View Full Version : Serializer & GPS

01-20-2010, 09:14 PM
I am looking for some help. I am determining if I should purchase a Serializer WL (http://www.trossenrobotics.com/store/p/5196-Robotics-Connection-Serializer-WL.aspx), and I need to have the ability of sending information to/from a PC. I would assume a ZigBee device such as the XBee Pro (http://www.trossenrobotics.com/store/p/5829-Xbee-Pro-60mW-with-U-FL-Connection.aspx) will allow me to send/receive data, of this data I will need to be able to send/receive GPS way-points. In the future I will be using Sharp IR Distance Sensors (http://www.trossenrobotics.com/sharp-ir-distance-sensor-gp2d12.aspx), and Devantech CMPS03 Electronic Compass (http://www.roboticsconnection.com/p-59-cmps03-electronic-compass.aspx)...

I would appreciate any help of any kind. Thanks in advance!

01-20-2010, 09:55 PM
The serializer's serial port would be taken up by the xbee, and the GPS would generally be serial.

If you get an I2C module GPS, you could use it to feed the serializer.

01-20-2010, 11:11 PM
I am having difficulty finding a I2C module GPS, is there a store or particular name I should look at to search for? Do you have any knowledge of a tutorial implementing this sort of sensor on the Serializer WL board?

01-21-2010, 03:25 AM
There really wouldn't be a tutorial. It would depend entirely on your device. You'd have to send it the wakeup from the computer driving the serializer, then you'd need to query the raw data back like a sensor, then wait awhile to get a whole packet, then reassemble it.

It's really not the right tool for that job, IMHO.

The ublox LEA-5S has i2c, I believe. As well as serial and USB. I don't think it's NMEA though, it's proprietary (protocol is available though). You're in for a lot of work for no benefit. Chipset here: http://www.u-blox.com/en/gps-modules/pvt-modules/lea-5s.html

I believe sparkfun has a board for it, but I'm having troubles with sparkfun.com again tonight.

Do you really have your heart set on the Serializer - and if so - why?

01-21-2010, 11:20 AM
I am not set on the Serializer, I just assumed it would be one of the easiest platforms to start off with. I essentially need a easy platform capable of using a generic programming language. I am most familiar with Matlab, which can compile in C# (I believe). I need to have versatility with this platform with multiple I/O, I am not familiar with micro-controllers so I apologize. This is going to be for implementing artificial intelligence, and eventually will be used for UAV research. Most importantly I will need encoders, IR, sonar, IMU, gryos, motion senors, cameras, in other words the works. Thanks! What would you suggest?

01-21-2010, 03:00 PM
For UAV research you're going to need processing power on the 'bot as well as at the ground station.

The Serializer is running a PIC chip, but is not programmable. So you're going to run into a case where your timings are non-determinant. I believe you're going to find that you need to start looking in to larger microcontrollers.

Matlab can generate C code. Do a help mcc . The downside is that it's pretty over-optimized and tough to read and port. But if you know C you'll manage. Most any decent microcontroller has a C compiler port.

If I were building a UAV, I'd probably be looking to the Axon (or other large Atmel processor) or the XMOS for serious DSP work, realtime AI, etc.

The Serializer is fine for what it is, but I think you'll be disappointed trying to get a UAV in the air reliably with all your processing power offboard.

The Axon has four hardware uarts (serial ports). The XMOS could, in theory, do a dozen or more software serial ports and still sample all the I/O you're listing.

When it comes to cameras, none of these devices are suitable. You really need to start looking at how to reliably get a video signal from the air to the ground. If you're using it just for human interface/visual, then there are a lot of options. If you're wanting flying target recognition, you're probably going to want to put something like a PicoITX in the air, interfaced to something like the other two processors I listed.

But then we have to start thinking about how much battery power we can realistically lift. And you haven't given us any ideas as to scale, wingspan, lift, loiter time, etc.

01-21-2010, 04:20 PM
Well I can't really discuss anything further really about the UAV research. These are completely separate project at this time with no intention of communication between UAV & robot, (does sound pretty cool though). If you would extrapolate a little more about using a camera as human interface/visual that would be great.

For this robot I was assuming a camera would open up a greater versatility/adaptability.... The XMOS/Axon boards you are speaking of look much more applicable to what I am looking to do, thanks for the information. Would you recommend at certain board specifically?

I feel at this time for my given budget the Stinger/Traxster will most likely be our platform. I am mainly looking to be able to take all of the sensor data from the robot and send it back a PC, and be able to control the robot from the PC. I am dealing with $500 for everything, just FYI.