There's a problem, though, with using more power. You increase interference with everybody else while making a small improvement with your intended recipient.
A directional antenna helps you when you receive as well as when you transmit. If you need to serve an area, you can still benefit from an antenna that concentrates radiation in a pancake shape so you don't waste power transmitting straight up.
High power conflicts with sharing.
Stock antennas are pretty unimpressive and leave a lot of room for improvement.
Antennas are cheap, especially if you build your own, and they don't burn up battery power.
Antennas work in two directions. An antenna with a better pattern improves your range for both transmit and receive. An amplifier on one side of a link doesn't help you hear the other side any better.
Antennas with radiation patterns that match where you need the network reduce interference coming in as well as interference going out.
The aviation industry's been looking into millimeter-wave systems for landing in fog. You can "see" a completely fogged-up runway with millimeter-wave radar.
Like brunes69 said, it's the same frequency (2.4 gighertz) as lots of other unlicensed devices, along with the leakage from your microwave oven.
WiFi uses clever spread-spectrum techniques in place of raw power to get its range. A typical WiFi card sends a twentieth of a watt to its antenna, and the inverse square law applies, and you're usually not holding a WiFi card up to your head like you do with a cordless phone.
That frequency is two to four thousand times higher than AM broadcast, and a couple of dozen times higher than FM broadcast. Each photon is proportionally more energetic, though still not energetic enough to break chemical bonds the way a photon of light can. Every time you change frequencies that much, you can get fascinatingly different interactions with matter.
"There's more to science than mistreating animals. But frankly it's the part I like best." -- Dilbert
If you want better precision than chocolate or ants provide, you could print a grid onto thermal fax paper, moisten it so it won't catch fire, and put that into the microwave oven. See http://www.amasci.com/weird/microexp.html#demo. The idea is from JE Slone.
>you can try repeating the measurement many times
Of course after each measurement you have to do something with the chocolate.
"Mmmm, I think we should increase the sample size some more. Got any more milk?"
Some ovens have the fan-type stirrer you describe, but not all. Others solve the problem of hot and cold spots by putting a turntable at the bottom which rotates food in and out of the hot and cold spots.
The experiment is for an oven with a turntable. The article talks about taking the turntable out and putting the chocolate on something non-rotating.
The answer to your question about cavity size and standing waves is I Don't Know. In fact, I've wondered for a while why microwave oven designers don't use the same trick as recording studio architects, and make the walls non-parallel.
It's possible that they have to make the cooking cavity resonant in order for the magnetron to "see" the right kind of load on its output. But that shouldn't matter much as long as there's food absorbing the microwaves.
Also, you can slightly change the frequency of a magnetron by varying the power supply. The resonant frequency of the cavities is a starting point, but the output can be "pushed" a little above or below it.
The power supply in a domestic microwave oven is designed to be cheap, not to be stable. It varies all by itself. So does the output frequency.
Sometimes ham radio operators will try to use microwave oven magnetrons as transmitter components. Hams have a spectrum allocation around 2.4 GHz, and it's tempting to think about getting high power cheap. One problem is that the output frequency jumps and drifts so much that you need to add circuitry to meet legal standards for a clean signal. We're talking fractions of a percent, close enough for measuring chocolate with a ruler but far too bad for radio communications.
Yes, and pulsing does make things even worse. The magnetron will behave a little differently during startup transients.
Oh, and don't play with magnetrons unless you can do it without killing yourself. The power supplies are immediately lethal and the kilowatt of RF isn't good for you either.
I'm certainly not an expert on this, but I have taken the American Radio Relay League's class on emergency communications for ham radio operators.
Remember Tip O'Neill saying all politics is local? All emergencies are local too. A widespread disaster like Hurricane Andrew is, in practice, a bunch of local emergencies ('cause you're sure not getting any spare firefighters from the next town over, and the bridge is out anyway).
