AIS GPS tracker for divers

AIS or Automatic Identification System is used by marine traffic to communicate their position to other vessels.
This communication is done via a digital VHF signal on +- 162mhz
This is not to be confused with a similar system DSC / Digital Selective Calling which also works on digital VHF signals.

Information that is transmitted includes GPS based lat/long, speed, heading, vessel name MMSI/Maritime Mobile Service Identity which is a unique identifying number for each vessel much like a phone number.
 Upon receiving this information, an AIS receiver can calculate information such as bearing to the other vessel, closest point of approach, time to closet point of approach etc.

With AIS you can have a transmitter, a receiver or a transponder.

A receiver  simply receives the data and may display this information on its own screen or pass it on to a MFD/chart plotter where it can be displayed. These are becoming more and more common on many VHF radio sets, most of which already have DSC.

An AIS transponder will receive and transmit  information. These are common on most commercial vessels and if you visit the site ( or have the app) marinetraffic.com you can track millions of vessels worldwide. How does this work? Well there are multiple VHF receivers based around busy ports that take the VHF signal and then upload the data on the web. So not only can you track an AIS equipped vessel on a boat via VHF but you can also track it on the internet assuming there is coverage.

Pseudo transmitters. These mark items such as navigational hazards such as rocks. The signal usually comes from somewhere other than the position marked. Such as a VHF repeater station ( like the skytower in Auckland)

An AIS transmitter is a bit more of a rare beast. These transmit information but dont receive any.
Uses include emergency location devices and location tracking of fishing nets etc.
This is where my interest come in! I have been looking for a tracking device for awhile.
I wanted something can you transmit Lat/Long without setting off an emergency event and be rated to at least 100 meters depth.

 The closest available product would be the Nautilus Lifeline which used both AIS and DSC. Its 130 meters rated and very small. BUT you can only active it on the surface and will only transmit on AIS in a distress situation.
The next closest product is the Matsutec HAB120 which is a fishing net AIS transmitter.
Its a lot bigger than the Nautilus mainly because it was designed to be positively buoyant and have a fixed internal VHF antenna. While being waterproof it is not pressure proof.
I decide to get one to try it out.
While I sat in the comfort of my elevated sea view home, one of my fellow Coastguard skipper ( Kerry) took it out for a test.
Once thrown in the water the HAB120  lay on its side. Not the best as far as GPS receiving or AIS transmitting goes. Kerry then duct taped it to a datum marker ( vertical floating marker pole) just above the water line.
I was over 2 nnm ( 4kms) away and picked up the signal on my vessels AIS transponder. It was also about 2nm away from a AIS base station receiver so I could see it on marinetraffic.com. We didn't get to test it  at greater distances than this. But I am also about 2nm away from said base station and both the HAB120 and my vessels AIS transponder have no issue reaching it.
The coastguard vessel motored away from the HAB120 but lost the signal at just over 2NM. Bearing in mind that the HAB120 was just above water level.

OK. so stage two was modifications!
I built a  canister out of schedule 80 PVC and mounted the bared PCB and battery from the HAB120 into it. I removed the reed switch /magnet that is used to turn it on and off and fitted a toggle switch than must be turned on pre-dive. This also allows for a pre-dive check that it is working. I removed the small whip antenna and fashioned an external 1/2 have wave ( 0.9meter) antenna using stainless steel cable.
The idea is that I can clip the free end of the cable to the top of an SMB and let the canister dangle below. This will allow me to deploy the transmitter at any depth I choose as long as I have a spool/reel long enough.
The HAB120 also has a "Mayday" button. AT this stage I have decided not to incorporate this in my build. It would mean an externally operated button on the canister which may easily get knocked during the dive. Anyway, the dive boat skipper can always to a Mayday call if the diver does not return to the boat at the allocated time and inform other vessels of the AIS system in place.

So why didnt I go with an EPIRB system?
EPIRB works differently. Once activated an EPIRB or PLB will receive GPS information then re-transmit this on 406mhz back to a satellite. This will activate a full on emergency response. This is fantastic and definitely the way to go..... if thats what you need. The AIS system  wont set off a full on distress....even if the mayday button is pushed. Someone picking up your AIS distress must manually inform the authorities ( RCCNZ, MNZ Coastguard, police) about the situation.
Interestingly, EPIRB's and PLB's are now available with AIS transmitters. So not only will they  transmit your position back to RCCNZ ( or the equivalent authority) they will also transmit it to all AIS equipped vessels in range!

