scuba dive compass with return function

Hi all,

I have very little electronic/programming knowledge, so this will be my introduction into a new and exciting (and probably frustrating) area of my life.

My idea is to make a heads up display (hud) compass for my scuba diving, i would like to have a function which would help guide me back to the shot line (bouyed line back to the surface).

I am assuming I can’t use GPS as im under water (usually between 30 - 100 metres)?

I am using a Arduino uno to prototype (because this is what i brought to try and teach my self). And i brought a lm5883 triple axis compass module.

So before I start buying more unnecessary bit I would like to ask for some of your advice.

I have been looking around google for GPS-less navigation, i see lots of intresting things about “IM” (inertia movement?) In which they use a LSM9DS0 board (3 axis mag, 3 axis accel, 3axis gyro).

So to my question, will the LSM9DS0 board do what I would like to achieve?

Many thanks for reading.

Regards

Barry

turtle dude:
I am assuming I can’t use GPS as im under water (usually between 30 - 100 metres)?

So to my question, will the LSM9DS0 board do what I would like to achieve?

To be short ... yes and no to the 2 questions above.

The problem with inertial navigation is drift. Hobby grade sensors have errors that can’t be compensated for or calibrated out. Over a short period of time the error in position will be small but I suspect your dives will last more than a minute or two. There’s a reason INS systems for airplanes cost many 10’s of thousands of $$s and Lord knows what ones used in a nuke sub cost.

Since your goal is to make it back to a vertical line dropped from the surface, might some “beacon” on the line and a receiver carried by the diver work ? Given that sound carries so well underwater, might the beacon be a pinger of some sort ? You carry a directional microphone ?

Thanks for the reply,

Dives are usually between 30-60min (on the wreck), so I’m expecting anything more than about 3-5m drift will be a no no, I get lost as it is without having something point me in the wrong direction.

I already attach a strobe to the line, so its a good idea of yours, i will look into a dual purpose strobe/beacon then maybe just a compass for general use.

If I did attach a beacon to the line could i display it as a mark on the compass?

I hope to have 1/2 an analog compass displayed and a digital display in the middle. When i turn i would like the pointer to stay still and the analog part to move.

I know I’m jumping in much deeper than I should be with a first project (actually this is only the project), but as I’ve often found if your doing something you want/enjoy I’m more likely to stick with it and be much happier when i achieve my goal.

Thanks for your advice, off to google for some research

Baz.

I think you’d need a rather costly gyro to do what you want. they don’t drift too much. Perhaps too hard to do for SCUBA.

People try to use accelerometers to get delta-X,Y,Z then integrate. That has way too much drift.

This would be a common need and if there was a cheap way, you’d see it for sale.

Hi steve,

Thanks for the input.

There is a dive computer on the market that offers these features, I seem to remember it being called the x-deep black (for around £400). It shows on there site a compass dial with an orange blob marked on it diplicting “saved” location.

My project is not to copy a comercial product but to pick my personal best bits form various products to make my own personal dive aid (not for sale/profit).

I feel it must be using some sort of 9dof board like the lsm9ds0, I’ve brefly looked at this and people use a kalfman filter (?) with the IMU to give a more accurate result.

There also use to be a comercial product on the market like what mee_&_mac surgested that used sound think it was called the xios eyesea.

So it looks like there are ways of doing what I would like, just would like to posibly get round hanging more stuff on the line, more batteries, more parts to get lost or damaged.

Thank you for your interest

Regards

Baz.

turtle dude:
There is a dive computer on the market that offers these features, I seem to remember it being called the x-deep black (for around £400). It shows on there site a compass dial with an orange blob marked on it diplicting “saved” location.

I feel it must be using some sort of 9dof board like the lsm9ds0, I’ve brefly looked at this and people use a kalfman filter (?) with the IMU to give a more accurate result.

