LEDs and Solar cells in place of tubes and FETs?

This TED talk is intriguing:

Harald Haas: A breakthrough new kind of wireless Internet | TED Talk | TED.com

Essentially, Harald Haas uses an off-the-shelf LED lamp and solar cell to transmit high-resolution video data by modulating the brightness of the LED.

So I know Gefell uses LEDs and solar cells to clean up phantom power on the way into the mic.

What about using them to transmit the modulated signal exiting the capsule, in place of a tube or FET? I would imagine a miniaturized version of Haas's system would saturate gracefully.

Any thoughts?

???

I'm not getting it.  Do you have any info on the Gefell thing?  We're not talking about an optisolator used for some sort of switching or control in those circuits?

We could definitely amplitude modulate an optical signal to encode audio information (and it has been done), but the primary job and single most important function of tubes and FETs in mic circuits is the impedance transformation.  We need something with the 1-10Gohm input impedance that can drive more conventional stages.  I didn't watch the video yet, but it sounds like they're just using commercial lights and solar cells to do an AM modulation, which don't do anything to meet our needs in a mic capsule amplifier.

Here's a quote on what Gefell is doing from a SOS article. Sounds like an optoisolator to me.

Microtech Gefell's maximum SPL (at 0.5 percent distortion) is 142dB SPL — that's 4dB higher than the Neumann. Part of the reason for the higher headroom is a rather innovative method for generating the capsule-polarising voltage. Apparently, the M930 uses a high-intensity LED powered from the phantom supply to illuminate a photocell, which provides a decoupled high-voltage bias for the capsule.

Click to expand...

source: https://www.soundonsound.com/sos/jan04/articles/microtechgefell.htm

I like the concept greatly. It is an awesome application of light as transmitter. I think it is a long way off even for the large scale applications he discusses. Also if you look closely at the reverse shot of the receiver pedestal the light source is sitting on there is a lot of bench gear sitting on those shelves. With I will wager a bunch of FET based data converter chips. So I am going to guess that this is not a miniaturization ready technology.  But it might be ready for in ear monitor systems instead of fractal antennae technology for live music production. He also unless I missed it does not say the receiver is a solar-cell. It might be but the fact that he said that the receiver only sees what is being sent and not the other light in the room speaks to some kind of filtering either mechanical or digital or both. 

As to the Gefell mic tech. Cleaning up Phantom power no, but decoupled DC to DC conversion very cool and yes cleaner, no switching frequency, and less involved. The SOS article refers to  "high powered LED"  or is it Remote Phosphor Technology and therefore high output? Very cool. No probably just DC booster circuit. I don't think remote phosphor technology existed in 2004.

http://www.futurelightingsolutions.com/en/technologies/Pages/remote_phosphor.aspx

Seems to me that "Talk on a light beam!" was published in Pop Electronics 1959.

The entire internet (except the last mile (10 miles here)) is "lights and solar cells" except confined to glass threads.

OK, so my desk lamp can transmit to my TV. And that's not really a technical problem: my TV remote does the same only low data rate, and faster chips are everywhere.

So I get a spooky movie on the TV and for effect I turn off the lamp. Oops!

If it is a pay-per-view movie the TV has to send-back billing info. Do I have another "60W" LED lamp on the TV?

The hospital: yes, everything is "wired" and this can be wire-less. But much of that stuff has to send-back. Your heart stops. The monitor needs to send a call. OK, instead of a "60W" LED on every gizmo they could use pinpoint LEDs, and perhaps the ceiling light also has the solar-cell to hear the signal. Then the nurse bends over and blocks the beam.

As mentioned, the last-mile is often the worst. You could daydream LED street-lights with sensors to pass data up and down the street. But on MY street there is just one light, I can hardly see it from my house; also I think it is privately owned. And 1 lamp per mile is relatively lush light for many Maine roads; and many non-urban/suburban areas around the US.

""this petrochemical plant -- you can't use RF, it may generate antenna sparks""

Antenna sparks? Didn't spark-gap transmitters go out of style in 1920? High-power outdoor antennas have spark-gaps but they never spark in normal operation; they are to divert lightning strikes which will hit any metallic (or wood, or tree) object.

The idea of cars talking to each other is thought-provoking.

PRR said:

The idea of cars talking to each other is thought-provoking.

Click to expand...

I bet it is already past thought stage...  in prototype systems for driverless trucks etc.

JR

Look at the date
https://en.wikipedia.org/wiki/Photophone

JohnRoberts said:

PRR said:

The idea of cars talking to each other is thought-provoking.

Click to expand...

I bet it is already past thought stage...  in prototype systems for driverless trucks etc.

JR

Click to expand...

BMW and other folks are already looking at car to car Wi-Fi type stuff. 802.11p. For safety, collision avoidance and whatnot. Sounds crazy to me...

