How 405-line television could be broadcast licence-free

What's the 3.5MHz bandwidth power source?

You'd need an optical quality dish or very large lens unless you going to fry pigeons.

Strictly Line Of Sight.

Renting some space on a Satellite Transponder and using maybe 10MHz bandwidth FM might be easier and cheaper per person. Uplink outside UK and no UK licence needed.

If it's not being broadcast at VHF I don't see the point. Internet streaming would be as valid as satellite broadcasting, and probably cheaper and easier for all concerned.

I had a maplin infra red video link kit back in the day. it worked ok for a few tens of meters ...

Thanks for the comments

What's the 3.5MHz bandwidth power source?

This isn't baseband... a full Ch1 carrier bandwidth is to be modulated onto the deep red light here... so say a bandwidth from 41 to 48 MHz.

If it's not being broadcast at VHF I don't see the point. Internet streaming would be as valid as satellite broadcasting, and probably cheaper and easier for all concerned.

The point is, it is being broadcast at VHF as far as the aerial downlead and set is concerned. But on the way to its destination, this 'VHF' is impressed on a light signal, not directly on the radio spectrum - so gets round the radio spectrum licensing issues.

In this way, all you need at the receiving end is a photodiode and pickup lens.. .no clever RF conversion boxes. Ergo... out of the photodiode an 'instant' authentic Channel 1 signal emerges. In some cases the lens and photodiode could even be hidden in an 'H aerial for still greater authenticity.

This got me thinking.. how much power would be required at the 'transmitter' - and how much aperture at the receiver - to get say ½ millivolt into 75 ohms at the photo-diode pickup? It turns out that you don't need to roast pigeons. (Arbitarily) assuming an overall system efficiency of 10%, at a range of 5Km with an omnidirectional beam of 100m thickness (height) at that distance, you could do it with 20 watts at the transmitter and a pickup lens of 8cm diameter. I can PM the calculation to anyone who's interested.

Of more concern would be obtaining sufficient bandwidth out of the emitters (perhaps a ring of line-pattern modulated lasers) and the signal-to-noise out of the detectors.

Steve

I'll have a look at the sums. I'm more used to terrestrial 10.2GHz downlinks and 10.65 GHz uplinks. I have a system running about 13km. Aperture is about 20 x 10 cm per array at receiver for 10km, which isn't enough.
So I have a 44cm dish. Bandwidth of signal is 8 MHz and it's QAM64

Output TX about 1W, but that's about a 60 degree sector and massive array to reduce the vertical beam width about 5 degrees. The Receiver is band pass filtered at 10.2GHz with two stages gain before mixer, then 485MHz IF amp.

The photo sensor has poor output though. I'll check some good ones.

For source you'd use an array of IR LEDs. Even 800 of them is probably under £25 inc postage. Much simpler to modulate than lasers.

Being that this is Ireland all my TVs work down to 42MHz, so I can test the idea at 625 Lines. I have some high power amps that will do 42MHz and loads of IR LEDs and good photo transistors. I'll use mini-circuits 1MHz to 800MHz ring diode mixer to downconvert the UHF modulator to 44MHz approx. I have some 44MHz IF filters (for cable Modems) if I need gain and filtering.

But I still think you'd need parabolic mirrors or big lenses to get that coverage map. Or 2kW. I'll need to figure photo detector gain. Also for lasers or LED, the RF power you need to drive is going to be a lot more than optical power!

RF Parameters
Downstream
Frequency		Power	-3 dBmV
Signal to Noise Ratio	27 dB	Modulation	QAM64
Upstream
Frequency		Power	50 dBmV
Upstream Data Rate	5120 Ksym/sec	Modulation	QPSK

Status
System uptime	64 days 13h:34m:45s
Computers detected	1
CM Status	Operational
WAN Isolation	OFF
Time and Date	Fri Dec 05 02:17:27 2014 

Parameters at IF levels, 485MHz downstream and 21MHz upstream on cable modem.

