slightly off topic but you guys will know this

[Roger Feeley]

I work in data communications and from time to time I give talks about how it works. I would like to build a "Can LAN" composed of two cans and string. I can build the electronics to produce noise at each end but I also want to be able to passively sniff it.

So imagine two soup cans with a guitar string in between. You talk into one can and listen at the other. If you put a guitar pickup next to the string, could you pick up even an indication of sound? I don't need to pick any particular note. I just need an indication that the string is vibrating.

The issue as I see it is that the string is not being used traditionally. In this case the vibrations are being transferred along the length of the string.

[Adam Herman]

a pickup should work. talking in a can will produce both compressional and longitudinal waves, even if you cant see them such as on a guitar.

my family had the first "phone" in Manton, MI. I have it in a box somewhere. It is a cardboard diaphragm in a round wood frame with a wire tied to the diaphram that went next door to the bank they owned. You would pull a string to ring the bell hanging above it and yell into the cardboard. Apparently it was not very good and they usually just used it to ring the bell and someone would run over and find out what was going on.

[paul cottingham]

As long as the string is metal, it should work.

[Dave Richards]

We used to make soup can phones with two soup cans and a spool of kite string. If you pulled the string tight enough it worked just fine.

If you could find a couple of old speakers, you could wire them together to make an intercom. When I was a kid, a friend and his dad rigged up an intercom with some wire and a couple of speakers. One end was in the attic of the two story house and the other speaker was in the basement. No batteries required.

[Jim Becker]

A pickup should work for a "man in the middle" tap if the transport is a steel wire. :) Of course, Quality of Service is going to be hard to enforce with a Can-LAN. LOL :D :D :D

[Malcolm Schweizer]

I believe what you want here is not a magnetic pickup, but rather a piezo pickup; the latter having a greater frequency range, but it is a contact pickup and therefore would best be suited to pick up the vibration of the can itself. If you want to pick up from the string, however, you could build a sound box which when placed under the string would harmonically resonate and then have a sound hole like a guitar and you could install one of these http://www.musiciansfriend.com/accessories/dimarzio-the-black-angel-piezo-acoustic-soundhole-humbucker-pickup?rNtt=piezo%20pickup&index=9#productDetail which combines a piezo and magnetic pickup.

[Alan Rutherford]

Since I don't have any money riding on this and no one's actual tin-can-phone experience seems to run much deeper than mine (one or two tin can phones a long time ago, a fair amount of electronics experience) I'll say no, I don't think the guitar or piezo pickup will work. That would be because the vibrations are running longitudinally in the wire. The only vibrations perpendicular to the wire (and able to be picked up) would be incidental/accidental and random.

I would expect the "phone" to work with insulated wire stripped only where it attached at the ends. You would not expect the guitar pickup to detect anything that way.

You could possibly tap the line by soldering a second wire somewhere along the middle of the first, perhaps with a few inches of overlap. Then add a third can.

I invite you to prove me wrong.

[Malcolm Schweizer]

Since I don't have any money riding on this and no one's actual tin-can-phone experience seems to run much deeper than mine (one or two tin can phones a long time ago, a fair amount of electronics experience) I'll say no, I don't think the guitar or piezo pickup will work. That would be because the vibrations are running longitudinally in the wire. The only vibrations perpendicular to the wire (and able to be picked up) would be incidental/accidental and random.

I would expect the "phone" to work with insulated wire stripped only where it attached at the ends. You would not expect the guitar pickup to detect anything that way.

You could possibly tap the line by soldering a second wire somewhere along the middle of the first, perhaps with a few inches of overlap. Then add a third can.

I invite you to prove me wrong.

I am imagining the setup being the same as the old-school string and can, which only works when the string is pulled tight and therefore resonates like a guitar string. I'm not getting where you are using an insulated wire- are you picking up electronically in your scenario? Is the wire not taught? Maybe tin can phones have evolved since I was a kid.

[John K Jordan]

I suspect anchoring one or both ends of the cable to a piezo transducer (or putting one somewhere in the middle) should work, but that might not be as passive as you want for your talks. (The piezo method is how some "ancient" delay line memories worked, at least in the one I salvaged from a suitcase-sized calculator.)