So what is the range of AIS? This depends on several things.
First, the strength of the transmitter. This is usually 0.5-2 watt.
The quality and the height of the antenna above sea level. Also consider sea height!
The quality and height of the receivers antenna.
What is in the way.

I have found that vessel to vessel, you can get 10NM reliably. With the advantage of about 80 meters of height at home, I have picked up vessels 30NM away.
Nautilus Lifeline claim 34NM for their unit. I am sure its true.....but we need to ask under which conditions did this happen? Given that its only a 1 watt AIS transmitter and a less than ideal antenna transmitting at water level.... I 'd suggest 34NM is probably a one off and definitely not something to expect in real world conditions.

The HAB 180 is rated at 2watt transmit power. I will do some further testing over the next month or so . But I would think 2NM should be a reliable range. Obviously in a distress situation a EPIRB or PLB is a much better choice with a range to the shy and back.


EDIT: HAB180
Battery life- 240 hours-Untested
Transmission interval- 3 minutes  ( Faster when traveling at speed)

UPDATE:

I just got hold of another Chinese AIS buoy. This is the SH-998 from Shunghang.
 This has a couple of key benefits over the HAB120.
1st is that you can adjust the reporting time from 10 seconds to 4 minutes. Obviously the more frequent the reporting the shorter the battery life, but I have bench tested it at 10 seconds and it has last several days which is more than plenty for my requirements. Although when set to 10 seconds the reporting time seemed to be more like 30 seconds when I tested it with my boats AIS transceiver. I will look into this further at a later date.
The second benefit is that the PCB is smaller. At 56mm wide it should fit into most common dive light canisters  or scheduled 80  PVC tube without boring the ID out as I had to do with the HAB120 PCB.
The battery is 8.4v max vs 12.6v but capacity is quoted as 3000mah vs 2000mah (On the battery vs sales material) and battery life is 15days vs 10 days.
 Another benefit here is that I can easily swap out the 8.4v LiPo (?) pack and replace it with 2x 18650 pack of known genuine 3400ma cells. I dount think I can do that with the HAB120 due to space restrictions.
Both units are listed as 8w transmitters, although I have seen a 15w version called the SH-098 that runs 12.6volt (2000mah) and uses an external aerial.

UPDATE:13/07/2018
After a slight leak in my canister the SH998 has a transmission problem. The range is greatly reduced so I assume its an issue with the VHF section.
I have now sourced a different AIS buoy again. This is the ONWA KS33-NTX
There are several KS-33 models
There is the KS-33R which is a dedicated MOB AIS unit. Its nice and small but issues with it include only sending a MOB distress signal, non rechargeable non replaceable battery ( easy solved with custom case), limited power of 2 watt.
KS-33NT is the same as KS-33NTX except it has a small whip antenna built into the casing, while the X has an external 1/2 wave 900mm+- antenna. Both units may come in different battery voltage and VHF power specifications.

Once again the KS-33NT and X are programable for MMSI vessel name and reporting interval. The program cable is a simple serial programmer and I was able to modify the one I got with the SH998.
One of the great features of this is you can set it to report every 20 seconds up to every 3 minutes but every time it reports it repeats 4 times each time 1 second apart. This means if it was in the trough of a wave at the first instance, it may be on the crest for one of the others.
I have not done a in water test with it yet but I have tested it from my house which has an altitude of 80 meters and I picked it up in a vessel from 11.7nm away ( 22kms). Now, the signal wasn't consistent but given that it traveled all that was over land and sea, I am quite impressed. Not only that, but it gets constantly picked up by a marinetraffic.com AIS base station 30km's away with Rangitoto directly in the way ( I am guessing their antenna has some height).
Battery life is quoted at 360 hours (15 days) but this would be based on 3 minute interval polling.
I have ran it at 20 second polling and got over a week.

I have been in touch with a fellow diver today about this. He has been speaking with Nautilus  who make the small DSC/AIS unit. Apparently they may be developing some firmware that will transmit a non distress signal at regular intervals  instead of the MOB distress signal it now sends. Given its size it will be a great tool. Just dont go believing their quote range of 34 miles ( statute or nautical). At 1 watt power and such a short antenna it may have happened once with elevated antenna in perfect conditions but it sure as shit wont happen in real world situations at sea level. I'd be very surprised if it managed to get anywhere close to 1/10 of that.  But 3.4 miles ( 2.9NM or 3.9km) is still pretty good range (if that what it can do).  I'd love for them to send me one for testing to prove me wrong!

