Optical-EME for licensed amateur radio operators

[Baluncore]

I am not saying your suggestion isn't reasonable. I am saying that I don't understand how to do it. Details, please!

Ionospheric reflection is measured by ionosondes, that sweep from about 1MHz to 15MHz, at a rate of about 100kHz per second.
https://en.wikipedia.org/wiki/Ionosonde

For optical ranging of the Moon, consider initially sweeping the modulator signal at 100kHz/sec, from 740kHz to 1MHz in 2.6 seconds. That should provide a sweep that is within the bandwidth of both the optical transmitter and the photo receiver.

The sweep should be generated by a single chip Direct Digital Synthesiser, DDS, that can perform the same sweep accurately every time for everyone. A DDS is composed of:
1. A fixed "sweep-rate" register,
2. that is accumulated into a (delta phase) = "frequency" register, by a sweep rate clock,
3. which is accumulated into a "phase" register, by the fastest clock,
4. that phase is used to look up a sine value from memory,
5. that goes to the DAC, that provides an analog sinewave output.

For square wave optical on/off modulation, the sine lookup and DAC can be ignored, by using the MSB of the phase accumulator as the modulator input to the optical transmitter.
https://www.analog.com/media/en/training-seminars/tutorials/MT-085.pdf
https://en.wikipedia.org/wiki/Direct_digital_synthesis

Analog Devices make a range of DDS chips that could be used to generate the reference sweep, for any good-looking-listener, when clocked by a GPS disciplined reference.
https://www.analog.com/en/parametricsearch/2419#/

DDS is an exciting world of its own. Ask more specific questions to fill in the bits that you don't get.

 

[Hop-AC8NS]

For 532nm it's the doubler xtal that's both the expensive and hard part.

The doubler crystal (typically KDP) requires a laser source to produce frequency-doubled green laser light. A diode pumped solid-state (DPSS) laser diode provides this at microscopic scale in a standard transistor-style package producing about a watt or so our optical power. It is not as nice, nor as powerful, as the rig used by the APOLLO group for Lunar Laser-ranging: it produces multiple output beams in an elliptical pattern, with multiple transverse modes of emission, rather than just one clean TEMoo green output beam. Some optical manipulation required before sending diode laser light toward the Moon.

I mentioned the possibility of repurposing a used tattoo removal laser if a DPSS laser diode should prove to be impractical. But you are correct regarding a DIY frequency-doubling crystal, such as KDP. No way a ham wants to get involved with such esoteric details unrelated to amateur radio. Some would call that "jumping the shark" but I call it a desperate last resort. I would probably just go do something else and give up on optical-eme if that was my only alternative. It's fine if someone else wants to give it a go... maybe a fellow ham in VK-land will do it. It's a big continent, and optical-eme could turn out to be a useful tool "down under" even if comms with the antipodes is impractical because of geometry.

I have no intention, nor the need, to duplicate what was done for lunar laser-ranging. It will be difficult enough to add digital modulation, using microsecond-wide PCM data bits to a short, millisecond, transmit window at a low duty cycle. I don't anticipate needing to leave the laser beam on for extended periods of time.

I also do not expect to receive enough signal reflected from the Moon to perform analog ranging operations using coherent multiplication of swept (chirped) RF modulation of the transmitted and received signals. The Moon is not a static target and the atmosphere is always somewhat in a state of turmoil and changing in density. These two things alone preclude a long-duration RF chirp being useful. This is a communications problem, not a ranging problem per se.

We do need to know an approximate range (expressed as a round-trip signal delay) to align delayed receiver bits with the previously transmitted bits. At one megabit per second bit rates, knowing the delay to within a few hundred nanoseconds should be sufficient. The bit-rate can be lowered, if necessary to accommodate errors in delay measurement, but I think one megabit per second is a reasonable goal.

Approximately one millisecond of transmit window, followed by an identical period of non-transmitting, will be all that is needed for a short, repeated, message to be recovered from the noise after a sufficient number of repetitions. The time in between these two intervals is irrelevant if a bi-static rig with separate transmitting and receiving telescopes is used, but this "dead" time interval does determine the transmitter duty cycle and affects the average power transmitted.

For reasons stated elsewhere, the size of the transmitting telescope is not as important as the size of the receiving telescope. What IS important is the amount of transmitted power (affects the number of photons per unit time that are sent to the Moon) and the receiver aperture area (affects the number of photons per unit time that are detected back on Earth).

Ionospheric reflection is measured by ionosondes, that sweep from about 1MHz to 15MHz, at a rate of about 100kHz per second.

I am familiar with using swept-RF (a chirp) to create, in effect, a narrow-width ranging pulse. I wasn't aware of its application to measuring ionospheric height, which is presumably the purpose of the ionosondes you mention. Fine business that, but I doubt it is relevant to measuring the round-trip delay of signals bounced off the surface of the Moon.