I looked at the user manual for the above DC and I don't think it's anywhere near as sophisticated as you think it is. You can preset the compass to prompt you for a set of bearings and times. It simply uses it's onboard compass to prompt you to stay on course and uses a timer to know when to swap to the next bearing. It doesn't seem to actually measure distance. You can then do the reverse to get back home. Even in simple compass mode, it seems the "back home" function merely reverses the compass heading to head you back. If you were drifting in a current, it doesn't know it and won't compensate for it. If you swim quicker or slower then planned, it doesn't know it. At least as I read the manual.

http://www.xdeep.eu/files/manuals/xdeep … ual_en.pdf

The digital BLACK TMX compass allows you to set the course (bearing lock) and then navigating it.
In order to set the course, set the compass bearing, so that it is placed in the axis between you and the destination you wish to reach. Then select “Lock bearing” in the compass context menu. The set course will be visible on the compass scale as a round point. Upon setting the course, the compass clock will start to measure the time you swim towards a given direction.
In case of course deviation, the compass will alert you by changing the colour of the bearing and the course indicator. In order to help maintain the direction in case of deviation, the compass will indicate the direction you need to take to return to the correct course:
‘Return’ function allows you to turn on navigation quickly in a reverse direction after you swam some distance with a set course. Therefore, you can take a defined course for a while and come back to same place using the opposite course.
The „Return” function is available in compass context menu.
It is possible to activate the ‘return’ function only when you set the course beforehand. Once „Return” option is selected, the compass will automatically set a reverse course (-180⁰ relative to the previous course). Once you take the new course, compass clock will start counting the time back.

As I said an IMU, at least one you could carry, let alone afford, won’t cut it in the case of currents and varying swim speeds. That then begs the question, how could you do it with some useful degree of accuracy. A 3D magnetometer can (does in the above device) give you a tilt compensated compass. A pressure gauge can give you vertical depth. What’s needed is a way to measure velocity, ideally ‘velocity over ground’ so currents can be accounted for.

Thanks mee and mac once again for your valid points.

It really looks like the only option I have is the beacon on the shot (unless I win the lotto and get myself a nuke sub haha).

I’ve seen some water proof 40KHz transducers on ebay but they don’t give any indication to range. One seller is selling a pack of 5 (they seem to be sender/reciecer in one) i was thinking of mounting 4 on a tube (slightly tilting down to wreck) and the other one on my Hud body?

will try to get more specs.

Thanks everybody

Regards

Baz

Just a thought but the on-body (?on-mask?) unit(s) only needs to be a microphone. That may, or may not, have some benefit as they could be smaller … though perhaps at the “cost” of being less sensitive. My initial thought would be to have 2 mics, one on each side of the mask, so as to form a directional mic. I’d have to run some numbers to see if having 2 mics, spaced at most a head-width apart, brings any real benefit.

Two topics here …

  1. Given the speed of sound in fresh water is 1497 m/s and 1560 in salt water (nominal, neglecting temp, depth, salinity, etc effects) I compute that I’d expect a wavelength of around 1.5" at 40 kHz. So 2 mics (more properly hydrophones) on a mask does look feasible for an underwater acoustic interferometer. You could even implement a 2 axis monopulse network (4 mics) to give you R/L and U/D info. A lot would depend on the hydrophone.

  2. Above I stated …

What’s needed is a way to measure velocity, ideally ‘velocity over ground’ so currents can be accounted for.

As it turns out there's a solution for this already, it's called a DVL (Doppler velocity log). Such devices have been used in underwater drones (AUVs) along with good IMUs to navigate while underwater. But it turns out there's 1 device I found that's made for the diver. I have no idea how much this thing costs or where you'd buy one but for giggles here it is.

http://www.rdinstruments.com/cobratac2.aspx

http://www.rdinstruments.com/images/cobratac_2_1108.jpg

I also found a paper describing a CVL, which like a DVL, use ultrasonic transducers to “track” the seabed and get velocity from that. Supposedly a CVL allows the use of wider beamwidths (DVLs need narrow beams, or so I read) and therefore smaller and lower frequency (and so less $$ ??) transducers. That said my guess is that a package like the above, but using CVL tech, would still be bigger and more $$ and more cumbersome to use than a typical diver would care to spend/use.