Most of what ham radio operators do in emergencies is short-range, immediate traffic ("Our generator has 2 hours of fuel left. Please deliver to loading dock left of Nth Street. Don't take Mth Street, it's blocked by an accident"). The most effective ham radio emergency service teams are tightly integrated with the served agency. One typical assignment might be riding along with ("shadowing") an emergency official and keeping a communications link going to the emergency operations center.
The exception is health and welfare traffic, the "I'm all right" and "I'm slightly injured but receiving good care and am safe in a shelter" messages to out-of-area relatives.
Emergency response agencies could benefit no end from having a backup TCP/IP network on which they could send email, use web-based order forms for supplies, or even use a regular telephone hooked up to VoIP. Email especially, because they *love* having written communication.
Seattle is a very good place for people to be hardening the 802.11 network against disaster. It's an earthquake area, which periodically has near-apocalyptic events like the 1964 Alaska quake or the 1960 one in Chile. Every few years a winter storm knocks out power so widely that repair crews have to be imported from Canada.
Seattle Wireless is doing something potentially very practical here.
Thanks for a well-written, informed and factual post.
One question. Isn't the capacitance problem going to be about the same regardless of the semiconductor material? A vacuum tube lets you reduce capacitance by putting many centimeters of vacuum between the electrodes.
Back in the Cold War, the biggest ham radio organization got some time on a military EMP simulator and tested this idea. There's an account of the results at http://www.qsl.net/n9zia/emp.html, along with references to the original and quotes from informed speculation about post-apocalypse communication.
If you're in a hurry, the executive summary is that there are two kinds of EMP mechanisms to think about. One is in the near vicinity of the blast. "Near" as in "near enough that it's the least of your prolems". The longer-range type is like a wide-area power surge. A semiconductor device that would die from static electricity is *more* likely to survive an EMP-induced surge than a tube is, because the lower impedance lets the current surge through.
The proposed frequency range for broadband over power line runs up to 80 MHz. There's a radio astronomy allocation at 25.55 Mhz, a couple more around 38, and another at 73.
Over-the-air TV channels 2-5 lie in the affected range.
There's also a subtler problem. That transformer near your house blocks high frequencies. To get broadband in and out of your house, the power company will have to buy and install bypass circuits to divert high-frequency signals around the transformer. OK so far...
But what happens next? Right now that transformer is isolating you from your neighbors X-10 system, your neighbor's cheap switching power supplies, your neighbor's made-in-Sweatshopistan dimmers, and the home power line network on the next block.
Bypass that transformer and every one of those sources might as well be plugged into your house.
Some of the WiFi channels are within the amateur radio allocation, governed by Part 97. They could have run a powerful tight beam legally by complying with the rules for the amateur radio service.
If both ends were run by someone with a ham radio license, and if they used channel 1, and if they didn't attempt communication with the general public, and if they didn't use obscene or indecent language, and if they turned off encryption, and if they didn't forward data for third parties from other countries that don't have third-party traffic agreements with the US, and if they identified transmissions with their callsigns every 10 minutes and at the end of each transmission, and if they didn't transact any business or communicate on behalf of an employer, then it could have been legal.
Not only did they need to know what to do, they had to do it to close tolerances.
You can ruin the pattern of an antenna by putting in irregularities a fraction of a wavelength high. One wavelength at WiFi frequencies is about 12 centimeters. The usual rule of thumb is to be smooth within 1/10 of a wavelength.
Like glonoinha said, they're better than hard disks for high-vibration applications.
A mountaintop router for a wireless WAN might not have reliable utility power. There's another niche. Systems that have to run on battery or solar power are much happier if they don't have to spin a platter. CF cards have the nifty property of talking ATA with a cheap hardware adapter, so they're drop-in replacements for hard drives.
Slashdot Mirror
There's a problem, though, with using more power. You increase interference with everybody else while making a small improvement with your intended recipient. A directional antenna helps you when you receive as well as when you transmit. If you need to serve an area, you can still benefit from an antenna that concentrates radiation in a pancake shape so you don't waste power transmitting straight up. High power conflicts with sharing.
