[time-nuts] Photodiodes for high frequency OPLL

[Attila Kinali]

On Sun, 31 Mar 2013 15:27:02 +0100
David Kirkby <david.kirkby at onetel.net> wrote:

> You can't modulate the sensitivity of a PIN diode by an adjustable
> bias like you can in an APD. For a PIN diode, at the wavelengths you
> are talking of, you basically get around 0.5 Amps/Watt from the diode.

Right.. Forgot about that ^^'

 
> APDs are not that expensive, especially the small ones you will need
> for this.

Well.. an APD is about 60-100USD while PIN photodiodes start from 10USD.
Ok, considering that the rest will be probably in the 2000-5000USD
range, this isn't much :-)

> But I still can't really see a solution - I was just
> clutching at straws.

I think the gain modulated approach to downconvert the signal should work.

Taking the beat signal of the two lasers and applying it to the APD
will get me an approx 7GHz signal. Gain modulating the the APD with
an 1-2GHz sinusoidal should downconvert it to something below 1GHz.
The frequency of the modulation signal is low enough to "pass" trough
the APD unperturbed (the APD is specified for 2.5Gbps).
That the gain of an APD is an exponential function of the applied
bias voltage should make the bias modulation "sharp" enough to approximate
it with a pulse train. Ie it should "nicely" downconvert the beat signal
to something below 1GHz.


				Attila Kinali
-- 
The people on 4chan are like brilliant psychologists
who also happen to be insane and gross.
		-- unknown