Seems like you could make the optical equivalent of a super-regenerative radio receiver, with a laser. Like the super-regen, it would be very sensitive but kinda noisy.
I'm theorizing, based on just guessing, that a long-cavity laser, not a VCSEL, would work best.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
I've wanted to do one for ages, but it's never been the right tool for the job, so I haven't had the excuse.
One problem is how to apply the quench. In a normal super-regen, the gain g oes like the grid bias, which you control directly, whereas in a laser it's set by the population inversion, which depends on the time integral of the pumping rate, and the pumping can't go negative. Good laser media have lo ng upper-state lifetimes, which makes this worse.
So the time response of an optical superregen is depressingly slow. Also, o f course there are fibre amplifiers with near-terahertz bandwidths availabl e off the shelf.
Given how much of early radio technology is directly applicable in optics, the apparent uselessness of optical super-regens is sort of surprising.
Cheers
Phil Hobbs
Instead of relying on decoherence to happen at its own pace, would it be possible to pump the lasing media into an alternate state?
-- john, KE5FX
If you hit it hard enough with an external laser tuned just right, yes. Pro blem is, the result would be an output pulse at the same wavelength as the weak signal you're trying to amplify. It's precisely that sort of accumulat ion of big/hard/expensive stuff that destroys the optical superregen's usef ulness.
Another one is that laser gain isn't directional, so the reverse gain is as large as the forward gain. A RF amp with _minus_ 20 dB isolation would be sort of hard to use too.
Cheers
Phil Hobbs
There have been lots of Carbon Dioxide lasers in front of far IR sensors. Generally something with a very high single pass gain but low chance of goi ng super radiant is used. One of the easiest ways is to detune one cavity mirror by tilt. A second method involves superposition of a third cavity mirror to a two mirror system.
As DC driven Co2 lasers are really sensitive to tube current, its quite po ssible to bias the plasma to near cutoff and use it as a regen.
Co2 have very broad gain bandwidth (100 or more nanometers to a whole micro n, and are more likely to have a lasing mode that can be influenced by an e xternal coherent source then most lasers. Most other common lasers have v ery narrow gain bandwidths that would make it difficult to tune a transmitt er to the receiver frequency.
It also depends on the polarity of what you'd call regeneration.
I have a short, green, very low gain, helium neon laser that runs in a si ngle longitudinal mode. I use it for aligning the mirrors of much larger lasers. When you introduce even the weakest feedback to it, by aligning a t hird mirror with it's cavity, It will start flashing and winking out slowly . If I have a mirror about a meter away or more, and I start to align its beam back in, the green HENE will start to diminish in brightness while las ing without pulsing. However it's lasing bandwidth is very narrow and you would probably have to lock the receiving laser and the transmitting laser using Pound Drever Hal l or an Iodine Adsorption cell to have a communication over any meaningful distance, unless your modulating a reflection from just one laser.
I now await Phil to correct my physics a bit. I'm going from what I've wit nessed happening with lasers I've had on the bench. Including watching two CO2 lasers become injection locked to each other from a miniscule stray ref lection.
Steve
I let my fingers do the walking on a scientific search engine and found four laser superregen papers off the bat, in just a few seconds. Most are early 1990s. One system described in a paper could have its phase changed by as few as five photons.
Steve
+1 on injection-locking and (more often) self-locking. However...
Regen != superregen.
A superregen is an oscillator that runs well above threshold part of the time, and well below threshold otherwise. The gain is modulated by the quench, and the output is a heavily low-pass filtered version of the cathode current. In the absence of an input signal, the oscillation builds up from noise on each cycle of the quench.
There are two situations: the 'linear' superregen, in which the self-oscillation doesn't reach saturation, and the 'logarithmic' superregen, in which it does.
If a correctly-tuned input signal is above the noise level, the oscillation gets a bit of a head start. If it's 1 sigma over the input noise, the oscillation gets going 1 time constant sooner. In the linear case, the peak amplitude is 2.178... times bigger than the no-signal case. In the logarithmic case, the envelope loses 1 TC of insignificant amplitude and gains 1 TC of full self-limited oscillation.
A logarithmic superregen can amplify and demodulate RF signals way down towards the thermal noise floor, producing easily audible headphone output--a gain of at least 120 dB--using a single tube whose maximum linear-mode voltage gain is probably 20 dB.
