[Laser] LED used as APD

[Garnier Yves]

Hi Charles
My first experiments with various colours LED with this device diagram gives these first results :
http://sd-1.archive-host.com/membres/images/22679775843705539/UNIVERS_RX_B.jpg
The quantum efficiency of a LED at low reverse voltage when used as photodiode seems very low comparing a classical silicon photodiode.
May be 1/4 of the Q.E. of a Si PD with the same sensitive area...
The best sensitivity I found for the moment is from a 5 mm ultra bright 660 nm "deep red" LED.
To know before if a LED can work in avalanche mode or not it is mandatory to test its inverse leakage at low reverse voltage.
The lower the leakage the best avalanche effect gain will occur.
In a 10 LED batch, one or two of them could be well better the others.
The avalanche effect is abrupt and critical.
It seems very difficult to stabilise it near a usable point.
During very short times it seems to produce a gigantic gain but bias oscillations can appears.
I will be happy to have some reports about your own experiments.
Yves F1AVY
http://f1avyopto.wifeo.com


--- En date de : Mer 27.1.10, Charles Pooley <ckpooley at sbcglobal.net> a écrit :

> De: Charles Pooley <ckpooley at sbcglobal.net>
> Objet: Re: [Laser] LED used as APD
> À: "Free Space LASER Communications" <laser at mailman.qth.net>
> Date: Mercredi 27 Janvier 2010, 16h16
> I read with interest the article re
> use of LEDs as avalanche detectors, but one fact seems
> missing:  The quantum effriciency of the diode
> itself.  For example, if at a voltage below the
> avalanche voltage so it operates as a photodiode, what would
> be the current from, say 10 microwatts of light focused so
> it falls on the sensitive area?  10%, 50%, more? 
> Si PIN diodes have typically 70%. 
> 
> And another:  the sensitive area.  Will there
> have to be a scanning experiment to map it?  
> 
> The article I got (from Tim, I think), says that some L:EDs
> work but blue ones did not because the dark current rose to
> high before the avalanche voltage was reached.  Would
> this work with various longer wavelength, and IR
> wavelengths?  850 nm, 940?
> 
> My interest is in the various wavelengths available with
> small diode lasers, both NIR and red.
> 
> Charles Pooley     Microlaunchers
> 
> 
> 
> 
> ________________________________
> From: Tim Toast <toasty256 at yahoo.com>
> To: Optical_DX at yahoogroups.com
> Cc: laser at mailman.qth.net
> Sent: Tue, January 26, 2010 11:35:43 PM
> Subject: Re: [Laser] LED used as APD
> 
> With all the blue and UV led's out there, i bet a good
> match 
> can be found for the 3371A nitrogen line. Relatively simple
> home
> built nitrogen lasers that work in open air (TEA) make
> short 
> nanosecond or even picosecond length pulses. A large beam 
> spreading lens would be in order of course since these are
> high
> peak power pulses.
> tim
> http://www.aladal.net/toast/exp.html
> 
> ------------------------------------------------
> 
> On Wed, Jan 20, 2010 at 8:54 PM, Garnier Yves
> <f1avy at ...> wrote:
> > So if I well understand this effect a blue LED used as
> detector 
> >could be able to detect UV A ?
> 
> looks promising, sparks' emission spectra in open air peaks
> at 
> 337nm, the  superradiant nitrogen line. A significant
> portion 
> of the energy could be detected by the blue led, which
> opens a 
> path for spark signaling.
> If UV leds can work in APD mode too one might want to try
> them 
> and nitrogen lasers for very long range communications.
> inverse <inverse at ...>
> pagaiard 
> 
> 
> Tim
> I noticed an error in my last mail:
> I want to write : "Unfortunately it seems that a LED diode
> used 
> as photodetector receives a wavelenght slightly shorter
> than 
> its own transmission wavelenght."
> I confused wavelenght variations and frequency variations
> :-) !
> 
> So if I well understand this effect a blue LED used as
> detector 
> could be able to detect UV A ?
> Yves F1AVY
> http://f1avyopto.wifeo.com
> 
> 
> Tim
> Unfortunately it seems that a LED diode used as
> photodetector 
> receives a frequency slightly lower than its own
> transmission 
> frequency. To receive a classical RED Luxeon you must use
> an 
> other "deep red" LED.
> The main interest seems to be only the low cost of the LED
> and 
> its filter effect.
> The avalanche effect seems too abrupt and instable for
> linear 
> photodetection but it could be usefull for high speed light
> 
> pulses detection if followed by threshold and pulse
> shaping 
> circuitries.
>