Wireless device converts 'lost' energy into electric power [Galt's Motor?}

Phys.Org ^ | 11-07-2013 | Provided by Duke University

[Still Thinking]

Hey, if you’re an RFE, I have an RF question for you. Doing industrial RF telemetry, and had a guy tell me the other day “I use 900MHz for near line of sight, 2.4GHz for LOS with no fresnel zone, and 5GHz if I have a clear fresnel zone and want more range than I can get with 2.4GHz.” Does that make any sense to you?

I have no formal training in RF, just self-taught and OJT, but the formulas I’ve seen have the freq-dependent part of Lfs being proportional to LOG(f), so I’ve always used 900MHz for longer ranges. Is there some other factor that often overrides that? Talking 2-20 mile range or so for purposes of this discussion. Is this guy on crack or is he making sense?

[RFEngineer]

You’re right. Free Space Path loss is higher for higher frequencies, however, there is a partial offset - directional antennas of a fixed size will have higher gain at higher frequencies.

Generally, the lower the frequency, the better the ability to to penetrate clutter.

In the presence of clutter above 900MHz RF losses are closer to 1/r^4 than the freespace 1/r^2.

So for point-to-multipoint applications, 900MHz is best.

For point-to-point applications (in clear LOS conditions, highly directional antenns) 5GHz has some superior characteristics/range

I think interference in all unlicensed bands is a wild card that can cause any of them to not function as expected, if at all.

It gets complicated....

[Still Thinking]

So for point-to-multipoint applications, 900MHz is best.

Most of my systems are point-to-multipoint, and most of the rest have the potential to become so, so I tend to have a reasonable gain omni at one end and a reasonable gain yagi at the other.

I think interference in all unlicensed bands is a wild card that can cause any of them to not function as expected, if at all.

True enough, though most of my systems are so far out in the sticks I don't have much background noise.

You’re right. Free Space Path loss is higher for higher frequencies, however, there is a partial offset - directional antennas of a fixed size will have higher gain at higher frequencies.

My philosophy on gain is more like "Use as much as you need to have a comfortable power budget, but no more" than "More is better". You can reach the moon with a laser beam, but you have to be super-careful with aiming, plus if Jupiter aligns with Mars, your beam can easily get obscured. My preference is to be less on the bleeding edge gainwise and get increased robustness in exchange.

Then that leads to the question "What is a reasonable budget?", or does the answer to that question depend on the required data rate? My stuff always works, but it's possible that I've been more cautious with the power budget than I needed to be.

And finally, are you aware of any Ethernet radios packaged industrial style (most of my previous data radios have been serial), bookshelf orientation, DIN rail mount, separate antenna, compatible with 24VDC power via terminals vs. POE? The best fit I've seen so far is the W900 from GridConnect. Ubiquity has some cheap ones but they're not very industrial.