The superregen has some very serious flaws. The worst is that it effectively samples the input signal for a couple of envelope time constants, once per quench cycle. Thus the effective bandwidth is several times wider than that of an equivalent superhet.
In the RF case, the extreme simplicity of the superregen gives it a place in the sun; in optics, though, it's a much harder sell.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
There were both self-quencing super-regens, and externally quenched ones. A laser could be pumped at a high frequency, and run in gain-switched mode. Just speculating.
Super-regen radios had the same isolation problem: they were receivers and transmitters simultaneously.
It was cool to get Radio Moscow with a 1-transistor receiver.
Gain used to be a very valuable commodity. Now it's cheap.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Yup. There are laser analogues to a squegging oscillator, where the transition rates among the four laser states don't allow a steady-state solution. For instance, there's no way to make a CW nitrogen laser--you get nanosecond pulses no matter how you pump it.
Not as bad, though. You could tell the difference between the grid and plate circuits.
Optical gain is less cheap, but still a lot cheaper than it was. It's all the optical foofaraw you have to add that makes optical superregens useless in practice AFAICT.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
pcdhobbs wrote
Not surea about superregs with light, but I found this recently:
The majority of the super-regeneration laser papers I skimmed either had feedback to the cavity length or an intracavity modulator in the form of an AO cell. One used an AO cell as a Q-Switch for the quench.
All had some pretty sophisticated electronics and very fast photodectectors attached to attain the desired state. Only one looked remotely practical. All would have been really pricy to build and tune.
A multi-pass Co2 amplifier is practical. Non of the other published schemes were.
Steve
A prof. I knew in grad school did intra cavity absorption spectroscopy... basically using the cavity gain to gain up some signal. Is that similar to what you mean?
George H.
e job, so I haven't had the excuse.
goes like the grid bias, which you control directly, whereas in a laser it 's set by the population inversion, which depends on the time integral of t he pumping rate, and the pumping can't go negative. Good laser media have long upper-state lifetimes, which makes this worse. Q-switch the laser? (I can't recall the name of the X-tal that is used.)
GH
of course there are fibre amplifiers with near-terahertz bandwidths availa ble off the shelf.
, the apparent uselessness of optical super-regens is sort of surprising.
We've been playing with some positive-feedback optical gadgets lately: gain-switched diode lasers, modelocked lasers, cavity dumpers, SOAs. All to make very fast optical pulses. They are in the vague class of systems with positive feedback and too much gain that have forgotten to go bang yet. There are lots of electronic equivalents.
In a gain-switched diode laser, one applies a fast-rise high-current pulse and waits a few ns until it fires, makes maybe an 80 ps light pulse. Then you have to back-bias it very fast to kill any tail or secondary blips. Thing is, the laser makers don't say (or know) much about how this works, so you have to somehow get expensive lasers and try it, without blowing them up too often.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
ed.)
Sure, a Pockels cell is one possibility, or reverse current in a diode lase r. But the spontaneous emission is so noisy that there isn't much point unl ess the system is super simple.
Cheers
Phil Hobbs
Always get one to use and two to blow up!
Cheers
Phil Hobbs
We met a diode laser physicist, really nice guy, who also likes to blow things up. Kindred spirit.
So far, it looks like the diodes make their pulse when they have integrated enough coulombs to fire. Driving them harder just makes the blip come out sooner, but not bigger. Maybe a bit less jitter?
We've found all sorts of strange things in laser diodes. Some Voldemort effects (which cannot be named.) Each maker seems to have their own recipe.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
No, I do not mean cavity ring down spectroscopy.
I do mean setting up mirrors for say ten to a few million passes down the cavity and using the gain medium as a non-resonant pre-amp instead of an oscillator.
This is easy to with tilted planar mirrors, remembering the laser tube is a bi-directional amplifier.
If you use spherical mirrors and proper coatings, you can get incredible amounts of passes down adsorption or gain cavities and still have direct entry and exit beams instead of a truly resonant cavity.
Even a single pass down a gain medium can some times be measured, aka the single pass gain test, which can be done either manually or with a lock-in and chopper, provided you can remove the mirrors from the cavity.
Steve
George, I should mention there is a difference between storing energy in the carefully aligned, parallel, resonant cavity vs. just a few passes between tilted mirrors in a gain medium.
Steve
SOAs can have a lot of one-pass gain.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.