Measuring speed of light / time of flight

My current project idea is to use statistical resonance to measure the speed of light. Actually not the speed but the time of flight. I have documented the current state of affairs here: http://blog.blinkenlight.net/2011/12/01/thank-you-world/. As it turns out this is much harder than I expected. The current obstacle is heat build up / temperature sensitivity in the laser control circuit.

Has anyone here an idea on how to improve my setup to deal with this? I already got the suggestion to switch to a blue laser instead of a green one. The reasoning is sound thus I already ordered a blue laser. Any further hints would be highly appreciated.

It looks like an interesting project; but I think you're making a big mistake by using a diode to drop the voltage to the laser instead of a series resistor. Laser diodes, like LEDs, are basically current-driven. The current is what controls lasing action, but the forward voltage depends on temperature. Therefore, if you drive the diode from what is more or less a constant voltage, the lasing action will vary greatly with temperature. The margin between the current at which the diode starts lasing and the (higher) current at which it overheats is quite small.

Use a series resistor, or (better) an adjustable constant current source.

Very good point. I will fix this with the blue laser module.

Have you investigated how they determined the speed of light in a pre-arduino era? Could give you some clues.

1850 - http://en.wikipedia.org/wiki/Hippolyte_Fizeau - but there were earlier attempts as early as ~1600 IIRC

Have you investigated how they determined the speed of light in a pre-arduino era? Could give you some clues.

In 1862 Leon Foucault measured the speed of light as 299,796 km/s, today's figure is 299,792 km/s so he did not bad. He used a rotating mirror method. I had a go at using the same technique, but the speed I was rotating the mirror at made it disintegrate :(

Of course you have not said what medium you are measuring the speed of light in. In a medium with a very high optical density it is possible to slow light down to walking speed. I believe experiments are being done where such a medium flows faster than the speed at which light can pass though it. This makes it possible to create a sort of optical black hole with light trying to shine out but the medium being sucked down into a vortex - but that is getting a bit off topic.

You can be assured that I investigated this. I also am fully aware that the speed of light can not be measured anymore. This is because it is defined as a constant. My point is to measure it with an Arduino and my Blinkenlightshield. Not because it is the most efficient way but because I think it can be done.

Also the mirror experiments are not easy to setup. They sound easy but in practice they require a solid adjustment and some reasonable good optics. As I describe in my blog I want to try the same approach as the "ping" experiment. That is I want to throw enough maths at it to extract the results. However as I said right now heat build up is an issue. As suggested I will improve the laser module. I will also refine the statistics. Maybe this is then already sufficient.

Is there a university or so nearby where you can get liquid nitrogen. MY nephew has used it for overclocking a Pentium PC in 1999 and after 3 or 4 boards it worked. He made a sort of funnel of copper to cool the CPU. The (non tech) article is in dutch - Cursor/ -

The advantage of using N2 is that it will absorb almost all heat and the laser will not heat up easily.
The risk is that the laser/electronics won’t work at that temperature so there is a serious chance to break things.

BTW, the description of the experiment sofar looks quite good.

Yes, there is a university nearby. However using liquid nitrogen would defer my intention to pull this off with a somewhat simple setup. If I would want to determine speed of light with non trivial means I would just use my DSO and look at the phase shift of the detected laser pulses while increasing the distance of laser and sensor.

somewhat simple setup.

ICEcubes ==> -18C / 0F [depends on your freezer] Note that water and electronics don't mix too well, but the cooling effect could be good enough as water is a good heatsink. Simpler than Nitrogen as you do not need to leave home ;)

Other idea reventing the heat from building up in the first pplace: - how much power does the laser use? - can you minimize it some way? - the intensity may drop as long as it is a detectable pulse...

Other thought - coolant liquid for cars? don't know if it conducts...

I know you want to keep things simple but maybe you could use a Peltier Thermoelectric Cooling module (seems to be abbreviated to TEC, TEM or TECM ) http://www.youtube.com/watch?v=3fYUVsydMmA

They look like they would be fun, less messy than icecubes, give you temperature control and not be too expensive.

The issue is NOT that the laser overheats. The issue is that it did never reach a steady state.

I was reading your blog where you said; "The heat build up is inside the laser control circuit. I might be able to even this out by running the experiment for days in a temperature controlled environment. "

I thought the Peltier might give you a way of actively stabilizing the control circuit temperature.