Cuda Li Ion battery packs

Im starting a new DPV battery replacement project. In fact Im building 5 packs including replacements for 400 ,550, and 650 Cudas.

A few years ago I built a 96 cell Li Ion pack for my Cuda 550 using Samsung 2600mah 18650 cells.

This time around I am using 84 Panasonic 3350mah 18650's.
This will give me 84 * 3.7 *3.350 = 1041wh of power.

The reason I went for 84 cells is based on the layout of the cells.
I am using the plastic 18650 spacers to build the holder that creates a tidy row and column layout as opposed to a hexagonal lay out that would allow a few more cells squeezed in. It allows for a much tidier and easier build.

Im still awaiting the cells but it looks like I can just still use the stock battery case lid and retaining system. The 400 will be only just high enough while the 550/650 packs leave quite a bit of height using the stock case parts that I can easily add more lead to the pack to make it weight the same as the stock pack.

The 400 pack is the same as the 550/650 except for the height and the way it is held in place.
With the 550/650 there is a radial support in the nose cone that the battery sits against. With the 400 there is a pin in the centre of the cone that supports the pack.

 The new base plate with centre to accommodate 400 alignment rod in nose












Plastic holders for 18650's. 42 cells per level.












Well, the cells finally arrived! All 420 of them!

Several hours of tab welding later I have 2 finished rows of cells.
I you are t building your own pack based on this build take careful note of the layout of the cells and of the routing of the nickle plate in the photo below.
 Both upper and lower level are laid out the same. But the top and the bottom of each pack has the nickel plate welded in a different configuration as shown.
The bottom left is the bottom row of cells and the bottom right is the top row of cells.

This photo below shows the connection points for the BSM
0 is 0 volt or negative, From there the voltage rises by 3.7volt nominal.
 6 is the half way point. Its not clearly shown but this tab is soldered to the bottom of  cell 5.
1`2 is of course the positive of the pack and sits at 44.4 volt nominal (50.4 volt full charge).
EDIT: bottom right cell on top left pack is upside down.

And the assembled pack nearing completion.

This was only just high enough to hold the 2 rows of cells and still fit in a Cuda 400.
Unfortunately for it to work the threaded rod had to protrude through the lid.
This will be sealed up on completion.

The pack weighs in at 1.5kgs less than the stock NiMh 400wh pack. So next job is to make up a weight for in front of the battery pack. While I would like to offset the weight to help with trim, Im not sure this will be possible. But it does give the option of mounting heavy gear on the scooter and removing internal ballast. Perhaps even mounting external ballast at the bast to help compensate for heavy camera gear?

I have managed a burn test and recharge. I had some issues with the watt meter so couldnt get a total watt Hour reading from it but I manage to calculate it out based on 15 minute volt and amp readings.
1007wh on first discharge and 1052wh on first full charge.

Time	Volts	Amp	Amphour	Watt		Watt hour
0	49.4	8.35		412.49
15	47.6	8.06	2.015	383.656	95.914	95.914
30	46.3	7.8	1.95	361.14	90.285	186.199
45	46.5	7.8	1.95	362.7	90.675	276.874
60	45.3	7.7	1.925	348.81	87.2025	364.0765
75	44.4	7.6	1.9	337.44	84.36	448.4365
90	43.5	7.3	1.825	317.55	79.3875	527.824
105	42.8	7.2	1.8	308.16	77.04	604.864
120	42.2	7.2	1.8	303.84	75.96	680.824
135	41.6	7.2	1.8	299.52	74.88	755.704
150	40.9	7	1.75	286.3	71.575	827.279
180	40	6.7	1.675	268	67	894.279
195	38.2	6.6	1.65	252.12	63.03	957.309
210	34.3	5.8	1.45	198.94	49.735	1007.044
			23.49		1007.044
Recharge			23.7			1052

3000wh pack anyone? Should fit in a Cuda 850!

Snoot!

Whats a Snoot?

A snoot is a type of lighting used to  highlight an object for photography.

In this case it will be used for macro underwater shots.

A started off with a basic 1000 lumen underwater video light.

I added an aspheric lens that helps focus the light and a barrel to hold the lens and help control the light beam.

This is my first prototype.

Photos shows a pencil with phone lighting and then using the snoot.

Next I will try building a snoot using a strobe and fibre optics.

EDIT:

Well here are the results.

Lets not judge my photographic skills, but judge the effect the lighting gives the subject.

All photos came out a little underexposed.