Yes, insanely I have a Cable modem and outdoor full duplex transceiver at 10.2 & 10.65 GHz on a 44cm cassegrain dish

I'm sure there is cheaper
£13.39 inc postage for 500x IR LEDs
http://www.ebay.co.uk/itm/Lot-of-500-X- ... 0592178104
http://www.ebay.co.uk/itm/Lot-of-500-X- ... 0592178030

(Edit they sell 1000 of at £24 inc shipping)

I suspect a ceramic filter and 20dB amp (or more) at the photo transistor would be good. 44MHz is a standard part available in 2, 4, 6 and 8MHz. The 8MHz would be cheapest and easiest to get. That MIGHT be 40MHz to 48MHz, (again I can measure the ones I have), which would suit fine. These filters are for Cable Modems. I guess they originally used USA TV IF as 2nd IF as Cable Labs Inc is USA. The Modem's 1st IF is 1.2GHz.

Same seller has 5W IR LEDs! £44 for 10.
http://www.ebay.co.uk/itm/10pcs-5w-940n ... 0596628201

It would be interesting to know what VB thinks.
Laser modulation across the channel here.

http://www.g0mrf.com/laser5.htm

Eddie

Let's remind ourselves this scheme is a flight of fancy - it would be unfundable if it were just to provide a service for a handful of vintage television enthusiasts. But it's an interesting theoretical exercise nevertheless.

I'll come to Michael's points in a moment. First though, a list of the Pros, Cons, and Unknowns as I see them.

PROS

Requires no radio spectrum licence - regulation expires at 275GHz.

Historically appropriate Channel B1 could be radiated, encapsulated in light.

In principle, simple to receive - ready-to-go RF channel emerges from photo-sensor.

Able to provide a decent service area given a suitable emitter location.

A vintage set could be used as a receiver without modification.

Ability to use historically correct aerials with hidden photo-sensor in some cases.

Omni radiation could be limited above and below main beam, just like a radio antenna.

Optical carrier has massive space for other one-way data applications.

Near infra-red easier to generate and detect than far infra-red or millimetre waves; less penetrating to flesh too.

‘DC’ component of daylight prevented from flooding sensor by filter. Would work in day or night.

Height of pick-up units often not critical.

CONS

Line of sight only - although reception would be possible over a wide area, it would be highly 'granular' in nature.

Badly affected by rain and fog. Use of infrared would reduce this characteristic, particularly in mist.

Would not work when photo-sensor was facing the setting sun!

'Baseband Flutter' would occur on signal in fringe areas due to rising warm air and instability in atmosphere.

UNKNOWNS

Provision of sufficiently fast luxeons or laser array modulator for radiating ring.

Legalities regarding aircraft lights on towers.

Hazard to vision of staring at emitter at close range (unlikely with envisaged radiation pattern except from helicopter).

Bandwidth, sensitivity and noise floor limitations on available photo-sensors. Additional amplification might be required.

Heating aspect of near-infrared light at emitter's working power level.

Innate noisiness of near-infrared band.

Need to ensure adequate vertical thickness of beam form radiating array to cover local topography.

-----------------------------------------------

Michael, thank you for your contribution. You're going to a lot of trouble here!

In general, parabolic mirrors of a given size become more efficient as a signal collector with decreasing wavelength. Especially so, when you consider 1 micron (300 THz) near-infrared light is 30,000 times shorter than 10Ghz. Think of radio and optical astronomy for example. I'm sure my old 10-inch reflector could have captured sources as faint as Jodrell Bank!

So even quite small optical collectors can have a large gain. Think of the gain provided by a 10cm diameter parabolic mirror focussed onto a 1 sq mm photo-diode. The collection area of the mirror would be 7854 sq mm. The gain would thus be 38dB.

Regarding LEDs, I believe the rise times of most of them are barely fast enough for operation at 45MHz. I need to do more work here.

With a line-pattern modulated laser as an emitter, most of the optical work would be done for us already. The catch is the one device I have seen will only modulate upto 200KHz ! I once successfully modulated a laser torch module - they really need to be current-driven (like LEDs).