What exactly is the transient response of a solar cell?

I don't recall them performing "fast" at all.

A.alden said:

Here's a quote on what Gefell is doing from a SOS article. Sounds like an optoisolator to me.

Microtech Gefell's maximum SPL (at 0.5 percent distortion) is 142dB SPL — that's 4dB higher than the Neumann. Part of the reason for the higher headroom is a rather innovative method for generating the capsule-polarising voltage. Apparently, the M930 uses a high-intensity LED powered from the phantom supply to illuminate a photocell, which provides a decoupled high-voltage bias for the capsule.

Click to expand...

source: https://www.soundonsound.com/sos/jan04/articles/microtechgefell.htm

Click to expand...

As most blurb you can read in ads and magazines, it's not completely false, but also not completely true.
A well-known way of increasing the SPL capacity of a microphone is to make the diaphragm more taut or increasing the distance between backplate and diaphragm. Both these variations decrease the sensitivity of the capsule; the easiest way to compensate the loss in sensitivity is to increase the polarization voltage. That's what generated the HV concept (130v/200V) promoted by B&K and DPA.
Since Gefell did not want to create a source of RFI inside the mic, they chose not to use the quasi-standard DC/DC magnetic converter; that's why they opted for optical energy conversion, which is all DC.
In short, the elevated SPL is not the consequence of the optical conversion, it's the consequence of a capsule with a different optimization, which led to the use of creative thinking for generating the polarization voltage.

BTW, I experienced about 15 years ago with fiber optics analog signal transmission. The optocouplers that have good enough linearity for acceptable distortion are slow and the S/N ratio is not up to professional standards. That's why all the good systems use digital transmission.

> What exactly is the transient response of a solar cell?
> I don't recall them performing "fast" at all.

I am taking it in context of a "TED Talk". Big on concept, thin on technical precision.

I will assume photodiodes/transistors, which are clearly fast enough for EtherNet links, and much more if you treat them nice. "Solar Cell" rolls off the tongue smoother than "photodiodes/transistors", and needs no explanation. The non-geeks get the dim idea, and the deep-geeks see what you really mean.

Probably the real speed-limit is a power driver to modulate the LED lamp. You can just poke in the LED's main power controller, but it typically wants to run 100KHz, so 10,000 bits per second may be as good as it can get. (This sure would cover many of his use-cases.) A series modulator (small shift of main power at more rapid rate) could do MHz or faster (not enough for say CAT-scanning your head in reasonable time).

Another issue: I gather most "room light" LEDs use phosphors. If you have used an oscilloscope, or turned-off a CRT TV, you know they don't go dark instantly. 'Scope phosphors came in many speeds, and IIRC there was one which decayed so fast your eye would hardly see a single-sweep (used with sync photography to capture fast transients). But I don't know how fast they get, and how much compromise that makes with the basic need (a lamp).

While the idea that light gives privacy (does not go through walls or closed doors well) is notable, this is also a big fault (the TV IR remote does not go through the dog well). And the KGB will just use a telescope to snag any small light under door or gap in the window shade. And who looks-around to be sure the room is light tight? Not the doctor getting data on your horrible secret disease. For such work, readily available and auditable cryptology would be more secure against non-KGB/NSA snoops.

PRR said:

'Scope phosphors came in many speeds, and IIRC there was one which decayed so fast your eye would hardly see a single-sweep (used with sync photography to capture fast transients). But I don't know how fast they get,

Click to expand...

Pretty durn fast. Literally a flying dot at NTSC sweep speeds. Otherwise, this wouldn't work:

http://groupdiy.com/index.php?topic=60830.msg771320#msg771320

That's gotta be at least 3Mhz bandwidth, I see no reason why they can't be a lot faster.

Is "persistence" an absolute requirement for a phosphor?

Gene

Gene Pink said:

Pretty durn fast. Literally a flying dot at NTSC sweep speeds. Otherwise, this wouldn't work:

That's gotta be at least 3Mhz bandwidth,

Click to expand...

No. For TV, remanence has to be fast enough that one frame does not stay too long in the next ones, so BW is like a few Hz. And that's asymetrical BW; rise time is faster than fall.
Persistence is analogous to delay; the notion of BW is almost immaterial. Depending on formulation persistence ranges from microseconds to several seconds. Typically a few milliseconds in TV CRT's. Due to persistence, I doubt phosphors can be used for simple analog transmission.

Is "persistence" an absolute requirement for a phosphor?

Click to expand...

Yes; it's an intrinsic property.

> NTSC sweep speeds

Slow.

The dot-rate can be in the MHz. But the glow should persist well into the next frame. Frame rate is 60Hz.

I have an old German book about related topic. Not a lot of use for specific modern interest. But it appears that the process "excites" atoms to a higher energy level, they fall-back to a lower level, with a possibility they never fall back. So kinda a half bell-curve. Much of the fall-back is "soon", but never instant. I don't remember, or can't understand, what controls the fall-back rate.