FWIW - just how good an IMU is used in AUVs ? Compare the specs of this to the SFE/hobbyist grade gyro and accel specs for drift/bias and repeability and stability thereof.

http://www.kearfott.com/images/stories/ … nu_kit.pdf

Wow, I’ve opened a real can of worns here haha.

mee n mac your sure know your beans! I’m afraid this is going right over my head.

Most people are either happy to reel off or are much better at visual navigating than I am, i assumed (the mother of all f**k ups) that this would be a simply job of fitting a couple senders to a waterproof housing (housing battery and board for ping approx 1-2 pings per sec) and a receiver to diver mounted unit with screen showing distance and direction to senders (accurate to about 2-3 metres).

Regards

Baz

turtle dude:
Wow, I’ve opened a real can of worns here haha.

The start of a project is the time to examine all the options, get all the info on the "table" so the best solution to actual problem is most likely to be the result.

turtle dude:
Most people are either happy to reel off or are much better at visual navigating than I am, i assumed (the mother of all f**k ups) that this would be a simply job of fitting a couple senders to a waterproof housing (housing battery and board for ping approx 1-2 pings per sec) and a receiver to diver mounted unit with screen showing distance and direction to senders (accurate to about 2-3 metres).

I don't know that it's really all that more difficult than you describe. One thing is that you won't get distance w/a simple pinger on the line and a receiver on the diver. Then again I would ask if this is really needed ? If it is then you need a 2 way communication channel btw the pinger and diver, similar in concept to how air traffic control gets distance from an aircraft's transponder (when in flight). I figure if the device tells you right/left bearing to the line and your Dc already tells you your depth then that's enough to get back "home" safely.

As for the device … a lot depends on how bling you want it to be. A heads-up display in your mask is uber bling and thus also uber hard and $$s. A box attached to belt (or ??) is much easier. The simpler the readout on the box, the simpler the design can be (and less $$). In it’s simplest form you could have an LED that says “I hear the pinger” and (perhaps) 2 other LEDs that tell you to “go right” or “go left” or (both on) “go dead ahead”. I might maybe even be able to do that kind of device w/o any microcontroller in the box (though I’m not sure that would be the smartest way to do it).

One thing I envision is that if you are completely lost as to which direction to go in, you’ll need to do a slow 360, aiming the box outward as you rotate until it’s pointed in the general direction of the pinger. Said another way, a simple design means the box can only hear the pinger and do it’s direction finding function when it’s pointed within +/-“X” degrees of the pinger, X to be figured out. If you want X really large, even omnidirectional so you don’t have to point the box, then yes you guessed it … it’ll be more complex (more hydrophones) and thus more $$s. And physically larger too … which is a good time to ask just how big a box would you like this to be ? It’s min size will be dictated by the hydrophones, those you’ll have to buy because I don’t think anyone here is versed in making such things.

One big design driver will be over what range must this work. What’s the maximum distance btw you and the pinger ? This will dictate how loud the pinger must be and how sensitive the box must be. It may also influence the ping frequency as high frequency sounds don’t travel as far a lower frequency sounds (I ran across an equation for that).

In the end I think your basic box would be something like 2 hydrophones, 2 pre-amplifiers for those hydrophones, either some analog circuitry or a microcontroller, a battery and it’s associated circuitry and whatever type of display you want. See, not all that complicated. I could see the addition of a 3’rd hydrophone to guard against certain potential problems and I could see all this eventually integrated into 1 dive computer.

The basic principle involved …

http://en.wikipedia.org/wiki/Acoustic_s … of_arrival

http://www.eng.uwaterloo.ca/~dcbickfo/i … zation.png

http://www.aquacoustics.biz/index.html

I met the developer of this several years ago while it was in an early prototype stage.

It wasnt perfect but it seemed to work well enough. Reflected signals was one of the main concerns, but I think if the transmitter was placed in the optimal place and you took readings away from reflecting surfaces ( walls of wrecks and large rock faces etc) then it should work good enough. Given the amount of time he has taken to get to this stage I beleive it would now be a much better product than before.