There is only slightly (veeery slightly) more than zero signal returning to Earth from laser radiation sent to the Moon. This is true for any laser power currently available to the public. The APOLLO group, whose intent was the very precise measurement of the round-trip delay time of their pulsed laser beam, did not use chirp RADAR techniques because it is a non-starter. Only statistical photon counting techniques are viable.

Why? Because of the vast distance involved; the poor reflective properties of the Moon's regolith; the severe loss of signal upon retro-reflection using arrays left on the Moon for that purpose; and its propagation as an expanding beam back toward Earth is why.

My proposed communications solution involving PWM pulses, encoding a message that is repeated millions or more times, is one possible, previously demonstrated, way to solve the communications problem.

The measurement of the delay between transmitted pulses and received pulses is required to allow transmitted bits to be aligned with received bits, after accounting for the delay. It does not require picosecond resolution or accuracy.

For optical ranging of the Moon

I am not interested, except peripherally, in optical ranging of the Moon. That is the job for professional astronomers with big telescopes and powerful lasers. I am interested in using the Moon as a passive reflector of communications signals. Folks like the APOLLO group demonstrated "proof of concept" by sending a laser beam to the Moon and receiving reflected information (the round-trip time delay) back here on Earth.

Amateur radio operators have been doing so-called "Moonbounce" comms for over fifty years at VHF, UHF, and microwave wavelengths using Yagi-Uda antenna arrays and microwave dish antennas. I read about one ham who does this with a hand-held antenna and JT8 wsjt-x digital modulation. The only thing "new" about it, or the modulation techniques used to perform it, is the operating wavelength.

 

[Baluncore]

I am not interested, except peripherally, in optical ranging of the Moon. That is the job for professional astronomers with big telescopes and powerful lasers. I am interested in using the Moon as a passive reflector of communications signals.

If you cannot detect the range with an FFT transform gain, then you will not be able to communicate. Ranging is the step that lets you know how well your equipment is working. The product of your transmit sweep, by your returned photons signal, is down conversion and photon counting by another name.

Once you know your signal strength, you can phase modulate the DDS output to encode data onto the sweep. You then use shorter FFTs to extract the phase data.

We do need to know an approximate range (expressed as a round-trip signal delay) to align delayed receiver bits with the previously transmitted bits. At one megabit per second bit rates, knowing the delay to within a few hundred nanoseconds should be sufficient.

Do you really expect to have only one patch of the Moon, about one hundred feet deep, reflecting the signal?
Because the FFT computes cosine and sine, you do not need to establish the edges of the photon-counter boxcar-integrator in time, the FFT does that automatically by returning an amplitude and phase for every possible transit time. That gives you critical information about the surface your ghost reflection is returning from, which in turn, will set your maximum data rate.

I expect the modulation sweep will need to be lowered in frequency to give the demodulator a ghost of a chance against the real Moon.

 

[Hop-AC8NS]

For 532nm it's the doubler xtal that's both the expensive and hard part.

The doubler requires a laser source to produce frequency-doubled green laser light. I have no intention, nor the need, to duplicate what was done for lunar laser-ranging. It will be difficult enough to add digital modulation with microsecond-wide data bits to a short, millisecond-wide, transmit window at a low duty cycle. I anticipate having to leave the laser beam on for extended periods of time for temperature stability, using an external amplitude modulator . Commercial laser diodes miniaturize the optics so I won't need a table-top Nd:YAG laser with frequency doubling crystal for optical-eme.

Approximately one millisecond of transmit window, followed by an identical period of non-transmitting, will be all that is needed for a short, repeated, message to be recovered from the noise after a sufficient number of repetitions. The time in between these two intervals is irrelevant if a bi-static rig with separate transmitting and receiving telescopes is used, but it does determine the transmitter duty cycle and affects the average power transmitted. I want to keep the power low to avoid regulation hassles, but keep the duty cycle high to minimize the number of repetitions. The goal is make optical-eme possible, not distracting.

For reasons stated elsewhere, the size of the transmitting telescope is not as important as the size of the receiving telescope. What IS important is the amount of transmitted power (affects the number of photons per unit time that are sent to the Moon) and the receiver aperture area (affects the number of photons per unit time that are detected back on Earth). I don't believe there is any lower limit to the transmit power if a user has enough patience, but real-time QSOs (like talking on your cell phone) are probably not possible.

Thank you for your comments.

 

[Hop-AC8NS]

Do you really expect to have only one patch of the Moon, about one hundred feet deep, reflecting the signal?

