DTMF-controlled robot design
Now that I've graduated (back on May 20th), I have had loads of time to work on various electronics projects. One of which is a small remotely-operated robot. It's in the design phase right now, but I've decided to post what I've come up with so far.
For signalling, it will use dual-tone multi-frequency (DTMF) tones. These are the tones that you hear when you press keys on a telephone keypad. The reason for using this standard is that you can buy a single, very accurate IC (usually used in answsering machines and the like) that takes care of all the filtering, amplifying and interpreting necessary to determine which key was pressed. And because it's just an analog audio signal that is being exchanged, I can use a number of ordinary, low-cost transmitter/reciever pairs for sending that audio signal.
The design of the circuit for decoding DTMF tones is pretty straightforward. The CM8870 IC interprets the tones as a 4-bit digital signal. The HCF4514 IC then decodes this into 16-lanes of output which will switch reed relays, which will switch automotive relays, which will operate motors that move the robot.
| DTMF (dual-tone multi-frequency) signals | ||||
|---|---|---|---|---|
| 1209 Hz | 1336 Hz | 1477 Hz | 1633 Hz | |
| 697 Hz | 1 | 2 | 3 | A |
| 770 Hz | 4 | 5 | 6 | B |
| 852 Hz | 7 | 8 | 9 | C |
| 941 Hz | * | 0 | # | D |
When all put together, I will send the DTMF tones over an old 900MHz cordless phone, providing a range of several hundred feet. Because of the modular nature of this system, in the future I could swap out the cordless phone for a newer one, or a pair of FRS radios, or some other transmitter/reciever pair. It doesn't really matter.
So then based on which key is pressed, the robot will be able to interpret the tone and activate relays and such to turn on motors and move about.
The 16-lane output will drive a set of reed relays, which will in turn drive a set of much larger automotive relays, to switch a pair of drill motors, as well as motors that will rotate the camera-equipped head left and right.
Speed will be controlled in steps using a few high-load resistors taken from the drills. Three speed settings should be plenty.
There will also be a couple buttons left over for arms and stuff that I might add later....maybe to toggle an series of IR diodes for night-vision...or to discharge a linear magnetic accelerator....who knows?
The robot itself will basically be a box (made of either wood or sheet metal, depending on how the funds work out.) It should measure roughly 1 foot by 1.5 feet.
At the center of mass will sit a 12-volt SLA (sealed lead acid) battery (the kind used in uninterruptable power sources and electric wheelchairs). Two 12-volt drill motors will be connected with bicycle chains and sprockets to four wheels, providing 4WD mobility for the unit. I have no experience in making chained drivetrains, so I'll learn as I go.
Power for the robot will all come from the 12-volt SLA battery. One regulated DC-to-DC converter will provide 9-volt power to the telephone reciever and to the wireless camera. Another regulated DC-to-DC converter will provide 6-volt power to the logic circuits. This power will then go through an L78M05CV 5-volt regulator to get the proper voltage to the sensitive ICs.
As I mentioned eariler, atop the robot will be mounted a wireless camera that will be rotatable left-to-right with the use of a motor. This will provide a remote view of the robot's surroundings, and should transmit a good 100-200 feet. And I may add a bank of IR LEDs to give it some nightvision. That would be sweet.
So that's the plan. I've ordered most of the electronics. Once I get that part of it working, I'll start working on the chassis and drivetrain. Should be fun. I'll keep you posted...
Printer-friendly version
Pretty interesting stuff! Good luck with building it.
Pretty cool, i must admit that much. but wood? you realize that wood will easily break when it's supposed to run around on wheels and stuff like that (i learned the hard way). especially not the chassis.
where did you order the parts? I was planning to also build at least a chassis (or maybe a BUZZ 12 recreation) for the past few months, but now that summer vacation is here and i have time off to actually build it, i have been checking sites for parts, yet I can't seem to find any with suitable prices. I've actually made a chassis before, yet I had to take it apart for parts for my comprinter (don't ask).
just a random question- why didn't you join robotics? you would've been able to get ideas from there.
about the phone- try a joystick or R/C controller. unless you're really good at text messaging or something like that. It will most likely take a lot of patience to go right, left, etc. good luck and have fun (i know i will- i can't wait til i have time off for summer).
Wood will be just fine. It's not going to be ramming into stuff, just driving around. It's worked very well on my past robots. Besides, I can get it for free and it's easier to work with.
