Simple, cost effective, and accurate measurement of the speed of light/radio waves

[resurgence2012]

Hi Everyone,

I am new to this forum.

I am teaching high school science and I have been trying to find a simple, accurate and cost effective way to measure th speed of light.

I know that it can be done easily with a 100 MHz oscilloscope, pulsed laser and photodiode. However, my school does not have these resources, so I have been trying to find an easier, and cheaper way.

I considered trying to recreate the experiment done by Fizeau, where he used a rotating cog wheel, as a beam chopper and shone a light some 15 km before reflecting it back at the same cog wheel. I like this apparent simplicity of this approach, and it is also fairly easy for younger students to understand.

However, this method still seems to present a number of difficulties: manufacturing the disc, obtaining a fast enough motor, setting and aligning the optics, etc.

So more recently I was considering building a circuit that could be used to pulse a laser at say 20KHz, splitting the beam, and after send one half of the beam 300 meters, using a phase detector circuit like an XOR gate to detect the phase difference of the pulses, and using that, together with a knolwedge of their frequency, to calculate the time of flight. Of course this defeats my aim of having younger students or ideed 99% of my older students understand what's happening.

Notwithstanding, given the suggested frequencies etc, and a time delay of about one micro second, could I use a CD4046 CMOS? If so could anyone point me in the direction of a reliable easy to follow circuit diagram?

Alternatively, does anyone have any other suggestions for doing this experiment, perhaps more easily, but more accurately than just sticking a bar of chocolate in a microwave oven and measuring the distance between the hot spots?

Thank you very much for your patience in reading this and any thoughts or ideas you share.

Cheers Peter

 

[chuckey]

have you any TV kit? I was thinking of driving the laser with the line sync pulses that also drive a camera whose output is on a large TV set. If you put some filter over the camera lens (so you don't burn it out!) and point it at the reflected laser spot, the delay will show a shift in the position of the vertical line on the screen. Depending on what system you are using you can calibrate the width of the screen in microseconds. With system I, the active line is about 58 microseconds (from memory). Is it worth fiddling about with?
Frank

 

[resurgence2012]

Thank you Frank. I could easily buy a cheap TV but I am not sure I know enough about the electronics to know how to hack into it.

I think that maybe the optical beam chopper idea might still have some life in it. I found one company that sells a brushless motor which is controllable that runs upto 120,000 rpm (2000 Hz) which is getting closer to the mark. I have no idea how Fizeau was able to do the experiment 150 years ago. He got one of the first fairly accurate measurements of c as well.

 

[chuckey]

So your motor rotates at 2000Hz, i.e. once every 50 micro sec. So you need at least 50 holes in it, 100 would be better. Fizea used a cog and a much longer path length. So 100 holes + 100 blanks, or 200 X laser spot diameter = 2 X PI X disc diameter?
A motor rated at 120K RPM sounds expensive, a more practical option might be a common or garden woodworking router, these are variable speed and rotate at 20 K RPM and are 1/2 HP, so they can swing a large disc (with suitable safety cover!!!).
Frank

 

[resurgence2012]

Thanks Frank - several months later!

I am still working on this project and to be honest have not made that much progress.

So I did eventually concede to purchasing a Unit Trend 50 MHZ scope.

I did an experiment using a signal generator and loudspeaker and two sound sensors, set about one metre apart, and connected to separate chanels. It was easy to see and measure the speed of the wave by looking at the phase difference.

So I thought, "why not do this with radio waves?"

But what is the simplest circuit to use? It is easy to find a cheap fm transmitter. But I am struggling to find a circuit that I can use to detect the radio waves and display the signal on the oscilloscope.

I tried building a couple of different rf pre-amps but I am nit sure how I should connect them to the scope. So far I don't seem to be able to pick up anything on the scope. Is tue signal too weak? Should I use more than one stage for the pre-amp? I tried a circuit using a JFET (tried the same circuit twice) and then another using a BJT and stil noohing.

I don't know a much about any of this so I am very much still groping in tue dark. Cheers Peter

 

[Baas Rietrot]

Why not use/hack a cheap radio set for your receiver?