Yes, I expect only one patch on the Moon, a minimum of two kilometers diameter but maybe larger because of Earth's atmospheric divergence of the probe beam, to reflect temporally-tagged photons sent by my telescope toward the Moon. The receiving telescope will use light reflected from that relatively small area, not the full face of the Moon. Note that is temporally NOT temporarily, although I do plan to do this only occasionally. It is the temporal tagging of transmitted photons that allows frame stacking for signal-to-noise improvement and reception of the reflected message.

The depth of the regolith is totally irrelevant. It isn't like water where absorption is a function of depth and turbidity. Even a few molecules of thickness or thinness will reflect somewhere between ten and twelve percent of the incident radiation. Maybe depth is significant for reflecting neutrinos or gamma rays?

I suggest a friendly contest: you try to do optical-eme your way and I will try to do it my way. Winner gets a virtual QSL card via email or this forum. No need for an amateur radio license if you do it with light-emitting diodes (LEDs) instead of diode-pumped light-emitting laser diodes (DPSS LDs).

 

[Baluncore]

The depth of the regolith is totally irrelevant.

I refer to depth in path length variation, if you do not hit a flat spot near the centre of the Moon.

I have the amateur certificate and licence, but am not paid up, because it allows me to have transmitter equipment for modification under Australian law.

 

[Hop-AC8NS]

I refer to depth in path length variation, if you do not hit a flat spot near the centre of the Moon.

Hmmm. Path length variation is important for a signal that has traveled 384,400 kilometers? Light pokes along at roughly a nanosecond per foot, so if there is a variation in distance of the reflection, that works out to about 100 nanoseconds per 100 feet. I don't think I will worry about that.

Distance variations caused by the Moon's spherical surface that is illuminated by my laser could be important if large enough. I asked my machine friend, Grok, about that as it could smear received bit positions and I have not previously considered it. No worries. The maximum variation in round-trip time-of-flight is on the order of two nanoseconds for two kilometer diameter illumination. Even much larger spots up to, say, twenty kilometers diameter only produce about 200 nanoseconds variation in round-trip propagation time. Curvature of the Moon's surface is not a problem.

 

[Baluncore]

Light pokes along at roughly a nanosecond per foot, so if there is a variation in distance of the reflection, that works out to about 100 nanoseconds per 100 feet. I don't think I will worry about that.

200ns per 100ft because in reflectometry, it must go there and back.

 

[Hop-AC8NS]

200ns per 100ft because in reflectometry, it must go there and back.

Ah, got it—the round-trip doubles the timing spread, which could affect bit sync in comms. For APOLLO's small, flat, point-like (from the POV of the Earth looking at the Moon) reflectors, this wasn't an issue. I apologize for my confusion.

I forgot that Special Relativity says it is impossible to measure the one-way speed of light, which—since that speed is believed to be constant in vacuum and not significantly affected by either Moon or Earth gravity—means measurements that effectively measure two-way time-of-flight are also measuring distance.

I learned about (time domain) reflectometry in the mid-1960s when I used it to measure both the length and characteristic impedance of a slightly used surplus roll of Teflon-insulated coaxial cable. I used that cable between my ground-floor barracks-room Novice transmitter and a barracks roof-mounted half-wave dipole antenna. Plus, I had been trained up by the Air Force in how our Ku-band search and track dual radars worked. My only excuse is it's been awhile.

Thanks for the correction on the round-trip—it makes total sense for the observable delay. For bit alignment in optical EME, that spread could introduce smearing, but frame stacking over repetitions should help alleviate it, much like in ranging setups. Grok tells me some of those craters are quite deep, leading to microsecond variations in round-trip delay if I happen to aim my telescope at one of them! That would certainly mess up a one-megahertz bit rate! Two possible solutions: reduce the data rate (does not affect the number of repetitions required, just the "length" of the embedded message) OR aim the 'scope at an area on the Moon with less variation in distance.

I like the second approach because it is easy to aim my telescope (it has about one arc-second optical resolution... not bad for a "beginners" model 'scope) but it is much harder (for me at least) to roll new code to create a shorter message with longer-duration bits. I think I miss not getting a good education in astronomy and physics.

Electrical engineering did pay a little more, but it was years later before I found that out.

Thank you, Baluncore, for your intelligent discussion and replies.

 

[Hop-AC8NS]

Thought I would update this thread and perhaps elicit some more responses. So far, only @Baluncore and @DaveE have participated with interesting suggestions instead of nay-sayings, so I appreciate their contributions to the discussion.

The first problem a new ham experiences is obtaining components for his or her station "rig". This was sometimes a DIY project back in the "good old days" but has largely been replaced by commercial off-the-shelf (COTS) rigs: just plug-and-go. No electronics construction knowledge necessary. Fine business, that. It saves a lot of time getting "on the air" and having fun with the amateur radio hobby.

