Is this radio transmission feasible?

[Dingo]

I was thinking of a novel (to me at least) way of playing back mp3 (or ogg etc) audio via one's robot.

If the robot is in communication with a PC (via UHF link etc) couldn't it be possible to build an FM (or AM I suppose) receiver into your robot and tune a FM (or AM) sender that is connected to your PC.

That way the robot can inform the PC which sound file it wants played back and then the PC transmits (via a FM bug type transmitter) that sound to the receiver on the robot and in effect the robot is playing back its sound but without having to decode signals etc.

You could also use hacked walkie-talkies (although proper ones might cost too much)

Any thoughts on the feasibility of doing this?

[Bones]

I was thinking the same Idea. I was looking at voice reconition with Dragon dictate or something for commands using a fm wireless head set going back to the pc to the mic and a seperate channel for voice as you said to transmit back to the robot.
I have a set of the data transmitter chips at home that I talking about in one of the other forums. They have signal strength, error checking, buffer and all the clear to send that serial ports use etc. They have a small microcontroller build in to do the job. Very nice
There are heaps of different types. there are two frequencys 400 and 900 Mhz and you can get a receiver and transmiter for one way communication and transcievers for two way comms.
The units are called easyradio. have a look on the following web site. www.lprs.co.uk for specs.
And there is an Australian distributer in QLD.
www.lprs.com.au/ for Australian orders.
The transceiver units are the ER400 and ER900TRS.
The transmitter and receicers ER400RS & TS and ER900RS & TS
Really easy to use.
Bones

[Dingo]

To off load voice recognition wouldn't you need to transfer really large amounts of data?

I wonder if it is possible as even WiFi could struggle to transmit in a timely manner really high quality sample sizes?

I think you'd probably be best to have the Voice Recog PC on board

[Bones]

You would use a radio data module for the control etc and a fm mic for the voice. Fm should have enough frequency range for voice commands.
I was looking at using my pda with wireless for the voice commands to be streamed through to the PC in my robot.
In your case have a laptop as a control unit in the room etc.
Or as you say install a voice recon board.
Bones

[Dingo]

oh yeah - didn't think of transmitting the voice without digitising first. That is a good idea - saves a lot of hassle

For my consideration I just want voice out of my robot (nothing as fancy as voice recognition is reqired but I do have Dragon dictate on my PC and my play with that later)

[ZapBrannigan]

Bones said:

There are heaps of different types. there are two frequencys 400 and 900 Mhz and you can get a receiver and transmiter for one way communication and transcievers for two way comms.

You had better be careful with those frequencies.... One of them is the GSM network's frequency and the other is the frequency of the Mobitex Network (what the Tyco wireless fire alarms work on). I can't remember which was which as one is the Aussie version and the Yanks have those two reversed. We had people come scanning our company when we were ( illegally ) broadcasting a signal on the GSM frequency when we were doing devices for the US mobitex network.

[Bones]

Yeh everyone uses those frequencys cause they are the free to air bands that you don't need a licence for as long as you don't go over 1 watt.
Same as the 2.4GHZ used in wireless networks and mobile phones. Unfortunately that is the same frequency that microwaves use. Cause it happens to be the frequency that water molucules vibrate at that produces friction.

[Dingo]

I've purchased a couple of 433MHz transceivers. I believe these are ISM and that is legal upto a watt without license.

Oh yeah - 2.4GHz is getting so crowded with all the WiFi and microwave ovens and almost any domestic wireless devices these days.

I think they should release a couple more narrow ranges because of the explosion in wireless everything these days.

BTW I read somewhere the Human brain's internal clock is 16MHz (or 16KHz) and if too much energy is directed at someone around that range they'll have an epileptic fit - anyone know anything about that?

[ZapBrannigan]

Out of interests sake, what kind of range are you expecting to get out a watt or less of power? How sensitive will your receiver be and are you going to direct the signal or have it radiate in a spere?

[Dingo]

Well my needs will only be one or two indoor rooms (with gyprock walls not bricks) so I won't need more than 10 or 15 meters.

They come with a basic copper spiral antenna (just a length of wire coiled comeing from the pcb. They did mention something about pcb or ic type antennas - didn't know what they were on about.

I think they claim if you run the receiver on 12v you can get over a k LOS. I also take such claims with a grain of salt. I know WiFi cards can do 300metres LOS (in fine print optimal conditions, which means not likely unless you add a highly directional antenna in the mix)

I've heard of some locals here in brisbane getting several kilometres with a one watt WiFi using a waveguide type antenna (I think it was waveguide).

Power consumed (ie runtime) will be a bigger factor for me than distance from the PC. Thus I'll only run the minimum power to have reliable operation. (The devices operate from 3v to 12v I think)

So yeah I'll leave the antenna configuration as omni-directional as possible (because I don't know where the robot will be in relation to the PC).

As I perceive my distance requirements not very steep I haven't looked at Signal to Noise etc. (I think this is the class of devices used for automatic garage doors)

Zap you'd obviously have more idea than me - do you think distances of 10 - 15 metres will be a problem?

[Dingo]

got the partds from Oatley this morning but too busy at the moment to play around with them.

Will advise how much distance when I set them up

[Bones]

Not a problem,
It's hard to search for stuff. It comes down to the wording and it is when you are looking for something else you will find exactly what you wanted a month ago.
Or it will turn up in a mag winin a month. Murphys law.
I ordered some batterys from Oatley, waiting on them. 2x12V 12AH for my robot just to have a play with proberly need another 4.
And got some PIC-AXE's today. Are they surpose to say anything about PIC-AXE on them. Cause these don't. Got them from Altronics. They were nicely wrapped up in a static foam and put in a static bag. NOT. No static foam and chucked in a normal zip lock bag. Lets see if they work. I got the 1 of each except the 40. They look good.
Bones

[Dingoatwork]

I've onlyu purchased 18x and 8m types - they had a gold X and a silver M respectively - no other picaxe specific marks

[Dingoatwork]

BTW - package of picaxe from MircoZed is great - just in tube and bags - but for project boards they are a bit iffy - ie too little packaging and bent MOSFETs etc

[ZapBrannigan]

I think that you'll easily get 10-15 metres if the specs talk about a k as being the max distance Spherical transmission (the easiest to consider from an antenna) radiates power as the area of the sphere->
POWER / (4 pi r^2).
So doubling the distance reduces power at the receiver to at least a quarter, and so on. Hence, to reduce your reception distance to 1/100 of the max, you'll have a signal 10000 times larger, and thus can use far less power to create a receivable signal.


Will the radio receiver will just be a straight to speaker effort?

What medium will you use to transmit the request signal to the pc, as I don't think you'll be clocking your robots CPU up fast enough to put data onto a 400 or 800 MHz signal