All photos have been edited with Microsoft Photos using only the auto enhance function. No cropping or vignette.

My thoughts..... could use a bit more light.  Perhaps a strobe will deliver more or have 2 lights instead?

R14 DPV battery monitoring device

As the name implies this latest project is a battery monitor for a R14 Halcyon DPV. Of course it will also work with any other ( Suex xjoy7) scooter that has a poly carbonate window for the LED to shine through.

The device has 2 PCB's sandwiched together. The top blue PCB is an off the shelf step down voltage regulator that takes the scooters Li Ion battery (29.4v max) and steps it down to 5volt .

The second PCB hold a very simple circuit consisting of a ATTINY85 microprocessor running Arduino. It monitors the battery voltage and drives a WS2812 full colour LED.

The voltage thresholds can be set via code and then fine tuned via the trimpot.

I have set it up as follows

>26 volt Green LED

>23 volt Blue LED

>21 volt Red LED

<21 volt Fast flashing Red LED

For a SLA or NiMh battey voltage thresholds would be lower ( except for the 21volt threshold).

When the device senses the motor is not running for more than 1 minute the device will go to sleep to conserve battery power.  It will wake up every 8 seconds to see if the motor has started again. If the motor is still off the LED will flash very briefly once to show the voltage.

The device is not turned on and off via the scooters power switch but is powered whenever the battery is plugged in.

The only modifcation to the scooter is soldering 2 wires on to the terminal inside the large red Anderson connector. Although with a stock R14  the battery pick up could be done via the smaller charging Anderson connectors so no modification would be required at all.

R14 scooter monitor

Yet another project.

Not happy with the stock battery monitoring system on my R14 I decided to build something that would give a little more information.

I used an ATTINY85 running Arduino.

Basically it looks for a voltage drop across the scooters battery and if it sees this will measure the voltage and display it in a some what modifies Smithers code.
Lots of green flashes= lots of voltage
less green flashes= less voltage
At half capacity the flashes turn red and the less capacity the more flashes.

If the motor has not been run for a few minutes then the monitor will go into sleep mode.
It will wake up every so often to see if the scooter is in use. When it wakes up it will also display its own batteries level via a single quick flash of the LED. Green= good, Blue= not so good. Red= time to charge.

I decided to use a dedicated battery for the monitor so I didnt have to worry about building a voltage step down circuit.
As it is the monitors battery should last a couple of years in sleep mode.

Halcyon R14 battery pack upgrade.

While I really like some of the Halcyon R14 ( Suex Joy) DPV design features I am pretty disappointed in its range/ run time and speed. Not a lot can be done to make it run faster but its not all that hard to extend its range.

The stock battery pack is made up of 20 x 1.2 volt 14 amp-hour NiMh cells wired in serried to give 24 volt nominal ( +-28volt max) which equates to ( 20 x 1.2 x 14=) 336 watt hour. In other words it will run a 336 watt motor for an hour. And as it turns out the R14 is rated at +- 1 hour run time on high, So we can assume it has a +- 336watt motor.

The obvious choice of battery to upgrade it to is Li Ion. Li Ion has a much higher capacity density than NiMh. It is also much lighter which is a slight issue ( more on that later).

I chose to use Panasonic NCR 18650 34B cells.

Quick look at the specs...

Weight 45grams

Volts 3.7 nominal 4.2 max

Amp Hour 3.250 min 3.350 typical 

Width 18mm

Length 65mm

Discharge rate C2 ( Capacity x 2 or 1/2 hour)

To get close to the stock packs voltage I needed to use 7 cells in series. This would give me ( 7 x 3.7=) 25.9 volt or 29.4 volt max. 

The more cells I can get in parallel the more capacity or run time I can get.

The obvious option is to put as many cells in as possible. There is quite a bit of space in the forward compartment of a R14. Getting (13x7=) 91 cells is doable. That would result in a pack of 1130 watt hour and still require extra lead to make it weigh as much as the stock NiMh pack!

The main problem with this is that it would require some modifications to the battery mounting system and this would mean that its not that easy to swap back to the NiMh pack if needed for travel.

The next option is a pack that can easily be swapped out for the stock pack. I worked out that a pack of 70 cells in a 5x7 x 2 rows would work well in this situation and could be swapped out  as easily as a second NiMh pack while supplying 870 watt hours or over 2.5 times stock!

I went for option 3. I wanted to have the battery pack in its own seal container and have it easily swapable for the stock pack.

