[time-nuts] Photodiodes for high frequency OPLL

[gary]

Reverse biased diodes are not very low impedance. You can't really drive 
a low impedance with them.

In fact, the high impedance of the photodiode leads to all sorts of 
ugliness in noise analysis. This is well documented in Graeme's book. I 
think TI also has an app note on feedback networks for the best SNR, 
though Graeme has a technique to get optimal bandwidth and noise 
performance.

A lot of people make a living just designing front ends for high 
impedance devices. They are really hard to do optimally since high 
impedance is bad for bandwidth, noise, and all things nice.

There is also a book by Hobbs on photodiode interfaces. Much of the book 
is unfortunately filler, though parts are very good. I found a pirated 
copy online and decided not to buy the real thing. Graeme's book is much 
more detailed. The derivations of the equations are shown in painful 
detail. [My dislike of Gain Technology comes from dealing with the 
company. I have no personal knowledge of Graeme other than buying his 
books.]

You might want to track down the pirated version of Hobbs before 
spending $150. I did, then deleted it since I already owned Graeme's book.
> Building Electro-Optical Systems: Making It all Work



On 4/1/2013 10:30 AM, J. Forster wrote:
> I think that you should be able to take the signal right out of the
> backbiased diode and run it straight into a microwave mixer w/ a fixed 7
> GHz LO and frequency discrimitate the IF output from the mixer to generate
> a control signal for your LASER.
>
> In theory, if your LASERS were very, very, very good, you might be able to
> achieve phase lock, but I'd not hold my breath. A discriminator can give
> you frequency lock.
>
> YMMV,
>
> -John