6 meter takes 20 ns to bridge for light. Electricity will take twice the time in a worst case scenario. The thing is, I can sum up a whole lot of other things that will have a far bigger influence on the signal than travel time. So lets make a short basic list without going into the more detailed aspects of high frequency signal transmission.
First of all to define the starting condition, we're working at 340 MHz and a maximum voltage change of about 400 mV (give it an additional 100 mV if it's an old variant). Any length of an electrical conductor shows some degree of capacitance and inductance. This will affect the maximum raise and fall times you can achieve and be the main cause of the cable's frequency response. If you've never seen a sinc function before you're going to see it come out of that cable most likely if you manage to hook it up to an oscilloscope. The thing is, that shouldn't matter much due to the way the protocol is built. The clock line will ensure that the receiver will only look at the data when the sender really tells it to. Else it's just going to ignore the noise.
Another important issue you'll run into at 340 MHz is the skin effect. This is where the story gets funny actually. They will often claim that gold plating will improve performance. The skin effect will actually have less effect on a poor conductor. And since we have a bunch of cables in the same area you can also start looking at proximity effect. But given the fact that you'll be using DACs and or ADCs probably combined with some high bandwidth semiconductor amplifiers designed to operate beyond 340MHz (keep in mind these components their bandwidth is often limited on purpose) that's going to be the least of your worries since neither of those components require large currents to do their job.
Not to mention 15 meters is pretty much the standard length ensured these days without any noticeable performance decrease. People often forget this isn't all that bad. Gigabit ethernet will only do about 24 meters using twisted pair cables. So I'm not sure what your concern is, it's not cause it might take a nanosecond more for a signal to rise (it'll be a lot less actually) that you're going to run into huge problems. This is the main reason to switch to digital transmission.
The thing is, there were pretty good attempts at similar things in the late 80s, early 90s. In fact there was a digital desktop (as in a real desk) where you could project things like a calculator and it attempted to find where your finger was using a camera. But computer technology wasn't mature enough to provide the processing power required to do it well. Not to mention projector technology wasn't that great either back then. Keep in mind that the current microcontrollers often outperform the best computers of those days in many aspects. Microsoft their solution isn't all that elegant if you take that into consideration as microsoft did what they couldn't back then. Throw a large chunk of processing power at it.
Well he does have a fair point. This is one of the basic things most electronics students will eventually build in college as a simple exercise and they then drop it due to the lack of real world applications. I've seen people who haven't studied any electronics build similar things and make 3D scanners out of them. It's not like it's particularly hard either. It's harder to drive a small LCD display than to build one of these...
Fact is that it's annoying to see 30 year olds still typing with their index fingers.
But actually, teaching them how to touch type would be better. That way they'll be done typing their homework faster so they can do more of it:P
You have a few options to do this.
If you use radio waves that happen to bounce back, in that extremely unlikely case I hate to break it to you but radar has been around for 70 years.
If you use RFID tags to track motion, that has been done before as well on several ways (signal strength, timing, etc...).
If you use small that sends out radio pulses, has been done before as well.
Oh yeah, fun addition. Look out what frequency you use, cell phone operators already have usage rights on just about every frequency out there. And they love to sue people for disturbing their network (even if they don't use that frequency all that much).
Slashdot Mirror
6 meter takes 20 ns to bridge for light. Electricity will take twice the time in a worst case scenario. The thing is, I can sum up a whole lot of other things that will have a far bigger influence on the signal than travel time. So lets make a short basic list without going into the more detailed aspects of high frequency signal transmission. First of all to define the starting condition, we're working at 340 MHz and a maximum voltage change of about 400 mV (give it an additional 100 mV if it's an old variant). Any length of an electrical conductor shows some degree of capacitance and inductance. This will affect the maximum raise and fall times you can achieve and be the main cause of the cable's frequency response. If you've never seen a sinc function before you're going to see it come out of that cable most likely if you manage to hook it up to an oscilloscope. The thing is, that shouldn't matter much due to the way the protocol is built. The clock line will ensure that the receiver will only look at the data when the sender really tells it to. Else it's just going to ignore the noise. Another important issue you'll run into at 340 MHz is the skin effect. This is where the story gets funny actually. They will often claim that gold plating will improve performance. The skin effect will actually have less effect on a poor conductor. And since we have a bunch of cables in the same area you can also start looking at proximity effect. But given the fact that you'll be using DACs and or ADCs probably combined with some high bandwidth semiconductor amplifiers designed to operate beyond 340MHz (keep in mind these components their bandwidth is often limited on purpose) that's going to be the least of your worries since neither of those components require large currents to do their job. Not to mention 15 meters is pretty much the standard length ensured these days without any noticeable performance decrease. People often forget this isn't all that bad. Gigabit ethernet will only do about 24 meters using twisted pair cables. So I'm not sure what your concern is, it's not cause it might take a nanosecond more for a signal to rise (it'll be a lot less actually) that you're going to run into huge problems. This is the main reason to switch to digital transmission.
The thing is, there were pretty good attempts at similar things in the late 80s, early 90s. In fact there was a digital desktop (as in a real desk) where you could project things like a calculator and it attempted to find where your finger was using a camera. But computer technology wasn't mature enough to provide the processing power required to do it well. Not to mention projector technology wasn't that great either back then. Keep in mind that the current microcontrollers often outperform the best computers of those days in many aspects. Microsoft their solution isn't all that elegant if you take that into consideration as microsoft did what they couldn't back then. Throw a large chunk of processing power at it.
Well he does have a fair point. This is one of the basic things most electronics students will eventually build in college as a simple exercise and they then drop it due to the lack of real world applications. I've seen people who haven't studied any electronics build similar things and make 3D scanners out of them. It's not like it's particularly hard either. It's harder to drive a small LCD display than to build one of these...
Fact is that it's annoying to see 30 year olds still typing with their index fingers. But actually, teaching them how to touch type would be better. That way they'll be done typing their homework faster so they can do more of it :P
In that case you are probably using something similar to RFID. In other words, it already exists.
You have a few options to do this. If you use radio waves that happen to bounce back, in that extremely unlikely case I hate to break it to you but radar has been around for 70 years. If you use RFID tags to track motion, that has been done before as well on several ways (signal strength, timing, etc...). If you use small that sends out radio pulses, has been done before as well. Oh yeah, fun addition. Look out what frequency you use, cell phone operators already have usage rights on just about every frequency out there. And they love to sue people for disturbing their network (even if they don't use that frequency all that much).