If you push a TV *at the same dot* faster than the decay, the new bit is blurred by the bit before it. At 60Hz with TV-type phosphors it works out nice. There are faster phosphors. No doubt 300 baud can be done with stuff laying around a phosphor lab. Maybe faster. But kitten-videos and other essential data transfers today tend to need MHz and higher (on "one dot", not the 10,000+ dots of NTSC).

Yes, if instead of an LED light-bulb we used a TV CRT, then a sensor-array could extract MHz data, if it focused very well. (We know it happens to be easier to capture the video signal from the hi-volt video amplifier through EMI; that's how Russians knew what Kennedy was watching.)

With all due respect, I believe both of you gentlemen are missing the point of how that device works.

It has a blank raster, and a transparency with a printed pattern is placed right in front of it. A photomultiplier tube picks up light coming from that general direction, there are no optics involved.

If there was enough persistence to turn the flying dot into a short dash, the horizontal resolution would suffer.

Thought experiment:
Walk down the sidewalk at night past your neighbor's house, the one with the picket fence. Shine a flashlight through the fence as you move along, and your neighbor's house will appear to blink on and off. The flashlight dot is small enough to "resolve" the individual pickets of the fence.

Now try it with a 4' fluorescent tube held horizontal, no house blinking, the light source is too long to resolve the pickets.

From this:

https://en.wikipedia.org/wiki/Phosphor

I found this:

"Blue: Combination of zinc sulfide with few ppm of silver, the ZnS:Ag, when excited by electrons, provides strong blue glow with maximum at 450 nm, with short afterglow with 200 nanosecond duration. It is known as the P22B phosphor. This material, zinc sulfide silver, is still one of the most efficient phosphors in cathode ray tubes. It is used as a blue phosphor in color CRTs."

If I did the math right, that phosphor can handle the digital stream coming off a CD player.

Gene

PRR said:

The idea of cars talking to each other is thought-provoking.

Click to expand...

I think that's a very bad idea, as my old pickup truck would stop and flirt with every Porsche it ran across, I'd never get anywhere.

"C'mon ol' boy, put your driveshaft back in the transmission where it belongs."


Gene

Gene Pink said:

With all due respect, I believe both of you gentlemen are missing the point of how that device works.

It has a blank raster, and a transparency with a printed pattern is placed right in front of it. A photomultiplier tube picks up light coming from that general direction, there are no optics involved.

If there was enough persistence to turn the flying dot into a short dash, the horizontal resolution would suffer.

Click to expand...

This is relevant to cameras, where MHz+ response is a requisite. A photomultiplier does not involve any phosphor at all. It's an electron tube.

"Blue: Combination of zinc sulfide with few ppm of silver, the ZnS:Ag, when excited by electrons, provides strong blue glow with maximum at 450 nm, with short afterglow with 200 nanosecond duration. It is known as the P22B phosphor. This material, zinc sulfide silver, is still one of the most efficient phosphors in cathode ray tubes. It is used as a blue phosphor in color CRTs."

If I did the math right, that phosphor can handle the digital stream coming off a CD player.

Gene

Click to expand...

Indeed, some phosphors are quite fast, but why bother with a non-biunivocal process when there are more adequate optotransmitters?

abbey road d enfer said:

This is relevant to cameras, where MHz+ response is a requisite. A photomultiplier does not involve any phosphor at all. It's an electron tube.

Click to expand...

I'm not following you here. The phosphor in question is on the face of the CRT, we know that.

Try this:
https://en.wikipedia.org/wiki/Flying-spot_scanner

Indeed, some phosphors are quite fast, but why bother with a non-biunivocal process when there are more adequate optotransmitters?

Click to expand...

The suitability of the concept is beyond the "how fast is phosphor" scope of my post(s).  Although it would be fun to play with.

Gene

Gene Pink said:

abbey road d enfer said:

This is relevant to cameras, where MHz+ response is a requisite. A photomultiplier does not involve any phosphor at all. It's an electron tube.

Click to expand...

I'm not following you here. The phosphor in question is on the face of the CRT, we know that.

Click to expand...

I stand corrected. Since the captation process is in base-band, the MHz+ constraint applies. This is not the case for display, as long as excitation time is sufficiently quick.

Indeed, some phosphors are quite fast, but why bother with a non-biunivocal process when there are more adequate optotransmitters?

Click to expand...

The suitability of the concept is beyond the "how fast is phosphor" scope of my post(s).  Although it would be fun to play with.

Gene

Click to expand...

At the time, there were no alternative light sources fast enough for the task, so they just used what they had and learned how to escape the pitfalls. IMO, trying to adapt this technology is akin to making a 21st-century steamcar

> akin to making a 21st-century steamcar

Still trying.