I figured that I could fit a tube of 134mm OD in between the 3 rods that hold the front weight of the R14 and it could hold 3 rows of cells. This allowed 63 cells (just) but I needed to add a BCM to protect the battery pack. So I needed to reduce the pack to 56 ( 7Sx8P). This would still give me 694 watt hour or over twice the stock pack.  To swap it out from the stock pack all that is required is to remove the front weight ( 3 hex screws) remove the stock battery mounts, put the Li Ion pack in and replace the front weight.

Bill of materials

Panasonic 18650's

Battery Holder Brackets

BCM

Charger
Nickle strip

Canister   132mm OD acrylic tube 5mm wall 250mm long plus end caps

50amp Anderson connectors

Bellow is a few photos of the semi assembled pack.

Next up is to weld the battery pack up using a capacitive discharge welder ( another previous project) and complete the battery canister.

With the use of my lathe I will machine up the end caps to the canister. One will be glued into place and the other will be a push fit/ Oring seal.

16/07/2016 Update:

Today I started the process of welding up the cells to for the battery.

To do this I used a Capacitive Discharge Welder that was another DIY project of mine some years back.
My CD welder is of the more basic variety and consists of only a few components unlike some of the more elaborate types out there some of which are micro processor controlled.
I use a single 2 farad car audio capacitor that charges via my lab variable voltage power supply and a pair of SCR's to discharge it via a pair of copper probes. It seems to work well with sub 0.1mm nickle strip but struggles with the thicker stuff.
The photo below shows one completed row of cells.

10/08/2016 Update:

By luck I did a job for a guy who runs an acrylic fabrication business and he was able to glue/weld the battery case for me and supply a sheet of acrylic I needed for the top lid. I soon added the two cable glands for the cables and finished off the battery.

I did a load test on the battery ( after it was charged 2 weeks ago) and the result was 555wh. This was some what less than expected. After checking the pack I found one string was lower by 0.5 volt that the rest and assumed this is what caused the pack to shut down prematurely.
After a recharge and a balance charge of the pack I dd another load test.

I will try and graph the results later but this is a synopsis of what I found.

No Load voltage fresh charge 29.37volt
Using a Silent Submersion load tester...
27 volt 494 watt 18 amps. According to Mr Ohms this is a 1.5ohm load some what a greater load than the stock R14 motor by 50%.
After 90 minutes I had 21.5 volt,310 watt,14.46 amp and a total of 24.9 amp hour/ 569 watt hour.
The pack shut down 8 minutes later with no final data. But I can extrapolate it out to +-
20 volt, 285 watt  13.8 amp 27 watt hour/620watt hour.

At first glance these are not quite the results I had hoped for.
The pack does not deliver 680+ wh. But it does supply the expected Ah of 26.4Ah.

Regardless I am confident it will supply twice the distance and or run time of a standard R14 NiMh pack.

After the discharge the pack was noticeably hot to the touch. While I had no way to measure the heat I can say it was not so hot that I couldnt hold my fingers on it but it did become uncomfortable after 10-15 seconds. This was pretty much a worse case scenario +50%. It should never get this hot even when run at top speed for 2 hours!

Final work to be done is to fabricate a lead weight to allow for the difference between the NiMh pack and Li Ion pack.

28/08/2016 Update


And finally its complete!

Dues to the light weight nature of Li Ion I needed to add about 2.4 kgs of lead. I also had to replace the salt/fresh water trip weight as the new battery pack fouls with it.

Out with the BBQ and lead pot. The trim weight was easy. I had some aluminium tube that I cut to the right length and then filled with molten lead. Once cool the lead dropped out of the tube.
I then put the lead rod in the lathe and machined it fit. The weight ended up being close enough to match the original trim weight but shorter.  If anything this may slightly affect the trim but I doubt it will be much at all ( nothing a snap bolt on either end wouldnt fix).

The main weight was a little more difficult as it needed to be curved to wrap around the battery pack.
First try I made a mold from wood. The problem here was the wood burnt and off gassed leaving the lead to bubble away as it cooled. This weakened the lead.
Try again but this time using a sheet of aluminium as the base of the mold. Perfect!

I then proceeded to beat the hell out of the 16mm thick lead plate and get some curve into it.
Then weigh it all up to find I was a 200 grams to heavy. So out with the hacksaw to trim off some excess weight and then its all done ready for testing.

I do hope its not too front heavy or I will have to do more cutting of the plate and move some lead around to the centre section.

For now I have taped the weight in place but come Monday I will be off to get a pair of suitable hose clamps!