[Laser] Re : PWM of LED in QST

[Tim Toast]

Hmmm, i hadn't thought of the PGP receiver reacting differently 
to the pulsed signal of PWM. But, as you say, it has that unique
responce to short pulses. Odd, since it has the fairly steep 
responce roll-off above a 1 khz or so. 
The only sense i could make from avoiding using more than about 
half of the full range of duty cycle with a PWM, would be; to 
avoid over modulating it on the voice or music peaks. I know the 
average to peak speech waveforms can vary quite a bit from one 
person to another. I guess you would need some sort of audio 
limiting to take full advantage of a wider duty-cycle swing 
without sometimes hitting the hard "edges" of the modulator.
I was wondering what people thought about the PGP day-time version.
- Does it have a more flat frequency responce than the night-time
version? And would it be a good candidate for a kit with full 
voice bandwidth? If this was discussed before i have forgotten.




[Laser] Re : PWM of LED in QST
f1avyopto at aol.com f1avyopto at aol.com 
Tue Sep 28 08:29:39 EDT 2010 

Previous message: [Laser] PWM of LED in QST 
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] 

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

If the ampltitude to pulse width convertion stays in a linear fonction, it seems to me the full range of 1/99 to 99/1 is possible but the PWM response of an optical receiver depends of its design.
On a very sensitive K3PGP RX or a transimpedance RX with its very high feed back resistor and with its small added capacity for stability, it is the photodiode capacitance that integrates the pulses train.
This effect is not linear and asymmetric because the capacitance is fast loaded each pulse, but this capacitance unload very more slowly versus the equivalent resistor viewed from the photodiode.
This produce a charge amplifying effect that give a very high gain on the VLF signal from the modulation envelope.
This gives a big audio flickering and increases so much the signal scintillation that then appears saturate "click" noises and cut periods.
It is not really a PWM problem because this phenomenon can be the same with a 100 % direct linear amplitude modulation on the Led or laser current.
To reduce this phenomenon one can reduce the load resistor of the photodiode in the RX, also reduce the inter-stage capacitors to reduce the VLF amplification and to reject the PWM frequency modulation one can filter with a dedicated stage.
Because the price to pay is a sensitivity loss, the best way to cancel this phenomenon could be to simply reduce the optical signal ?



-----E-mail d'origine-----