The paper said they used a DC power supply, and only speculates about what would happen with a real el-cheapo microwave oven power supply.
This matters because you can shift the frequency of a magnetron slightly off nominal resonance by varying the power input.
Microwave ovens ship with the crudest imaginable high-voltage source and the magnetron voltage isn't even approximately constant.
If the oven's frequency is bouncing around the spectrum, other users may not be able to stay out of the way.
Stock antennas are pretty unimpressive and leave a lot of room for improvement.
Antennas are cheap, especially if you build your own, and they don't burn up battery power.
Antennas work in two directions. An antenna with a better pattern improves your range for both transmit and receive. An amplifier on one side of a link doesn't help you hear the other side any better.
Antennas with radiation patterns that match where you need the network reduce interference coming in as well as interference going out.
The aviation industry's been looking into millimeter-wave systems for landing in fog. You can "see" a completely fogged-up runway with millimeter-wave radar.
Like brunes69 said, it's the same frequency (2.4 gighertz) as lots of other unlicensed devices, along with the leakage from your microwave oven. WiFi uses clever spread-spectrum techniques in place of raw power to get its range. A typical WiFi card sends a twentieth of a watt to its antenna, and the inverse square law applies, and you're usually not holding a WiFi card up to your head like you do with a cordless phone. That frequency is two to four thousand times higher than AM broadcast, and a couple of dozen times higher than FM broadcast. Each photon is proportionally more energetic, though still not energetic enough to break chemical bonds the way a photon of light can. Every time you change frequencies that much, you can get fascinatingly different interactions with matter.
"There's more to science than mistreating animals. But frankly it's the part I like best." -- Dilbert
If you want better precision than chocolate or ants provide, you could print a grid onto thermal fax paper, moisten it so it won't catch fire, and put that into the microwave oven. See http://www.amasci.com/weird/microexp.html#demo. The idea is from JE Slone.
>you can try repeating the measurement many times Of course after each measurement you have to do something with the chocolate. "Mmmm, I think we should increase the sample size some more. Got any more milk?"
Some ovens have the fan-type stirrer you describe, but not all. Others solve the problem of hot and cold spots by putting a turntable at the bottom which rotates food in and out of the hot and cold spots.
The experiment is for an oven with a turntable. The article talks about taking the turntable out and putting the chocolate on something non-rotating.
The answer to your question about cavity size and standing waves is I Don't Know. In fact, I've wondered for a while why microwave oven designers don't use the same trick as recording studio architects, and make the walls non-parallel.
It's possible that they have to make the cooking cavity resonant in order for the magnetron to "see" the right kind of load on its output. But that shouldn't matter much as long as there's food absorbing the microwaves.
Also, you can slightly change the frequency of a magnetron by varying the power supply. The resonant frequency of the cavities is a starting point, but the output can be "pushed" a little above or below it.
The power supply in a domestic microwave oven is designed to be cheap, not to be stable. It varies all by itself. So does the output frequency.
Sometimes ham radio operators will try to use microwave oven magnetrons as transmitter components. Hams have a spectrum allocation around 2.4 GHz, and it's tempting to think about getting high power cheap. One problem is that the output frequency jumps and drifts so much that you need to add circuitry to meet legal standards for a clean signal. We're talking fractions of a percent, close enough for measuring chocolate with a ruler but far too bad for radio communications.
Yes, and pulsing does make things even worse. The magnetron will behave a little differently during startup transients.
Oh, and don't play with magnetrons unless you can do it without killing yourself. The power supplies are immediately lethal and the kilowatt of RF isn't good for you either.
I'm certainly not an expert on this, but I have taken the American Radio Relay League's class on emergency communications for ham radio operators.
Remember Tip O'Neill saying all politics is local? All emergencies are local too. A widespread disaster like Hurricane Andrew is, in practice, a bunch of local emergencies ('cause you're sure not getting any spare firefighters from the next town over, and the bridge is out anyway).
