SparkFun Forums

[stevech]

Hi stevech,

Thanks for your reply! I'll have to look at the performance figures for the radios I used and see if I can figure out what sort of improvement I'll get.

Perhaps you can correct me here, but this is my understanding which, if this is true then I'll be in good shape. The attenuation of the signal is going to be some sort of a cubic function vs the power of the transmitter. So if i was to increase the power output of the transmitter by a factor of 9, I would get 3 times the penetration into the water.

Is this correct? If it is then having something like 9" of submergance would be more than enough for my needs and I'll be quite pleased.

Cheers

I know a lot about the laws of physics but not so for water. I do know that fresh and salt water differ greatly. And the military uses acoustics or ultraviolet for 100's of meters or more. Many years ago I worked on one underwater acoustic data com project - it used a very long spreading code and the receiver operated at a large negative signal to noise ratio due to the post-detection correlation gain.

I know that RF in free space has odd special cases to the usual frequency dependent inverse square law (which is why we can hear Voyger's signal from outside the solar system)... e.g., at about 20GHz, the wavelength versus rain or fog water droplet size causes high losses in a narrow band of frequencies. But this isn't in a pool or sea!

Navy for years used VLF (very low freq) transmissions from giant shore antennas fed by huge amounts of power, and digtal coding at just a few bits per second. They had other, er, non-disclosed schemes too, as did the Russians, during the cold war era. http://en.wikipedia.org/wiki/SOSUS

[Moggity]

Thanks for your reply stevech and others,

This project has sat dormant for a while now but as it happens I have just re-started it.

Based on my initial testing and your advice I have made some changes to the whole setup so I think for the most part I won't have to contend with any water penetration issues which will be nice.

I did do some experiments with an acoustic based system but I was well and truly beaten by reflections from the environment. I think in the open water an acoustic system would be ideal, but in a swimming pool it was horrible.

Regards

Andrew

[Dave Mueller]

A quick note to answer some of your other questions. Transmitted power falls off as the inverse square of the distance. Twice the distance, 1/4th the power etc. It's easiest to express that in dB. Every time you double (or halve) the distance, the power decreases (or increases) by 6dB.

Also, be careful about which formulas you use for path loss. IIRC, path loss is actually independent of frequency. Most charts state otherwise because they assume a fixed capture area of the antenna - i.e. if you have a 3m dish at both ends, they will have higher gain at higher frequencies which makes the path loss appear to be less. If you use antennas that have a fixed capture area compared to a wavelength, then you'll see that path loss only depends on distance.

If you dig deep enough, you should be able to find a path loss formula that includes the dielectric constant of the medium. Maybe a college physics textbook.

[stevech]

Yep, the higher the frequency the less depth the RF will be received. Conversely the lower the frequency the greater the depth. Try doing this with an AM broadcast radio (550to 1600kHz), you will be able to receive stations from the bottom of the pool but not with an FM broadcast (88 to 108MHz) radio. The US Navy has a radio system to communicate with submersed subs that transmits at around 10kHz.
So find or build a radio that operates at 100's kHz.

Ah, Navy's (soon to be retired) VLF shoreside transmitters are like hundreds of thousands of watts of power. Somwhere on the web there are photos of the antennas.

[markaren1]

I have heard about commercial underwater modems using a multi-tone technique in the high audio range, and maybe ultrasonic ...

Saying that out loud, how about good-old DTMF - 4 bits per signal-pair, readily available hardware. Might be able to manage 40 bits per second, and you will need some error detection, and retry mechanism.

-mark

[stevech]

DTMF is too low freq and too much difference in the low vs. higest freq tone.

[markaren1]

Interesting you should say that. I just did a quick Google on "DTMF underwater" and got this. Are you going to tell NASA or shall I ?

http://www.fastcompany.com/3015636/scub ... -from-1963

BY ALICE TRUONG

Could the key to improving underwater communication be a technology developed for landline phones in the '60s?

At a hackathon earlier this month in San Francisco, oceanographers and scientists from NASA, the California Academy of Sciences, University of California Santa Cruz, and other institutions gathered to figure out how to improve on the current tools of choice for marine researchers: waterproof paper and pencil. The event was hosted by iDive, which is launching an underwater iPad housing this fall. The company is founded by a marine biologist, and the technology originated from a research lab in Saudi Arabia funded by the king.

The winning team of four--all with science and tech backgrounds, including a 15-year-old developer--came up with ScubaTone, an iPad app that relies on old-school technology created by AT&T in 1963 called dual-tone multifrequency signaling, or DTMF for short.

"DTMF was invented 50 years ago for telephone lines before we had rotary disks," explains Roberto De Almeida, a Brazilian oceanographer and developer at Sunnyvale, Calif.-based Marinexplore and member of ScubaTone. "DTMF is nothing more than the sounds a telephone makes when you press the different buttons, each one sending a signal composed of two different frequencies."

[stevech]

DTMF isn't the issue... the problem is in the receiving and sending transducers for water - and the attenuation in water at the lowest vs. highest DTMF frequency. Loud enough, and with the right bandpass filter on the receiving side, one can get some range, as does a beeping sonar. But just a small number of "digits" per second.