 

[BradtheRad]

Since you have a fast oscilloscope...

You can blink a laser diode, which is aimed toward a reflector 500 feet away, and see the reflection 1 microsecond later.

Each pulse is picked up by a photodetector.

To show students the reflection is real, you can modulate audio into a continual laser beam, and have a student 'catch' the delayed reflection in a detector, send it to a demodulator, then send the audio to speakers.

The student can move the detector away, the sound stops. Then he or she puts the detector in the beam, the sound returns.

 

[resurgence2012]

Why not use/hack a cheap radio set for your receiver?

Yes - buy I am still not sure about how to connect up the oscilloscope - probably sounds like ssomething really obvious to anyone with more experience but I am baffled.

I have built a superegen receiver which work quite well but I don't seem to be able to detect the carrier wave without killing the signal. Somewhere near the audio amp I can pick up what looks like a very nicely frequency modulated IF - haven't got a clue actually how that works - but it is another long story.

I am guessing that this might be an issue to do with impedance. I was going to try doing something like building another good quality tuned rf preamp followed by another stage but with no filtering to preserve the raw rf signal. Maybe even cascading a couple of the rf preamps before further amplification - but really I am still just groping in the dark. Tried for a long time to find some decent material on the net but no success so far.

Actually, I probably need some more experience using the oscilloscope. Cheers peter

 

[chuckey]

You can compare the phase of two sine waves on a CRO by means of lissajous (Sp.?) figures. basically you feed one into the Y input and the other into the X input (normally used for the time base). From memory, if the sinewaves are in phase you get a circle on the screen, as the phase changes , it can be measured directly by comparing the X and Y dimensions of the figure, when the sine waves are at 90 degrees you get a line at 45 degrees on the screen.
So get hold of a frequenncy generator, wind up its level and frequency, to say 50 MHZ. Split the output with a restive pad,and view the two signals as above with equal length coax cables, insert an extra length of cable into one leg, note phase change.
using BradtheRad's idea: build/get a fast pulse generator, say a pulse of 100 nS every micro seconds. Split this signal, use one to drive the CRO timebase "sync" input with the time base running at a fast speed (10nS/cm?). The other output should drive a fast LED. Using a Photo diode connected to the Y input of the CRO, you should be able to see a signal which originated from the LED. As the photo diode is moved away from the LED, the signal will decay in amplitude but should also drift across to the right of the screen depending on the propagation delay of the light, as the CRO is being triggered from the drivnng pulse.
Frank

 

[sky_123]

Regarding radio waves, I'm not sure of the best method, you could generate a wideband signal by generating a sharp impulse with a transistor, wait for it to reflect off a surface, and then look for the response. I remember seeing some paper where the received response was detected by using a tunnel diode (because if a tunnel diode has a current so that it is close to the first 'knee' of it's I-V curve, then just a slight increase due to a received signal will cause it to 'snap' to a higher voltage at the same current, and that higher voltage can easily be detected by a comparator. However, this scheme requires a fair bit of circuitry to ensure it works reliably and reset the tunnel diode as soon as the pulse has been transmitted. Not to mention that tunnel diodes are hard to find (when I looked, there was a US company but they wanted a lot, maybe $70 from memory, for a single one. There are low-cost ones from Russia available on eBay, I've got a few but not got round to trying them).
What would be a lot easier is measuring the speed of an electrical pulse in a coax or twisted pair cable, using a time-domain reflectometer, which could be possible with your oscilloscope, and the same generator which you would use for the LED experiment.

 

[sky_123]

Hi Peter,

Curiosity got the better of me, to try to power up the tunnel diode (spent all evening on it, I was dying to see what it would be like),
and I must admit it's very interesting. I'll document it tomorrow (need to draw up the circuit),
but basically I used the diode called 1I305A which was advertised as "Russ*** Mil***** Germanium Fast-Switching Mesa-Alloyed Tunnel Diode"
(I got it from e-bay a few months ago, but I have several so if you send me your address I'll post one to you, since it is for a good cause).
I'm not sure what a complete experiment to measure the speed could look like, but at least this is a way to detect radio waves.
I used a constant current source to power it up. This was the diode:

