[Laser] EME Experiment

Tue Jun 16 21:25:39 EDT 2009

Tim & Yves,
Very ambitious!

Isn't the moon more illuminated during eclipse than at new moon?
Based on my direct eye observations, new moon should be significantly darker.

Jim

----- Original Message ----
From: Tim Toast <toasty256 at yahoo.com>
To: laser mailinglist <laser at mailman.qth.net>
Sent: Monday, June 15, 2009 11:07:42 PM
Subject: [Laser] EME Experiment

Hi all,
Recently Yves, F1AVY and I tried an optical experiment in an attempt to receive echo's from the Moon by transmitting high power pulses from a modified xenon strobe. Despite good weather conditions on both ends, good tracking and an hour of recording time, no signals were detected after analyzing the wav file. 

Here is a link to a PDF file report written by Yves on the experiment:
http://www.savefile.com/files/2130118
or
http://www.savefile.com/projects/808615436

The transmitter used a 10 Joule flash energy and a pulse width of about 1 millisecond. This is a peak power of about 10 kilowatts per pulse. At a rate of 3 pulses per second this equals 30 watts of average power "input". A fresnel lens 22 x 28cm with focus of 33 cm was used to collect this light into a beam of about 0.5 x 1.5 degrees. 

I was never sure exactly how much of the light my lens collected from the strobe tube. A small metal foil reflector wrapped halfway around the tube ensures most of the light comes out on one side in a 180 degree arc. Xenon itself is about 50 percent efficient converting input power to light. Only a certain fraction of that is collected by the lens to be formed into a beam. The size of the strobe tube and the focal length of the fresnel meant the beam would be about one by three or four moon widths. So at best, only about one third or one fourth of the beam would actually hit the moon. On its way, the beam also gets absorbed a bit by the earth's atmosphere - maybe about 10 percent. All this added together (or subtracted) determines the actual power that was hitting the moon. My guess is about 1 joule per pulse made it to the moon or about 3 watt-seconds, which illuminated one entire hemisphere. 

On the receive end, Yves used a 24 cm telescope and PGP detector with a low pass IR filter (700 nm) and the program Spectrum Lab with its time domain scope. Both the transmitter and the receiver time domain scope were synchronized with the GPS system 1 pulse per second time signal. The time domain software allows the signals to build up over time as long as they remain "lined up" or synchronized with the highly accurate GPS time standard. The exact delay time of the lunar echo does not matter, it is the fixed phase relationship between the echo's and GPS which allow the integration to work and noise to be subtracted while the signal amplitude is added over time. 

I think using the time domain/pulse detection approach is similar to the way CCD devices can integrate light over time, building up a picture from a darkly lit scene, But instead of actually having the photons collect in the CCD cells over time, we have pulses repeating regularly over time and being integrated by the software. CCD's are "two dimensional" devices being literally an array of many cells arranged in a grid pattern. The PGP detector or any other single element detector is "one dimensional" in the sense there is only one of them. And when repetitive pulses are detected with the time domain mode, it is also one dimensional in the sense that all the pulses are treated as a single pulse, which builds up in amplitude over time. In order to get this "single pulse effect" high timing accuracy is required. So GPS is a natural choice for this. If the transmitted pulses or the receiver software synchronization drifts more than a few percent of the
pulse width, the pulses will no longer "line up" properly and wont be added together as well. If the timing drifts by more than the transmitted pulse width over the time it is integrated, then it will not be detected at all due to this "blurring" effect. I think this is a similar effect as image movement with CCD's, and results in a weaker, trailed/blurred image. Most noise signals are rejected this way. With the time domain setup the noise levels seem to be really low considering all the background earth-shine present. Is it possible the time domain setup is actively subtracting this noise?? and not just averaging it??

Although this experiment was conducted at a time (May 27th 20-21 utc) when the moon was near perigee and any changes in echo delay time were minimal. At other times in its orbit, the moon's distance and the resulting delay time changes more rapidly. At some points between apogee and perigee, the time delay can change by 200 microseconds or more in an hour. This is enough to "blur" or weaken the pulse signals integrated over an hour, although not enough to make them disappear completely since the transmitted pulse widths are about one millisecond and the delay time change is only about 20 percent of that. 

The moon's rounded surface can lengthen a pulse up to about 11 milliseconds but it is dependant on the receiver's field of view. Assuming the transmitter is illuminating a large area or the whole lunar surface, small RX fields of view limit the length of the pulses lengthened by the "echo depth" effect. One odd effect a small field of view can have when combined with manual tracking is; as the FOV drifts toward the edge of the moon or from edge to center, the delay time of the received signals will also drift - by up to that 11 milliseconds of echo depth. So the phase of the received pulses will favor the delay time of whichever part of the moon that is in view by the receiver. This could displace the position of the received pulse by up to 11 milliseconds relative to the GPS standard. With manual tracking, a periodic re-positioning of the receiver FOV is made. I think this would introduce a cycling phase shift that the integrating software would not
allow for, and treat it as a non-synchronized signal to be rejected as noise. So instead of a relatively constant phase relationship over time between GPS and the signal data, the phase would cycle through a certain range of time delays depending on which part of the moon has reflected the signal - drifting from zero delay near the center to 11 milliseconds near the edge then rapidly back to zero as the scope is re-positioned. For a 1 millisecond pulse, this would "smear" or blur the signal severely when using the time domain scope. Having a small FOV and either manually tracking more frequently or the use of a clock driven scope mounting would probably eliminate this problem or reduce it to well within the pulse width used. Having longer pulses with more noise in them may not be as much of a problem as having shorter pulses which drift in phase rapidly due to manual tracking errors. Having a FOV as wide as the whole dark part of the moon would combine
all the echo's into one long pulse or a series of pulses, which is OK but there would be more noise too...

Since the powers used in this experiment were relatively low, (3 watt-seconds?? - about the same as a 3 watt CW laser with a half degree beam) i still think it is worth a try to do other experiments using the same setups at higher power/better tracking or wider RX FOV's and other setups like frequency domain and other modulations/pulse rates etc. I think a modest laser array could easily out-perform my strobe especially since the lasers can be focused in a much tighter beam on specific areas of the moon and take full advantage of small receiver FOV's and good tracking. 

If anyone would like to try an experiment and set up a schedule with me just send me an email direct and i'll see what i can do. The best times to try this for northern hemisphere locations is in the March through June moon phases between new and first quarter phase. Other times of the year favor north-south and southern hemisphere locations or other moon phases in the northern hemisphere. I definitely plan to try this again during the next total lunar eclipse in December 2010.(not visible from Europe) I think there are a couple partial eclipses between now and then but they don't look very promising.

http://www.aladal.net/toast/gpsx.html

best regards,
tim toast
toasty256 at yahoo.com

______________________________________________________________
Laser mailing list
Home: http://mailman.qth.net/mailman/listinfo/laser
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:Laser at mailman.qth.net

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html