I would like a suggestion on a price-efficient way of experimenting with spontaneous down-conversion and single-photon counting. The simple dual-slit experiment could be one part of an application although there are many, but I'm mainly asking about the key components which are similar - the laser that pumps a SPDC-crystal, the crystal itself, and photon-counters (the other standard optical bench items I'm more familiar with).

For example, is a 405 nm (blue) diode-laser module the way to go, with a BBO-crystal, and what to use for detectors - are avalance photo-diodes good enough? What kind of quality does the laser need to have (wavelength stability etc)?

I know there is a Journal of american physics teachers or something which have articles covering these fundamental experimental setups now and then, but I thought I'd ask here in case anyone who works at a lab (with limited funding :) with a good setup could share some ideas.

If this gets many answers, it should be turned into a community wiki entry I guess.

  • $\begingroup$ Since I don't know the particulars of the experiment you are asking about I can't provide a good answer. Decent quantum efficiencies for high optical frequencies are easy to get with PMTs but they can run to a lot. Multi-channel plates are cheaper, but you loose some QE. And have you considered a bare CCD? That's probably the cheapest instrument going. $\endgroup$ Commented Jun 23, 2011 at 15:35
  • 1
    $\begingroup$ I basically want equipment to be able to reproduce some of the fundamental experiments with single-photon interference through a dual-slit, interaction-free measurement in an interferometer etc. Some don't require a downconversion setup of course, but I still would like to re-use as much equipment as possible if I need that later. There are some 4-channel APD devices available for example. It doesn't need to be super-cheap but it should be good, useable equipment. $\endgroup$
    – BjornW
    Commented Jun 23, 2011 at 18:19

2 Answers 2


The traditional method is with a photomultiplier tube or PMT. These are very expensive but are easily recognized and you might try hanging around the auctions at the local research university to see what comes up.

More modern is the Avalanche photodiode or APD. Amazingly, my favorite electronics supplier, Digikey, has them, at this time as cheap as $942. These might be harder to spot at the auction (as they are quite small).

Never fear, this is another alternative. The maker of the Digikey APDs is Advanced Photonix Inc which offers them only through Digikey, at least according to their website.

If I were an aspiring young scientist, I might consider contacting API and seeing if they were willing to loan me a "solder sample" (out of spec) APD. Before contacting them I would do the following:

(1) Get the data sheet for their most common and cheapest part.

(2) Study the data sheet enough to completely understand the part.

(3) Design a circuit that will use the part in a manner so as not to damage it.

(4) Send the proposed design to an API sales / support engineer using their contact form.

(5) Apply prayer.

(6) If it didn't work, repeat with another APD maker.

The sad fact of engineering is that the sales reps give away free samples to engineers who almost never use them. If you happen to known someone in an industry which uses APDs, they may have some sitting around in their desks that they couldn't care less about. I'd be willing to bet that there are LHC engineeers / physicists who would have a few lying around.

To get one, you might consider using your contacts to find someone connected to you who works at one of the colliders. If you're at a big school, it should be easy to use google to find a prof who's in a collaboration.

  • $\begingroup$ Thanks Carl, some very useful experience you share here! May I ask if you think APDs have enough QE and low enough dark counts to extract useable results from correlated single-photon detections? I mean, the usual signal/idler -> detector setup you see in many published experiments. I guess the answer is yes as I've seen them in the eq lists of some. Another thing that bugs me is the quality of the laser, the line width obviously has to be small enough, is this the main point differentiating diode lasers for quantum optics? $\endgroup$
    – BjornW
    Commented Jun 24, 2011 at 9:10
  • $\begingroup$ Hmm seems like you need more heavily biased APDs for single-photon use, the quenching-types. I found this (freely available) article which describes one of the experimental setups I wanted to recreate, along with a list of materials: people.whitman.edu/~beckmk/QM/grangier/Thorn_ajp.pdf $\endgroup$
    – BjornW
    Commented Jun 24, 2011 at 12:39
  • 1
    $\begingroup$ @Bjorn; Good! I think you already know more about APDs then me. This means you are well on the way to getting what you need. All I can contribute is a short description of the human engineering that might help. Please keep us apprised of your progress. $\endgroup$ Commented Jun 24, 2011 at 15:16
  • 1
    $\begingroup$ Here is a short paperhttp://spie.org/etop/ETOP2005_006.pdf describing a Bell-inequality set-up used in an undergraduate lab. Other more detailed papers are available for the same set-up, but they're in French. $\endgroup$ Commented Jun 27, 2011 at 13:16
  • 1
    $\begingroup$ @Bjorn: The following references are in French: What is given to students is paristech.iota.u-psud.fr/site.php?id=63&fileid=957 (p117 of the pdf) and the experiment is described to french physics teachers in institutoptique.fr/content/download/1093/7847/file/… . Those experiments are set up in Institut D'Optique graduate school. On this page, you have the contact email of the teacher who built the experiment institutoptique.fr/formation/Ingenieur-Grande-Ecole/… $\endgroup$ Commented Jun 28, 2011 at 9:30

Berkeley's undergrad lab should have a competitively cheap setup. Here's the wiki. You might be able to contact the lab manager for more information, which can be found through here.

As predicted in the last answer, it's APDs and BBOs.

  • $\begingroup$ Both links are dead $\endgroup$
    – Cheng
    Commented Feb 7, 2023 at 9:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.