Finally, thanks Eddie for that snippet. Most interesting!

Steve

Yes, same dish on Ka Band is nearly x4 gain compared to Ku.

LEDs do handle MHz. I suspect the low power types faster. LEDs HAVE been used for 200Mbps "Access" points experimentally in offices (on the ceiling). Low power lasers for fibre of course handle GHz.

IRDA uses LED. Originally only 115k baud, then up to 16Mbps, current versions can do 1Gbps.

I'll see how fast I can drive them. Possibly a DAC with pre-distortion look up table is needed to drive the LEDs due to non-linearity. If so oversampling and about 170MHz clock will simplify filter to drivers.

What form of modulation are you proposing?

Digital signals as used for telephony can be generated using cheap lasers but AM as used for cable TV is another matter entirely.

Unless prices have come down in the last few years you could be looking at around £10,000 a pop - and not much power, either ...

PS
Commercial 400GHz gear exists. Pricey though. Trying to put 625 PAL at 42MHz + over IR will be a bit of fun though.

What form of modulation are you proposing?

Digital would be easier. No, the idea is AM, so "receiver" is just photo transistor, optionally with 44MHz filter and preamp. No demodulation or frequency conversion.

There are a couple of ways to improve overall link linearity, if it's a problem.

Good luck with this Michael! Will await your practical results with interest.

Steve

We had a Birthday party here today, likely the coming week. Hope you and all family keeping well Trevor.

Interesting idea!

I have here some Philips infra-red cordless headphones, which work somewhere in the FM band. I used to use the transmitter part as an FM transmitter back in the days before little ipod transmitters were available, it only worked at close range. I also used it to transmit to a TV22 and TV62 once, many years before I got the Aurora too, I shoved 625 line video up it through the audio inputs, and wrapped a load of bare wire round the transmitter and connected it to the aerial sockets. The TV22 got the signal, but couldn't display much of anything, but the TV62 managed a picture with the line hold wound right up! It was a bit fuzzy to say the least, but it did work! I thought it was going to be the best picture I'd ever see on an old TV at the time...

Regards,
Lloyd.

Optical 'point-to-point' has been successfully done in the past, as distinct from the optical 'broadcasting' we are considering here. Eddie's example above is one such instance. Chris Long in Australia has great expertise in this... see here: http://www.bluehaze.com.au/modlight/index.html (link kindly supplied by DrZarkov over on the NBTV board).

Steve

That's why I was assuming maybe 500 LEDs for transmitter and narrow vertically and probably needing 20dB to 40dB of filtered amp on the photo transistor, maybe even Peltier cooler on it to reduce noise.
The point to point link gain is totally massive.

A point to point link over short distance will be relatively easy and reveal the real signal levels and powers involved.

Remember that the effective antenna gain on a point to point system is very high compared to an omnidirectional radiator. For optical the difference will be huge. Even if its not true omni but just a sector the difference will still be huge.

Absoloutely true, but consider the following as an illustration. I mentioned above 20 watts of omni output would do the job. This is actual light power. Now imagine placing an incandescent lamp (say 5% efficiency) of 20 watts 'light power' - so this now becomes a 400 watt bulb - at the top of the Crystal Palace tower.

You could see this on a clear night, even with the naked eye, as a brilliant pinprick, for miles.

For a receiver, you'll recall I cited a 10-cm (4-inch) concentrating lens at a distance of 5Km (3 miles). Now I happen to have some binoculars with 4-inch objective lenses (25x100) and I can tell you a 400 watt lamp at night at 3 miles would be easy to spot through them. It would like like such a lamp would look like with the naked eye at 211 yards, with an eye pupil of 4mm diameter. I'll have to work out what stellar magnitude this would be equivalent to, but I suspect it would be a negative number!

The above uses a true omni incandescent bulb as an example. But the light from the emitter I am suggesting would be concentrated in a circular (fat) line, giving much more light concentration than the bulb.

Steve