I was in OCCRA. Didn't join FIRST due to time constraints.
The real beauty of this design is that because its driven by DTMF tones, I can use any interface I want. There are several inexpensive DTMF generator ICs out there, or alternatively any computer can be easily programmed to send DTMF tones via line-out to a transmitter. So switching to another kind of controller would be really easy.
But for now, using a phone as the actual controller is by far the easiest implimentation.
Uhm, it has nothing to do with text messaging.
As it is right now, I would press 2 to go forward, 5 to go back, 4 to go left and 6 to go right, like in the picture. Very simple, much like the arrow keys on your keyboard. Plus, it's quite portable, and the range is very decent (~200 feet). And it has room to grow. At this point, I see no need to change it.
yeah... well, it depends what wood it is. IF you do it too thick, it will either a) be hard turning or more work to turn or b) not even turn at all.
And yes, I agree that DTMF does have good frequency. I've worked with it a lot and it's easy to use on literally any electronic you choose.
But when did you build more robots? this is the first time that i've heard that you have.
about the text messaging. i was using that as an example, as in you have to be quick about it, especially depending on what speed you plan on using for your motor. You would
Uh, no. The material that the chassis is made with will have no influence on turning capability. How well it turns is all related to the configuration of the motors, the kind of wheels used, etc. If it works as planned, it should be able to rotate one set of wheels forward, while rotating one set of wheels backward, yielding a turn radius of essentially zero.
What on earth do you mean by "DTMF does have good frequency"? That just makes no sense.
I built one last year that utilized an optoisolated relay controller interface that I designed to work over WiFi. It worked really well. It was just bulky. I might make a post about it sometime.
I tested out the interface in general the other day and it's very accurate and responsive. I don't anticipate any problems.
Hey Mark I just wanted to inform you that Im pretty sure your anonymous commenter is none other than Woody.
I doubt it. Woody actually knows what he's talking about.
Nope, not me. I rarely even come here, just happened to poke around recently. Best of luck Mark, sounds interesting.
Just read your write-up again, Mark. For your drivetrain, you may need to consider an over/under chain loop for your drive motors, like we use in OCCRA. Since most drill motors are wound to output slightly less power in reverse, you need to run them both in forward and use the mechanics of the chain drive to invert the output of one side... If you don't the robot will most likely travel in a large arc rather than a straight line. Contact me if you need more details, it's difficult to put into words.
how does it stop? I see change to full speed, 1/2 speed, etc. But where is the "stop!!!" button?
just something that you might want to consider before you fire this thing up... otherwise, it looks very cool.
You could even use a cell phone to control it. idk how good the cell reception is in your area, however.
It only keeps the motors running for however long the buttons are pressed.
So when you hit "2" to go forward, it will keep going until you let off the two.
The full, half, quarter speed selectors toggle resistors in between the motors and the battery, changing the overall speed for when the buttons are pressed.
You are doing a great job....
I just have a question: How can the output of DTMF turn on the motor?How do you interpret the output of DTMF so the output will command the electric motor to start running?
This image explains it: http://markbowers.org/cms/files/Image/robot/signalflowchart.png
The DTMF decoder outputs a 4-bit number. That goes into a 4-bit decoder, and relays are hooked up to the outputs of that.
Hi Sir
I really really liked your project, but the coincindeince that I'm making something really similir to yours,anyway, just wanted to advice you something, do as I did,replace the receiver with a cheap mobile phone like N3310 and call it from an individual transmitter and you'd get 10000 Km range right?,even I will be able to control it from here!.
but again,FANTASTIC.
Hi Mark
Nice project. I'm needing to do something similair but for a theatrical cue light system - several outstations with "standby" and "go" LED's on them. These can all be connected onto the same line by one XLR cable (standard in theatre audio). The idea is that a stage manager can send silent, visual cues. It's quite a common system but i'm looking at building my own.
The reason for using DTMF rather than digital switching (RS232 and the like) is that i'm keen for this to be able to run within a standard audio multicore. I've had problems sending digital data down these due to induced noise and audio should be a lot better (it's designed to have audio on adjacent cables within the multicore).
In my system, the outputs from the HF4514 will feed into a set of dipswitches where the outstation address is set. The first two outputs being for the first outstation, the second for the second, and so on. The dip switches will connect the outputs to the correct relays.
To cut to the quick, do you have any ideas on where to source the necessary IC's? I'm in the UK, not sure where you are? Nobody seems to stock them.
Many thanks
C