How Can I Transmit Stereo Audio Signals Using a Laser?

[n00bhaus3r]

I came up with this idea purely because I had some free time over the weekend. Basically, I want to transmit audio signals (i.e. from a standard 1/8" audio jack) by laser (in essence, streaming it by laser to a speaker setup somewhere else). Most of my interest are not on the circuitry side of things, so I'd like some advice on how to go about doing it.

I decided to start simple by AM modulating the laser, by connecting a mono input to a 8-1000ohm audio transformer (connecting it to the power supply [2x11.5V AA batteries in this case] and then to the power terminals of a modified run-of-the-mill 5mW laser pointer. For the receiver, I just used a Cd-S photocell connected to one 1.5V AA battery (because the headphone jack RMS is 1.5V if not a bit under), which. of course, is connected to the output mono jack. This setup is quite cheap and comes with a few obvious problems:

(a) Environmental conditions such as humidity, precipitation, etc. degrades the intensity which has a direct impact on output levels.
(b) The setup only works for mono, 1-channel sound.

I was wondering if any of you could offer any guidance on cheap-to-implement modulation methods that could allow stereo sound in two channels to be sent (solving (b)). Also, is there any way to normalize the signal, so that the effects of factors mentioned in (a) would be minimized?

Out of curiosity, is there any kind of documentation that gives the intensity to resistance relation of Cadmium-Sulfide photocells? I was thinking of using higher intensity lasers because the ones advertised on eBay are garbage. I ordered a couple of 532nm (green) 50mW lasers for another idea of mine, and their measured intensities were around 20-30mW with roughly 20-40% of the light emitted being IR (1064nm).

Anyhow, any help in improving my idea would be appreciated.

 

[mgb_phys]

You could use FM if the laser has enough bandwidth.
You could modulate some other component eg polarisation that isn't effected in the same way as total power.
Or you could encode the audio digitall - PWM might be easiest.

ps. Class III Lasers (ie over around 5mw) are now heavily restricted in the USA and you certainly don't want to be using 50mw invisible beams for free space links.

 

[n00bhaus3r]

I briefly looked into digital modulation methods, including PWM, but I ended up deciding that ADC and the DAC chips would be too much trouble. You're welcome to prove me wrong.

I ruled FM out immediately because I don't think the quality of lasers on eBay are particularly reliable, as mentioned in my previous post. A "50mW" laser is around 30 bucks on eBay, but from a reliable lab equipment outlet, they run $100+.

I'm not sure how I could modulate some other component like polarization, some more details would be nice.

Keep in mind that I'm trying to keep the implementation costs down. Students like me don't have much money to begin with. :P

(Honestly, even quality 5mW lasers with good beam coherence can be more expensive than the "50mW")

 

[mgb_phys]

If you only want telephone quality then an8bit adc at 8khz would be enough - you can do this with a very simple PIC micro then you can transmit the data digitally however you want.
You aren't going to be able to send a purely AM signal more than a very short distance in ideal conditions. Look at using a telescope (or just a simple lens) at each end to produce a wider collimated beam.
It's called free space optical communications if you want to hit google.

ps You do not want to be doing this with more than class 1 (<5mW) laser pointers.

 

[n00bhaus3r]

Standard audio CDs do 44KHz/higher rates. Is it naive of me to think that it can be done cheaply at a reasonably higher quality?

Free space optical comms., as I understand it, use really bright LEDs, or in the case of IrDA, infrared beams. Are there any existing laser implementations of FSO systems?

It is my understanding that class I-IIIa cover up to 5mW. Is there any particular reason, barring safety issues, that a higher intensity laser should not be used?

 

[Averagesupernova]

I've done this before. No need for ADCs or DACs. Go with PWM using a triangle wave generator and your audio source into a comparator. The detector cannot be a CDS photo cell since its response time is in 10s of milliseconds. Use a phototransistor and just low pass filter the output to recover the audio.

 

[n00bhaus3r]

Ah, Averagesupernova, that's a good approach. Now I have something to do over the weekend. Haha.

PWM through a comparator is still a one-channel modulation, albeit a better approach than my original AM modulation. Is there any way we can extend that to two channels?

 

[Pumblechook]

IR linked headphones used to be common. Still got some here. ..including stereo. Most seem to use RF now.

 

[n00bhaus3r]

IR is a method, but the issue of modulation (and the stuff I listed previously) are still important. Will digitizing the signal and then using IQ modulation work/be realistic in terms of my design objectives?

