[time-nuts] Photodiodes for high frequency OPLL

David Kirkby david.kirkby at onetel.net
Sat Mar 30 16:45:43 EDT 2013

On 30 March 2013 19:50, Attila Kinali <attila at kinali.ch> wrote:
> On Sat, 30 Mar 2013 18:43:03 +0000
> David Kirkby <david.kirkby at onetel.net> wrote:
>> I would contact Hammamatsu Photonics
>> http://www.hamamatsu.com

> This might be a good idea. Thanks!

>> I played tricks during my Ph.D. of gain-modulating an APD with a step
>> recovery diode.

> You mean pulsing the bias voltage at about 7GHz and see whether
> i can get a something out of that?

I did not really have anything in mind specifically. I was just
wondering if there was anything could do modulating the gain of a
detector so forming a non-linear detector.

I pulsed the voltage on an APD to buid a cross correlator. By doing
that I was able to get better temporal resolution than the basic APD.

Quick summary here:

Fulll thesis here

I have seen systems where one sinusoidally modulates the voltage of a
photomultipler tube. You could do the same with an APD. Then it acts a
a mixer. Essentially

1) Modulate laser at 100 MHz
2) Modlulate the APD bias voltage at 100.01 MHz
3) Detect light from laser on an APD.
4) Output of the APD is a sine wave at 10 kHz.

> There has been a similar idea around, using a third laser that
> is pulsed with the desired offset frequency.

I was wondering if it could be done in more than one stage. A bit like
a superhetrodyne receiver.

I've had quite a few beers tonight (OK, I am ****ed) so my mind is not
thinking too straight.

>> Have you considered using a photomultiplier tube? I think there are
>> some fast PMTs around, though perhaps not fast enough. 20 years I used
>> to know what was state of the art in this area, but not any more.
> Not really. I am mostly looking at stuff i can easily get my hands on.
> If possible without using ebay (other should be able to build the
> system as well).

PMTs are available new. Hamamatsu has almost 100 different types.


so there is no need to trawl ebay. Using PMTs you can easily get sub
ns temporal response.


More information about the time-nuts mailing list