But could you use a microphone on or near the end of the receiving can to pickup transmission or is that cheating? Or if you want to sense linear tension waves in the cable itself without contact can you mount a small disk or plate somewhere on the line, oriented perpendicular to the line and sense sound or motion in that? Laser light reflected from a mirror mounted perpendicularly could probably sniff tiny longitudinal motion. In fact, I'll bet you could optically sense something as small as a tiny crystal glued to the cable or even a knot tied in the cable. With a magnet on the cable, a Hall probe or even a simple coil might work.

I'll try to call my friend Joe on Saturday. He's the guy I go to for questions about optics, magnetics, sound, and electronics since he knows everything.

JKJ

[Alan Rutherford]

I am imagining the setup being the same as the old-school string and can, which only works when the string is pulled tight and therefore resonates like a guitar string. I'm not getting where you are using an insulated wire- are you picking up electronically in your scenario? Is the wire not taught? Maybe tin can phones have evolved since I was a kid.

I'm not picking up anything - I'm assuming the wire is not vibrating like a guitar string. Insulation would make it even more difficult for the guitar pickup to detect anything but if I'm right, the "phone" would work anyway.

Everything has evolved since we were kids.

[Roger Feeley]

A pickup should work for a "man in the middle" tap if the transport is a steel wire. :) Of course, Quality of Service is going to be hard to enforce with a Can-LAN. LOL :D :D :D

Jim, For what I have in mind, QOS is largely moot. In data communcations, you deal with two states: 1 and 0, mark and space, Plus voltage and minus voltage. Here, our states would be noise and silence. Frequency doesn't matter.

My idea is to use black and white balls as bits. Black ball would make noise and white ball would not. A mechanical device would cycle the balls into a scanner at 1 second intervals (1 baud) where an Arduino would read the balls and produce the noise (or not). That reduces a process that generally happens too fast to see to something visual and tactile.

So the balls provide the states, the Arduino produces the noise and the sound travels across the the wire to another can where another Arduino UART reads the noise and produces the character. Easy right.

Now the really sweet part. I work for Frontline Test Equipment, a division of Teledyne Lecroy. We make sniffers. I would not want to create this demonstration rig without being able to sniff it. That's where the pickup comes in. It represents a passive sniffer. It would read the vibrations in the wire without interfering with the transmission. The pickup talks to a third Arduino which produces a USB output to what we call a Data Source which talks to our sniffer.

[Roger Feeley]

My idea would involve using an actual guitar string which is not insulated. I don't pretend to understand electric guitar pickups but I assume they work on capacitance. The vibration of a plucked string has more amplitude than a string carrying sound between two soup cans. hence the question to this group.

[Roger Feeley]

Maybe what I need to do is find get a couple of soup cans and find a friendly music store. I buy a guitar string, put it between the cans and we hold it close to a guitar pickup and see what happens.

[Alan Rutherford]

Find a hardware store and buy 100' of any cheap steel wire. I'd bend a short "L" in each end after you poke it through the hole in the can and then solder it to the can. I bet you won't have to worry much about keeping it taut. In my limited experience, a guitar pickup is not sensitive enough for what you want to do. I'd look for the piezo device, a turntable cartridge, or build something for the Arduino. The laser was not a bad idea.

You might have an un-tappable communication system. That's not what you want, of course.

I'd love to spend today working on this but my wife would kill me. Keep us posted.

[Rick Albright]

Do you think a piano string would be better or does the guitar carry sound better?

[Jim Becker]

Jim, For what I have in mind, QOS is largely moot. In data communcations, you deal with two states: 1 and 0, mark and space, Plus voltage and minus voltage. Here, our states would be noise and silence. Frequency doesn't matter..
For "this" project I would agree with you and I said what I did in jest. But even in pure data networking (which is a part of the business I've been in for decades) QoS is absolutely important because different applications have different expectations relative to data flow and latency across the network. It used to be that folks could get away with "just throwing more bandwidth" at the problem, but ultimately, application demands got to the point that important data flows were still getting stepped on with apparent randomness and causing issues with their associated applications. Kudos to those of us in the telecom world who primarily focus on real-time communications (voice and video) because our application requirements have basically forced QoS on everything and everything has actually benefitted from that. (IMHO, of course. :) ) And now with SDNs (Software Defined Networks for those not in the business), it gets even more interesting with how we can segment, direct and isolate traffic to insure that not only is performance top-notch, but absolute security can be enforced including what my particular line calls "stealth networking". (If you watched the Sochi Olympics, you experienced that without even knowing it)