 

[Pumblechook]

I don't understand the point of the exercise (as in many of these threads). Sounds like people want re invent something that already exists. How does a lowly TV remote control work?

How far are you trying to transmit and for what purpose?

 

[n00bhaus3r]

Well, for starters, it's not an "exercise." It's just some idle musing on my part.

I'm trying to go building to building and maybe extend the implementation to go about a mile. Purpose? Just for fun.

 

[Averagesupernova]

I don't understand the point of the exercise (as in many of these threads). Sounds like people want re invent something that already exists. How does a lowly TV remote control work?

How far are you trying to transmit and for what purpose?

Oh puleeeez. Just about everything that happens on this forum is a reinvention of something that already exists (as you have hinted). Sometimes 'reinventing the wheel' plants a similar idea in ones head as how to adapt the idea into something else. There is still nothing that beats hands-on, doing-it-yourself, real-world-experience to get a complete understanding of something. If noob wants to do this I will be glad to put my 2 cents in.

 

[n00bhaus3r]

Thanks, supernova. Do you have any circuit diagrams of your PWM design?

I figure that since my idea is already some idle musing, I might as well put some more effort into it. I'm considering a stereo implementation using IQ modulation, and I'm also reading up on AGC gain control circuitry for the receiver. Is this a good idea/feasible?

 

[Averagesupernova]

I don't have a diagram for it. Not sure why you want AGC in the receiver. The beauty of PWM is that a weak signal will not cause amplitude drop in the audio. The received signal quality will be similar to FM as when the signal strength drops (beam attenuation) the noise floor will come up. If you want an AGC I would put it in the transmitter. Why send a weak signal down the beam just to have to amplify it at the receiver along with noise encountered along the way? Have a good method of squaring up the received signal with enough hysterisis to help with noise imunity. Put some serious thought into how to avoid problems with ambient light (contstant) falling on the sensor, things of this nature. My circuit consisted of mostly op-amps and probably a transistor to drive the IR LED. I did not build it up from an existing schematic. I pulled it out of my head and most likely tossed the schematic that I probably scribbled on a notepad. If it seems a bit much to tackle, start with an optocoupler and the PWM circuits. When you get that working, switch from the optocoupler to an LED and phototransistor for some distance.

 

[russ_watters]

I don't understand the point of the exercise (as in many of these threads).

Are you an engineer? Engineers often do thing just to prove they can. Ever seen the show "Home Improvement"? I once killed three $100 webcams trying to modify them for long exposure before buying a $300 astro-cam...

 

[famousken]

To transmit in stereo, you need what is called a multiplexer for your transmitter and a demultiplexer for your receiver. These are the same chips used in wireless headphones. they are still analogue and they are designed to be used with a small number of additional components. I would then take your multiplexed signal and feed it to either a vco for fm, or into a pwm controller for pulse width modulation. You can find the pwm chips in most switchmode power supplies, just feed the multiplexed signal into the duty cycle input with a pot so you can adjust the gain.

 

[n00bhaus3r]

@famousken. I'm not familiar with analog multiplexers. Perhaps you could elaborate?

@Averagesupernova. Interesting. I shall try it with the optocouplers first. Thanks!

 

[famousken]

Here is the datasheet for one such ic, http://www.datasheetcatalog.com/datasheets_pdf/N/J/M/2/NJM2035.shtml

It is a pretty simple device, basically what it does is take one channel (usually right) which is left as is, except the second channel is superimposed on it via an internal chopper at a rate of 38khz. each channel is actually broadcasted 50% of the time. think of it as switching between left an right channels vary quickly. This signal is then added to a carrier signal (you don't need one) for fm transmission. This method has the added benefit of being backwards compatible with older mono fm receivers because they simply do not respond to the 38khz modulation and only respond to the non-modulated channel (R). For your receiver, I believe the easiest thing to do would be to use a pre-manufactured stereo receiver, such as in a boom box, take it a part and locate the demultiplexer, look up the datasheet for it, and then disconnect the input from the RF receiver and just tie into the appropriate pin on the Ic. Hope this helps!

