Thanks for the nice comments chaps.
Harry Dalek wrote:BTW have you tried a night test by reflection off different surfaces ?
I have been using a small mirror in the workshop to fold the optical path so I can work with both the transmitter and receiver next to each other on the bench. Yes, it would be instructive to try bouncing it off windows etc. This should work just as well during the day as at night. The advantage at night is you can more easily point the beams by eye.
Harry Dalek wrote:i suppose it would have to be close depending how much IR your led or leds can do ,i like the idea of light skip like Radio propagation .
I've been steering clear of infra-red for now, because then you *can't* see the beam at all and judge it's of safe brightness. As it is, the deep red 660nM light I'm using has what's known as a 'luminous efficacy' of only 40 lumens per watt. This means the eye is only 1/16 as sensitive to it as peak yellow, which in turn means it's not a good idea to stare at it - there's more energy in the beam than it appears.
Harry Dalek wrote:As i remember head amps or LNBs on satellite dishes work better if you can cool them to stop thermal noise....I wonder if you did the misty light transmitting test and got some cold spray and cooled your detector would the noise decrease
It might do - slightly - I suppose, if it were really cold. I could try squirting on some freezer... in the meantime I have now improved things by tinkering with the circuit.
Harry Dalek wrote:IS IR light or heat ? or the same
It's all the same stuff, different wavelengths. The higher the temperature of the radiating body, the greater proportion of shorter wavelength infra-red, which then merges into light.
Steve Anderson wrote:I wonder, over longer distances of several km would atmospheric disturbances affect the signal?
Yes I believe they would. Based on data from how stars and various planets twinkle (or do not twinkle) when at low altitudes, I reckoned this effect would be partly determined by the angular size of the radiating array, and it would kick in only after several miles. I called this 'baseband flutter'.
Harry Dalek wrote:Chris had the advantage of a hilltop-to-hilltop path away and above the rubbish in the air that a large city creates. I also think these voice links were conducted at night when maybe the atmosphere is more settled.
In astronomy, the most stable 'seeing' is often accompanied by haze, while the most transparent weather conditions can be as unsteady as a jelly!
Klaas Robers wrote:Steve, I can't find your circuit diagrams back, but you will know that the cumbersome capacity of a photo diode (of ALL diodes) decreases if you apply a voltage in the non conducting direction. This is the effect that is used in varicaps. In reality all diodes are rectifiers, varicaps, zeners and photocells, but the effect is only specified (measured) for diodes that are named in that way. A semiconductor diode is just a semiconductor diode.
Yes I tried this, for the reason you mentioned. However in this case, and with this diode, it seemed to give no advantage over photo-voltaic mode, which - as I understand it - also offers the advantage of the best linearity. Since I'm amplitude modulating with an analogue signal, itself amplitude modulated, linearity is an important factor.
Klaas Robers wrote:Secondly try to find a coloured filter that is transparent only for the colour of light that you use. This will remove the noise introduced by all other colours of light that might hit the detector diode. For near infra red these filters exist in remote controls.
I'm currently using visible light of course. A gelatin filter would also introduce significant losses, even within its passband. An interference filter might be better but would have to be broad enough to encompass the (fairly narrow) bandwidth of the LED.
With high frequency modulated, weak light - 'quantum noise' can become a factor too.
Steve O