How small of a distance do you think this experiment could run in - provided you could get it stable? I mean, if you can get it to measure time-of-flight for distances between .5 meters and say 5 meters, with a good accuracy, fast enough - well, you'd have a somewhat simple laser distance measuring device (although it might still be better to measure doppler difference using a pulsed laser and a vfast timer circuit).

Interesting experiment, nonetheless! Good luck with it!

:)

radman: I was reading your blog where you said; "The heat build up is inside the laser control circuit. I might be able to even this out by running the experiment for days in a temperature controlled environment. "

I thought the Peltier might give you a way of actively stabilizing the control circuit temperature.

He'll have to correct me if I'm wrong, but I'm betting what/where he means by "laser control circuit" is that little circuit board sticking out of the back of the laser module; that, and "heat build up" being some really tiny amount, but still so large as to swamp what he is trying to measure (I mean, he's trying to measure something at the threshold of the noise level via statistical sampling methods - so it's gotta be a pretty small thing, if I am understanding this correctly).

Adding a peltier might just make the whole thing worse, instead of better, because of extra "noise" introduced by control of the peltier to try to achieve temperature stability with it - but then, I'm talking out my rear here, I really don't know...

:D

Thanks cr0sh! You seem to be one of the first who really understands the whole issue that I am dealing with. Over at hackady most people did not even read the article before they started complaining.

Thanks to cr0sh I think I understand your problem. You will not get an answer to it from me. However I have often found that when I have been stuck with something just talking about it to somebody else makes me look at the problem in a different way (instead of going round the same old thought loop) and often a blindingly obvious solution pops into my head.

Is the temperature sensitivity of the laser control circuit the same at all ambient temperatures? I can understand why actively controlling the temperature could cause more problems than it solves but if you just run at a much lower ambient temperature, even if it is not stable, would that reduce the sensitivity of the circuit.

Probably going off at a complete tangent, as I know nothing about electronics, here is a quote that might (or very possibly might not) trigger some thoughts;

Temperature compensation Virtually all semiconductors and the circuits comprised of them exhibit a temperature coefficient. Owing to their high positive temperature coefficient, NTC thermistors are particularly suitable for compensating this undesired response to temperature changes (examples: working point stabilization of power transistors, brightness control of LC displays). Resistors in series or shunt plus suitable voltage dividers and bridge circuits provide an excellent and easy-to-implement compensation network.

Maybe I misread your original post, but when I suggested using a series resistor to drop the 5v to something the laser can handle, I wasn't aware that the laser had some control electronics and that the problem is the control electronics getting too hot. The electronics is probably a constant current circuit, and will dissipate (Vi - Vl) * 0.3 where Vi is the input voltage and Vl is the voltage needed by the laser diode. So to avoid overheating it, you need to reduce the input voltage. Currently you are feeding it with 5v less the drop of your switching transistor (which I guess to be about 0.2v) and the drop of the series diode (around 0.7v). So it's probably getting about 4.1v, whereas it is designed for 3v to 3.6v.

You could try putting another 1n4001 diode in series with it. However, the constant current circuit is unlikely to be designed for fast switching, so you may find it difficult to generate pulses of laser light that are sufficiently well-defined for your experiment.

What I would do is to trace the control electronics to work out exactly what it is doing, i.e. is it just a constant current source or is it doing something else such as temperature compensation. It looks simple enough, I can see a transistor, two resistors, a pot and a capacitor, although there could be more components on the other size of the board. Then build a replacement controller to replicate the control electronics, adding the ability to pulse the laser quickly.

Alternatively, buy a laser diode without control electronics, and make your own controller. Once simple way of driving the diode with a current that is near enough constant would be to run it from higher voltage (say 12v, preferably regulated) with a series resistor chosen to get the right current. Then use the transistor switch you already have to pulse it.

PS - in case you haven't found it yet, there is some good information about laser diode controllers (including modulating the laser output) at http://www.repairfaq.org/sam/laserdps.htm#dpsldd.

By now I understand how solid state laser control works. Seems that it is not the control circuit but the laser itself that will heat up. Anyway I think it is a good idea to switch from a DPSS to SS laser. I think the next attempt will use more mathmatics to deal with the heat build up.

[quote author=Udo Klein link=topic=81097.msg615614#msg615614 date=1323103905] By now I understand how solid state laser control works. Seems that it is not the control circuit but the laser itself that will heat up. Anyway I think it is a good idea to switch from a DPSS to SS laser. I think the next attempt will use more mathmatics to deal with the heat build up. [/quote]

Can't you drive the laser in short pulses for your experiment? That would avoid heat build-up.