 

[Averagesupernova]

Hold it! Stop the misinformation famousken. The standard FM broadcast band uses a multiplex scheme that goes as follows:
-
The left and right channels are added in a summing circuit. This signal is known as L+R. Then, the left and the inverted right channel are added in a summing circuit. This signal is known as L-R. The L-R signal is then applied to a balanced AM modulator with a suppressed carrier frequency of 38 Khz. It has nothing to do with any chopper circuit. This signal is the subcarrier signal. The subcarrier and the L+R signal are added in a summing circuit along with what is called a pilot signal which is half the subcarrier frequency which comes out to 19 Khz. This composite signal is applied to the FM modulator to achieve about 75 Khz of deviation. At the reciever the detected FM composite signal is run through a low pass filter to recover the L+R. The composite signal is also run through a bandpass to recover the double sideband suppressed carrier signal (subcarrier) which contains the L-R signal. The 19 Khz pilot is filtered out and doubled with frequency doubler to obtain a 38 Khz signal used to demodulate the subcarrier to recover the L-R signal. The original 19 Khz pilot does not deviate the main carrier very much. Just enough for a phase locked loop in the receiver to be able to recover it. So now we have 2 baseband audio signals. L+R and L-R. They are added in a summing circuit to get 2L (+R and -R cancel). We now have the left channel. To get the right channel we simply invert L-R (turns into -L+R) and add it to the L+R signal to obtain 2R. (-L and +L cancel).
-
noob, if I were you I wouldn't go to the trouble of using the scheme that the FM broadcast band uses. The reason the FM broadcast does it that way was so FM receivers built before stereo which were only able to receive mono would be able to detect L+R with no modification. They had to be compatible. What I would do is look into a rising edge/falling edge scheme to send 2 channels using PWM.

 

[famousken]

are you sure? I am trying to remember info I researched about five years ago, so you could be correct, however I do know that many use a single chip to do the multiplexing for fm broadcast. I am aware of the 19khz pilot, but in he would not need this with his circuit. Here is what I read, http://transmitters.tripod.com/stereo.htm.

 

[Averagesupernova]

I didn't read the link in your first post and somehow I just don't remember it being there when I posted but I'm sure it was. One other purpose of the 19 Khz pilot is to 'tell' the receiver that it is receiving a stereo broadcast. I don't think the stereo decoder ICs will work without the pilot.
-
And yes, I AM sure of what I posted. It works the same with NTSC broadcast TV except the pilot is the horizontal scan frequency of 15734 Hz. I've worked in this industry on video and stereo TV generators and cable TV field strength meters. I am very familiar with the format.

 

[Pumblechook]

The mono part is L+R not just R or L.

The system can be looked at in two ways ..suppressed carrier with L-R at 38 kHz OR simple switching L R L R at 38,000 times a second...single samples. It is the same thing. It is usually generated and decoded by simple switching. The pilot tone is needed to synchronise the switching. The audio is limited to 15 kHz because you have to sample at a frequency which is at least double the max audio frequency.. Nyquist rule.

http://en.wikipedia.org/wiki/Nyquist–Shannon_sampling_theorem

 

[rbj]

I didn't read the link in your first post and somehow I just don't remember it being there when I posted but I'm sure it was. One other purpose of the 19 Khz pilot is to 'tell' the receiver that it is receiving a stereo broadcast. I don't think the stereo decoder ICs will work without the pilot.
-
And yes, I AM sure of what I posted. It works the same with NTSC broadcast TV except the pilot is the horizontal scan frequency of 15734 Hz. I've worked in this industry on video and stereo TV generators and cable TV field strength meters. I am very familiar with the format.

Average, Pumblechook is correct. the FM stereo can be done either way. i remember working this out when i was an undergrad, that synchronized time multiplexing the bandlimited L and R (at a rate of 38 kHz) was the same as L+R at the baseband and L-R heterodyned up to 38 kHz. it's mathematically equivalent. the 19 kHz pilot was to set the clock for this synchronization.

 

[Pumblechook]

The standard stereo system is called Zenith/GE and suffers from poor signal to noise. If you are receiving a weak signal on FM it will be much noisier in stereo. If you have a receiver which will switch to mono you will notice a vast reduction in noise (hiss) on a weak signal.

You could use a standard 10.7 MHz carrier frequency on the laser. There are plenty of chips available, filters and all the coils needed for the quadrature demodulator. You could just modify a radio.. feed in at the intermediate frequency.

Building a stereo coder might be a bit more tricky.

The other way to get stereo analogue is to use two separate subcarriers. The old analogue satellite TV used various subcarriers for sound... 7.02 and 7.20 MHz being the most common. Chips for 10.7 will work at 7 MHz but maybe 10.5 and 10.9 could be used say. Depends how easy it would be modulate a laser at these frequncies. You could go lower to the 1 MHz region.