This is his personal projects website. The direction finder has since been removed, probably for commercial reasons.

http://www.turtlesarehere.com/

I’m confused … the Amigo product sounds (pun intended) a lot like the transponder device I mentioned above. In some ways similar to the device under discussion, in some ways different. I note that the Amigo seems to depend on a single, highly directional hydrophone vs using the 2 as I discussed above. That can also work, perhaps I was too quick to dismiss that idea ? :?:

Lastly I wonder if the OP is also the same ‘turtle’ person behind the website (and product) above ? I don’t mind that but if my engineering “services”, given here for free, are being used to aid an existing commercial product, I think it fair that I (“we” here) should be told so ahead of time.

Packhorse:
Reflected signals was one of the main concerns, but I think if the transmitter was placed in the optimal place and you took readings away from reflecting surfaces ( walls of wrecks and large rock faces etc) then it should work good enough.

Yup, multipath is beotch to handle in any environment. Radars have had to combat this for decades and there are ways (I hinted at such a problem above) to mitigate the problem. That said, diving within a wreck, and then trying to accurately locate your buddy would seem to be a problem w/o a known solution.

Mee_n_Mac:
Lastly I wonder if the OP is also the same ‘turtle’ person behind the website (and product) above ?..

That I doubt very much.If you look at his web site you can see he is pretty clued up and as I said has been developing this for a number of years. AFAIK he is a Kiwi still living in NZ.

Hi, def. not the same “turtle” for sure, as my electronic knowledge stretches to gcse level (from 15yrs ago) to being able to play games on a nintendo ds. I know not quite the qualifications to start a project like this!

I have been chatting with ‘packhorse’ on dive forums as the other half of this project is a pp02 oxygen monitor. ‘Packhorse’ has achieve an extreme nice unit so was asking if possible to combine the to, his display has space for possibly left/right arrows and poss distance (in whole metres).

Regards

Baz

(Next time im in work - tue, ill try and draw up what i had in my mind, and try to workout how to put up here for you to see)

Good to hear that you’re not him.

‘Packhorse’ might want to search the forums here re: PPO2 gauge. IIRC the topic has come up 2x, though w/o much in the way of a conclusion.

EDIT: You might find this of some interest.

http://www.benthos.com/_doc/main/Brochu … 4_pg_1.pdf

http://www.benthos.com/index.php/produc … 75-locator

This is what I had in my mind???

Regards

Baz.

I will try to look up those threads about PPo2 monitoring, Thanks for the heads up.

Two quick thoughts …

  1. Do you intend to mount the receiving hydrophone(s) on/in the main electronics box mounted on the rebreather ? If so will there be blockage of the direct waves by the diver/tanks/etc ?

  2. The “speakers” on the “pinger” are listed on your diagram as having a directivity (beamwidth) of 60 deg x 15 deg. Is that 60 deg +/-60 deg or +/-30 deg ? If it’s the latter then you don’t have the full 360 coverage I think you want. And I’d ask if the 15 deg is wide enough as well. How far off the bottom is the beacon going to be ? How far out from the shot line are the divers going to be ?

Hi mee n mac,

the a average wreck lengh is around 100-150m long, the skippers normaly shot mid wreck, but of course they could shot either end.

I see the “xios eyesea” had a range of 1000m, i would only be looking a range of 180-200m if possible?

The receivers would be mounted on the out side of the rebreather case (i might even mount them in seperate little housings with leads going back to the control unit to give more flexibility with location).

the transmitter would be mounted approx 5-10m above the wreck attached to the shot line. The sensors i found on ebay were picked at random because thay looked nice? as i have no idea about spec other that a lower freq travels further. (Haha i sound just like my wife who saw a really nice car the other day, when asked what it was, she said a red one?) If the strobe is an issue I could replace it with another transmitter. As you said the spec shows 60° +/-15, i have been trying to search google for 90° beam angles but could not find any, so would it be a case of fitting more transmitters (ie six instead of four) to get the 360° required cover? Also if i had more transmitters if there beams crossed would this start to cause issues?

when possible i put up a pic of my unit with ideas drawn on (i would prob go for mounting both receivers on the back instead of one on the HUD because at times my thick head could block out the sound waves!?)

many thanks

Baz.