Optical EME proposes to put the DIY vibe back into ham radio... mostly. COTS is not a boogeyman when it comes to this: I don't expect hams to begin building transmitting telescopes, for example, but it may be necessary to build a receiving telescope perhaps using large Fresnel lenses to obtain a greater photon flux. OTOH, if patience is a virtue, my small transmitting telescope can also be used as a receiving telescope.

This month (February 2026) my wife and I attended the annual Florida HamCation held at the fairgrounds near Orlando. This was our third visit since moving here in December 2016. On one of the two previous visits I scored a set of interlocking aluminum poles, and a canvas bag to carry them, from a flea-market tail-gate vendor. These were originally used to support camouflage netting by the U.S. Army. I thought they might be useful for portable antenna operations such as Parks On The Air (POTA) activations. POTA was started as a one-year event by the ARRL (American Radio Relay League) but it has spread into a separate niche that continues to attract hams.

On the other HamCation visit, I scored a surplus vacuum variable capacitor that was part of the final tuning stage of a military transmitter. I am planning to use it to tune a large (twelve foot tall) magnetic loop antenna for the 80m amateur radio band.

The HamCation flea-market is first-rate as these things go, easily competitive with the flea-market at the annual Dayton Hamvention. But for this third visit my intentions were modest: obtain a one-foot long feed-through with SO-239 connectors on each end. I found and bought one within minutes of stepping off the chartered bus that the Tamiami Amateur Radio Club provided for the trip.

We live in a house constructed with concrete blocks and impact-resistant windows, standard here in Florida because of annual (usually) hurricanes and sometimes tornadoes. It makes it difficult to string an RG-8 size coaxial cable from inside the house to an antenna outside the house. I will have to drill a hole through an outside wall for this purpose, but I don't want to thread coax through that hole. The feed-through allows me to separate the inside-the-house coax from the outside-the-house antenna feed. And I can seal the hole with expanding insulating foam after inserting the feed-through.

The other reason for attending HamCation was an attempt to elicit interest in Optical EME from fellow hams. I passed out flyers to anyone who expressed even the slightest interest. See attached file. So far there has been no response. I did find a small company, Rocket Machine Worx, whose owner and wife had a small, inside, vendor's booth. They expressed an interest in performing job-shop machine work for my Optical EME project. I will eventually need this type of service to construct telescope tracking equatorial mounts and optical component mounts.

Overall, it was definitely worth the trip. However, next time I will reserve an electric cart instead of walking. My wife's admission ticket won an hourly door prize: a nice three-position (A-GND-B) HF-VHF-UHF antenna switch with a built-in lightning arrestor. The club had about twenty-five passengers on the chartered bus, but I was the only member who came back with a door prize.

There was a waning crescent Moon rising above the eastern horizon the morning of the HamCation bus trip. This is exactly the type of Moon phase that I would like to use. The slim crescent provides an eyeball target that I can use to "see" the dark moon phase. My telescope has to be able to track the Moon as it moves across the sky. With an online ephemeris, it should not require visual observation unless it is to pick a relatively flat area (as @Baluncore noted) at which to aim the telescope. But what do most hams know about astronomy? Not much, at least for this one.

Something that I have not talked much about is the time required to detect a weak signal reflected from the Moon's regolith. This is most influenced by (1) transmitter power and (2) receiver aperture area, where more is better in both cases. Only the receiver aperture is easily increased by using a larger telescope. The FAA restricts the amount of transmitter power to avoid blinding airmen who wander into the laser beam. However, the modulation and detection paradigm (frame stacking) allows Optical EME to occur at arbitrarily low transmitter power.

Reducing the transmitter power does is increase the time required to detect the signal. But, if a ham has enough patience, and the right tracking equipment, a two-way QSO can occur over a span of days, perhaps weeks or even months! Some automation is required to do this unattended, but hams are allowed to use remotely operated rigs.

This has legal implications, too. If the laser power is small enough the beam will be invisible to even the dark-accommodated eye! No permits, no regulations required to just "view" the Moon with a telescope. And if you should decide to "view it back" by transmitting a low-power, invisible, green laser beam that cannot be seen (except perhaps as a faint green glow) even if you stare directly into the beam... well what are "the Authorities" going to say about that? I think this would "fly" even in Australia. Patience is the key. Just ask the JPL scientists who "talk" to and receive data from deep-space probes like Voyager.

Thanks to everyone who has read this far!

And thanks to @Baluncore for mentioning Gold codes earlier. I may use one of those for initial acquisition of the round-trip time delay since I doubt that I will be able to "chirp" my laser transmissions for ranging. Grok 4 tells me that an online ephemeris will get me in the ball park of a "distance to the Moon" measurement, but a sequence of Gold code transmissions might narrow it down further if that turns out to be necessary.