Most of what ham radio operators do in emergencies is short-range, immediate traffic ("Our generator has 2 hours of fuel left. Please deliver to loading dock left of Nth Street. Don't take Mth Street, it's blocked by an accident"). The most effective ham radio emergency service teams are tightly integrated with the served agency. One typical assignment might be riding along with ("shadowing") an emergency official and keeping a communications link going to the emergency operations center.
The exception is health and welfare traffic, the "I'm all right" and "I'm slightly injured but receiving good care and am safe in a shelter" messages to out-of-area relatives.
Emergency response agencies could benefit no end from having a backup TCP/IP network on which they could send email, use web-based order forms for supplies, or even use a regular telephone hooked up to VoIP. Email especially, because they *love* having written communication.
Seattle is a very good place for people to be hardening the 802.11 network against disaster. It's an earthquake area, which periodically has near-apocalyptic events like the 1964 Alaska quake or the 1960 one in Chile. Every few years a winter storm knocks out power so widely that repair crews have to be imported from Canada.
Seattle Wireless is doing something potentially very practical here.
Thanks for a well-written, informed and factual post.
One question. Isn't the capacitance problem going to be about the same regardless of the semiconductor material? A vacuum tube lets you reduce capacitance by putting many centimeters of vacuum between the electrodes.
>They are immune to EMP.
Back in the Cold War, the biggest ham radio organization got some time on a military EMP simulator and tested this idea. There's an account of the results at http://www.qsl.net/n9zia/emp.html, along with references to the original and quotes from informed speculation about post-apocalypse communication.
If you're in a hurry, the executive summary is that there are two kinds of EMP mechanisms to think about. One is in the near vicinity of the blast. "Near" as in "near enough that it's the least of your prolems". The longer-range type is like a wide-area power surge. A semiconductor device that would die from static electricity is *more* likely to survive an EMP-induced surge than a tube is, because the lower impedance lets the current surge through.
The proposed frequency range for broadband over power line runs up to 80 MHz. There's a radio astronomy allocation at 25.55 Mhz, a couple more around 38, and another at 73.
...
Over-the-air TV channels 2-5 lie in the affected range.
There's also a subtler problem. That transformer near your house blocks high frequencies. To get broadband in and out of your house, the power company will have to buy and install bypass circuits to divert high-frequency signals around the transformer. OK so far
But what happens next? Right now that transformer is isolating you from your neighbors X-10 system, your neighbor's cheap switching power supplies, your neighbor's made-in-Sweatshopistan dimmers, and the home power line network on the next block.
Bypass that transformer and every one of those sources might as well be plugged into your house.
Some of the WiFi channels are within the amateur radio allocation, governed by Part 97. They could have run a powerful tight beam legally by complying with the rules for the amateur radio service.
:-)
If both ends were run by someone with a ham radio license, and if they used channel 1, and if they didn't attempt communication with the general public, and if they didn't use obscene or indecent language, and if they turned off encryption, and if they didn't forward data for third parties from other countries that don't have third-party traffic agreements with the US, and if they identified transmissions with their callsigns every 10 minutes and at the end of each transmission, and if they didn't transact any business or communicate on behalf of an employer, then it could have been legal.
Simple, really
Not only did they need to know what to do, they had to do it to close tolerances.
You can ruin the pattern of an antenna by putting in irregularities a fraction of a wavelength high. One wavelength at WiFi frequencies is about 12 centimeters. The usual rule of thumb is to be smooth within 1/10 of a wavelength.
Pretty good for aluminum foil.
Like glonoinha said, they're better than hard disks for high-vibration applications.
A mountaintop router for a wireless WAN might not have reliable utility power. There's another niche. Systems that have to run on battery or solar power are much happier if they don't have to spin a platter. CF cards have the nifty property of talking ATA with a cheap hardware adapter, so they're drop-in replacements for hard drives.
In case anyone wants a sample, here's something I snagged from a club newsletter: DE NW6P. FB ROBERT. RST 579 579. OP TOM TOM.QTH NR SF NR SF. HW? BK