input or suggestions. I’m also looking for specific parts that would work. The connector on the bot is currently a RS-232 plug.
I’d suggest taking a look at the WiPort, its basically a WiFi enabled serial port and probably closest to what you’re looking for. I personally don’t have one, but people here have used it successfully and there are tutorials for Lynx bots on the main site. It also probably has the advantage in range. I have gotten 1km WiFi links myself (and since the WiPort is just a normal piece of WiFi gear assume it could do similar), and know that people have easily obtained over 7km omnidirectional with just some inexpensive custom antennas. However, the downside of the WiPort is price. The modules are rather expensive and you more or less need a carrier board which is also somewhat pricey. The going rate I found online for one with a board was $250. If anybody can find it cheaper, please post since I’d love to get one.
If range isn’t that important, I’d suggest using a bluetooth module from Sparkfun. They have a tutorial that shows just how easy it is: sparkfun.com/commerce/present.php?p=Bluetooth%20Primer I personally have a BlueSMiRF module (basically the same one as the tutorial, but it can go faster) and followed the directions in the tutorial. It probably can’t do much more than 100ft (Sparkfun says 350ft but that is an exaggeration or at least at school with interference from stuff it can’t) but its $60 price point is a big advantage. Truthfully, I’m not sure how far bluetooth can go. It’s 2.4 Ghz so using the same antennas used for WiFi it should be able to get similar range, though nobody has been able to do this. My bluetooth adapters are dongles which like the module has a little internal antenna. I haven’t tried to hack them for range, but my guess is that its not really possible or somebody would have done it already.
The only module I suggest you avoid is the Bluetooth Appmod from Parallax. Besides a hefty price of $130 for basically the same functionality as the Sparkfun module with a bigger footprint, a number of people here have had trouble with it, as the documentation is extremely poor. I think the idea was it was designed to be used to teach the bluetooth protocol so if you want to plug and play, you’re going to be disappointed.
This sounds like how I will be using my small Octabot (BOE-Bot type robot). I just need to get a “brain” for it. This one will be a sociable people follower also.
It really depends on how much data you plan to transfer between the robot and the PC. If your data requirements are low, check out Bluetooth from Sparkfun. For higher throughput and data load, there is the WiPort on a WiPort Board.
I am excited about the Sparkfun Bluetooth. Take note their Bluetooth module is Class 1 which has a range of around 100 meters instead of 10 meters like for Class 2. I am definitely planning to experiment with this Bluetooth setup, and it will be a great match for the Parallax Propeller chip since they are both 3.3V parts. I will also eventually work with WiFi on a robot, but it will be with a Linux based microcontroller.
Yes, I didn’t use those exact modules, but a few years ago I got a relatively cheap ($25) set of 433 Mhz modules which was almost exactly the same and I got rather poor performance, a little better than spark-gap style wireless. Some modules basically take an on/off signal and filter out all but resonant frequency with a choke, resistor, and capacitance, then use a similar bandpath filter on the receiver. The higher-end ones use an AND operation on the output of a crystal, fixing some of the band issues with the transmitter. However, generally the frequency output is over a much wider band than it could be, hence wasting a good deal of the limited output energy of the transmitter.
The cost looks nice, but to get anywhere near the functionality of a bluetooth link (forget the speed) you’d need two sets, a stage before the antenna with an op-amp (or at least a transistor amplifier), not to mention two pics or other small microcontrollers to handle modulation of the signal (RS-232 signals aren’t clean multiples of 433 Mhz after all) not to mention that error correction is usually done in the radio. Including all that, the price looks a lot worse.
Basically you need to do a lot of work on those to get anything useful. Sparkfun already does this and sells it too: sparkfun.com/commerce/product_info.php?products_id=560 This unit is a lot better than the individual links (note the Atinys that do modulation and error correction, and the Faraday shielded amp), however, for the price of two of these you can get a bluetooth module and a USB bluetooth dongle for your computer and get better range, faster speed, automatic compensation for dynamic interference with channel migration, and wide hardware compatibility. If you need pure RS-232 (i.e. no PC) bluetooth master serial ports exist too, albeit they are a bit more expensive. Note, I honestly don’t believe the Sparkfun range they cite. They claim 3000 ft, which is laughably bogus. Considering the commercial version of this is what does 900 Mhz cordless phones, and RF remote controls, etc. I’d say 30ft is much more realistic.