Basically, it was just attached to the current source and also to the oscilloscope.
This was the current source, it really does need that multi-turn trimmer to fine-tune the current setting:

It was powered up (9V PP3 battery), and then just turn the trimmer and observed the voltage across the oscilloscope go up to about 55mV.
Any further turning, and it instantly 'snaps' to about 0.49V (quantum mechanics is awesome : ). This is such a huge jump that it is
easily observable on a scope (or it could be used to light an LED or something). After that the battery needs to be disconnected and
reconnected to reset the diode.
Anyway, in order to try to trigger it with radio waves, an extremely simple circuit was tried - wire and battery, and hopefully this was
not just a coupled inductor effect. I havn't really experimented much with it:

It works very well from 1-2 feet away (I did not try it any further), which is quite interesting because there was no antenna connected to
the tunnel diode.
This was a couple of pics of the scope output (timebase set to a long and short value - the trigger was a bit higher than the 55mV setting):


Anyway, I suppose more experimentation is needed (time is so short : ( But it was nice to see something that I'd only read about in theory
all this time.

EDIT: Below is the circuit (I used the highest-speed op-amp that I had at hand, but something like AD8061 would
possibly be a far better choice, to get the scope to show a sharper response, although the actual response doesn't matter too much since it's just a time offset for distance measurements). Maybe a simpler discrete current source may be better.

(1% or better accuracy resistors are needed where marked on the diagram). The trimmer allows the current to be adjusted between 5mA and about 15mA (the tunnel diode needs about 9mA to be on the threshold to trigger).

 

[resurgence2012]

Thanks Frank and others. I just posted a reply to this and lost it! But l like the way you are thinking about this. I particularly like the piece of wire and battery to produce an em pulse. That is exactly the kind of thing I was thinking of and it reminds me of the earliest experiments performed by Hertz.

In fact I did save a file of a very nice experiment someone had done with microwaves. I had completely forgotten about this. On the paper I am pretty sure that they draw up both the transmitter and receiver circuits. They were doing low power micro waves over a distance of a few meters. Well really that diode you have described looks exactly like that kind of thing we are talking about. I wonder if there is a way to persuade it to produce a pulse.

Any way, I am in India now and embarking on a trek for ten days starting tomorrow. So I can't play with any apparatus right now.

I would still like to find out how to just hack a cheap transistor radio and display the original carrier wave on a scope - or better than that, just to build one's own circuit with a couple of well chosen transistors.

Cheers

PS. A battery and a piece of wire - I like it!

 

[keith1200rs]

I cannot see where you are going with radio waves. To measure the speed of light you really need a reflective system. With radio waves a phase shift system will be a struggle because you will not be able to distinguish the transmitted and received signals - they will be mixed together. If you try to do a line of sight measurement - transmitter at one end, receiver at the other - you will then have long wires connecting to your measuring system and adding delay so you will be measuring the speed of the radio waves less the speed of propagation of the signals in the cables. So you really need a pulsed, reflective system - think RADAR. With RF that will probably be beyond your capabilities and equipment.

I would suggest an optical method. As Brad said - a laser and reflector and you get roughly 2ns per foot (round trip) so it is easy to get a signal which you can see on a cheap oscilloscope. Retro-reflective tape/sheet does a pretty good job of sending a significant proportion of the light back where it came so you don't need expensive corner cubes and the alignment is not an issue. You can use a road sign as the reflector, or buy a sheet of reflective material. Isolation of the transmitted and received signals is not that difficult in an optical system.

You only need to modulate your light source at a few MHz.

Keith.

 

[resurgence2012]

About a year and a half later, and finally started to make some real progress with this.

I conceded in the end to buying a 50MHz oscilloscope. Built a simple avalanche transistor circuit and sent pulses of IR around the lab. Just used a 9 volt batter, one resistor and a PIN photo detector diode as the receiver. I was able to get a fairly clear measurement of about 75 nano seconds, just comparing the distance between the two pulses as they were rising. The pulses from the long and short path receivers overlap a lot. I would like to
work on finding a way to reduce the rise and particular the fall time. I suspect that what is needed is a high speed amplifier on the receiving end.