Distance in Tracking System

I'm attempting to build a tracking system that will consist of a robot containing a receiver that will follow and track a transmitter that will be attached to a mobile user. I want the robot to maintain a 2 meter distance from the user so if the user accelerates, the robot accelerates, or if the user decelerates, the robot decelerates. I'm wondering which type of sensor I should use to measure the distance between receiver and the transmitter. I've done some research on rf, infrared, and sonar sensors, however, I'm not exactly sure which would be best for this application. Would it be possible to measure the distance with just a transmitter and receiver or would I need transceivers?

Distance Finding With RF

I am currently working on just such a project with my son. We are using the Linx ES TX/RX Modules. Very easy to use just power and an antenna basically. The receiver module has a built-in RSSI signal strength that varies from about 1 volt background noise to about 2.730 vdc at 3ft. The signal at 70 feet is about 1.4vdc. We are going to use two receivers with directional waveguide type baffles on the front of the robot to enable it to seek out the transmitter. We are currenlty testing antenna baffle designs for the best rejection. A simple grounded foil baffle has given us about a 50% rejection in received signal from the side. So we have a focused antenna for the the left and right side of front/center. Our drive controller pack has analog inputs available so the calibration and controll will be in software rather than descrete components. The ES series uses 916Mhz and the antenna length is about 3.25 inches. So, the baffle does not have to be huge. That's all I can tell you now but it looks good so far. Hope this helps.

Thanks for info, it helped alot. I'm going to look into those Rx/Tx modules with the RSSI ouput. I was thinking of determining the orientation of the robot by using a Dinsmore compass. I was going to incorporate the compass into the transmitter and then send a PCM signal containing the compass heading to the receiver every couple seconds. At first I was thinking about using the Dinsmore 1490, which is a digital compass, however, it only has a 45 degree accuracy which probably isn't sufficient for a tracking system. The Dinsmore 1525 is analog compass that is accurate to about 2 degrees so I'll probably end up using this instead. How exactly does a directional antenna work? I wasn't even aware that there was such a thing. Do you think it would be easier to use in a tracking system such as this than a compass?

I have to admit that I am not familiar with the Dinsmore compass. My background for the last 20 plus years has been in analytical and nuclear electronic circuits. I am new to RF transmission projects myself. This project came up with my son at school and he asked for help from dad. I worked through several internet projects but with no luck. I am several months into trying to find a solution for our direction finding robot. I am confident that I am now on the right track. However, the most interesting thing about electronics to is learning about different sensors and how they function. So, tell me about how this works. Is it magnetometer based? I used to work on an analyzer in the oilfield that incorporated magnetometers and accelerators to give 3d position anywhere in the world. ( the calibration was a bugger but it was before GPS and was all that was available) Besides, it was designed for subsurface detection where GPS would not have helped much anyway.
Either way, on with it. The antenna on the transmitter is effectively vertical for maximum received signal the receiver also needs to be vertical. If you orient the receiver antenna horizontally you will loose distance because of the orientation but you will make the receive area directly in front and to the rear of the antenna better than at the sides. (The modules work for several hundred feet and I am only concerned with 100 foot.)Since the frequency is fairly high, the antenna ¼ wave is just 3.22 inches and shielding the antenna by at least 105% of it’s ¼ wav will further decrease reception. Put it in a soup can so to say and ground the soup can. It is effectively like cupping your ear to hear. Our idea is based on 3 antennas with approximately 120 degrees coverage each.(2 looking forward and 1 rear) There will be overlap but after a simple distance calibration of the RSSI signal of each receiver we will develop our plot data for the controller. Directly in front of the antenna we see about a 1000mv change over about 70 feet. From the side we see about 500mv over the same distance. So, our resolution is going to be a few mv/foot. At any given time two of the receivers (after calibration corrections) will be higher than the third. The controller will rotate to fix position in front of it and then advance. Since the signal currently saturates at about 2.5 feet we will have to keep about 3-4 feet away (but that was our original goal). The Linx module is a single chip solution for this application. You can use encoder/decoder chips to send addressable data additionally but I just needed the RSSI for my project.
Let me know what you find out on the compass data.

I can't really tell you much about the Dinsmore compass since I have just recently started considering them. They seem pretty popular so if you were to do a search on google for "Dinsmore 1525 compass" it should return many results.

Here are a couple sites that explain about the compass:
http://www.dinsmoresensors.com/
**broken link removed**

"The Dinsmore compass works by using four hall-effect sensors. There is a subminiature rotor jewel in combination with these sensors. The compass can efficiently detect the weak magnetic field of Earth and therefore give an accurate direction back to the host computer."

What would be the advantage of using encoder/decoder chips to send data over other chips? I noticed that you mentioned that your tracking sytem will be dealing with a distance of 100ft. The distance that I will be trying to maintain is much shorter (approx. 6ft). Do you know if the RSSI method of determining distance would be accurate enough for 6ft? You mentioned that the signal difference between 3 ft and 70ft was just over 1V.

I'm pretty new to electronics and i'm doing this as a school project also. Anyways, I'll try the compass method and if it doesn't work I may have to give the directional antennas a shot.

By the way, do you know of any good antenna designs that I could use to send a PCM signal containing the compass heading from the transmitter to the receiver?

Hello Again. I missed your reply, just saw it this morning. I can't help you with the antenna design. It is my weak spot also. The RSSI is effective at 6ft. As you near about 3 feet the receiver RSSI saturates at 2.730VDC. as soon as you back off it starts going down again. It was just perfect for my project. However, I am counting on "finding" direction at a distance and then just keeping it in my sight. I hope to follow a moving transmitter at a range of 4-5 feet once we catch up to it. The first goal is just to find it in a fixed spot and go to it.
The ES series is a digital/analog repeater. It works with audio and digital signals. An encoder/decoder would make it addressable and able to read data if you needed that function. In a remote control situation you could encode several command lines and then decode them to relays or drive circuits. You can also use a Max232 on each end directly into the ES chipset and computer control with it. You can send files or just control data. They are really functional and use few discrete external components.
Since this is overkill for my project I didn't need this functionality. To Linx, the RSSI signal was to help provide for a squelch circuit. Although it works really well for distance, there is no calibration of the signal and it differs from one chipset to another. The VCC is 3VDC so I expect the saturation is fairly consistent but the Voltage/distance would vary by several factors including antenna design. The RSSI as an indication of field strength is usable for distance and is repeatable but you wouldn't want to design and sell products around it. It works fine for home projects.
I do plan on starting my own project after we complete this one. I want to build a transceiver using the Max232 and RX/TX on each side so that I can get feedback from my sensors. Hopefully, Video as well.
There are several new devices to help take the bite out of the design work that were not there 20 years ago when I used to play with projects like these.

Just the other day, I called Linx to order a few transmitters and receivers from their ES series, however, they told me that they were backordered for about 2 months. Would you happen to know of any other places where I could order these modules? If not, do you of any similar chips with the rssi capability, that I could use for my tracking system instead? Currently, I'm looking into the Atmel at86rf211 transceiver, although, it seems like overkill for my project and it is much more complex due to the external circuitry required for it. Btw, how is your son's tracking system coming along? Thanks in advance.


Graham