[Laser] PIN diodes and Noise floor

[David D. Rea]

Andy,

There are definitely two separate noise components that you've got to
deal with, and you nailed both of them. Part of the noise - the part you
can't easily deal with - is due to thermal excitation in the PIN diode,
and is a function of temperature. I don't recall whether there is any
forward current dependence, like there is for a MOSFET, with a diode.

The second component is due to ambient light. Take a look at the two
circuits on the K3PGP page, one for daylight and one for night time use.
With the night-time circuit, the diode is used to pump electrons into
the gate of the MOSFET. Since the FETs gate is around one diode drop
above ground (when any appreciable current flows), and the
drain-to-source voltage is maintained at some comfortable level by the
two biasing resistors, it's fairly safe to say that Vds > Vgs-Vt, thus
the transistor is almost always in saturation. Since it's saturated,
current changes in the FET and thus voltage changes at the receiver's
output are a function of gate voltage, which (since we're running the
PIN diode in photovoltaic mode here) are dependent on light.

As for the daytime receiver, you'll notice John has reverse-biased the
detector. With a reverse bias, the detector operates in photoconductive
mode rather than photovoltaic - so instead of actually creating
electrons from the light that falls on it, it allows them to pass when
light falls on it. More light means more current flows (even hundreds of
milliamps with a big enough PIN diode) and thus the drop across the 82K
resistor increases. This voltage change is capacatively coupled onto the
gate of the FET so that you don't get a big nasty DC component screwing
up your biasing - which appears to also be saturation due to the 22M
resistor at the gate.

Anyone - please correct me if I'm wrong with this analysis...I'm doing
mental napkin-level work... :)

Looking at these two circuits, its easy to see how John compensated for
the ambient light noise that he would inevitably see with a daylight
receiver...by using capacative coupling, only the AC portions of the
signal get through. As long as the capacitor is big enough, and a 2.2uF
is definitely big enough, the lower frequencies like 60hz from light
bulbs, 120hz from Flourescent bulbs and DC from the sun won't get
through. Unfortunately, the trade off is that you have two additional
resistive lumps before the gain stage of the amplifier (i.e. on the gate
side of the fet) and so you're generating two more resistor noise
components. They will be fairly high since resistor noise is directly
proporational to both temperature and resistor value.

Hope this helps with your understanding of noise, and how the 'PGP
frontend works. I still would like to get together for coffee sometime
since we're in the same city anyway!! Let me know if you're interested,
and we can compare notes...I'd like to learn more about the slow CW
you're working with!

73
Dave
RIT EE

On Fri, 2003-04-18 at 18:12, Andrew T. Flowers, K0SM wrote:
> I have a question for folks who have played around with PIN diode 
> detectors--
> 
> I've put together a few of the K3PGP front ends now and I'm really 
> looking forward to getting my laser diodes going for a complete 
> tranceiver system.  Anyway, I was wondering if the PIN diode dark 
> current is much of a factor under actual nightime use?  Is the noise 
> floor coming from ambient light or from the diode itself?  I imagine 
> this will depend largely on how big of a lens you are using, and how 
> clear the sky is, but I'm trying to get a ballpark figure.  John 
> mentions that cooling the diode improved the NF, e.g. putting it in the 
> freezer for a while, but does that really matter practice?  Does ambient 
> light drown out the dark current?  I'd test this myself, but it's not 
> easy for me to get away from city lights.
> 
> Andy K0SM/2
> 
> _______________________________________________
> Laser mailing list
> Laser@mailman.qth.net
> http://mailman.qth.net/mailman/listinfo/laser