Connect with us

Man-over-board alarmsystem

Discussion in 'General Electronics Discussion' started by isis91, Nov 23, 2012.

Scroll to continue with content
  1. isis91

    isis91

    9
    0
    Nov 23, 2012
    Hi! I'm new to this forum and im hoping for help. We are sig guys who are going sailing for a year. I want to build a system that sets of an alarm if anyone was to fall overboard. Im new to working with microcontrollers etc. but i have tried to read up on the subject and i think ive come up with a working idea. I would very much appreciate any help or ideas you can give me. I can program C fairly well.

    The system has to be extremely low-cost as we are on a very tight budget. I'm thinking that everyone onboard will wear a radio transmitter that constantly transmits a "code" to a base unit. If the base unit hasnt received one of the six individual codes in 15 seconds it sets of an alarm.

    Im planning to do it this way, with everyone wearing a module that transmits to a base. If you have abetter idea of how to do it, im all ears.

    The module

    Ill get this microcontroller as it is cheap and has a low energymode that seems suitable for my needs. There are both cheaper and more eneryeffecient uC ive found, but the atemel programmers are so much cheaper that they get a lower total anyhow.

    The uC will be connected to this radio transmitter. Ive found that the transmitter would require an antenna of about 30cm length. Can the antenna be any normal piece of copper wire? can i have it coiled? i want to wear the final unit as you would a wrist watch.

    I will power the components with a buttoncell battery at 3v and 1Ah. Hopefully this will get me about a year of batterylife i think. I guess. I hope?

    Would i need a crystal to get the uC runnig? would i need a voltage regulator? Im thinking that it doesnt matter very much that the unit transmits at perfectly regular intervals, nor that the voltage is stable.

    Fynally im going to encapsulate it all in epoxy to make it waterproof and resistant to shocks. Could this get me overheating-issues? The battery will be outside of the epoxi and sealed in some other, unknown, way.

    The base unit would consist of the same kind of uC and this radio receiver. It would also be connected to this MOSFET. The MOSFET will be connected to the boats 12v system and a beeper of some kind. I picked a P-gated MOSFET to have it sound the alarm if the base malfunctinos for some reason.

    I was thinking that the base needs a crystal, and as i dont have any powerissues here it might aswell be a 20MHZ one. I havent quite figured out how to run the uC and receiver from the boats 12v system yet. Ideas?

    Are there any flaws in this design that you can see? Ive heard that crystals need to be paired with capacitors, is that correct?


    I posted this question in a swedish forum but i ended up getting barrated about why harnesses should be worn. So ill say raight away that we have plenty of methods for staying onboard, and weve discussed this through and through. We will ofcourse be using harnesses aswell. What i really need help with is the electronics.


    Thanks in advance,
    Joakim.
     
  2. duke37

    duke37

    5,364
    772
    Jan 9, 2011
    Welcome Joakim,

    I would like to make some observations.
    The module would not need a microprocessor, you need a transmitter switched on for a short time and modulated with a tone to say which one it is. The time between each transmission should vary randomly so that it will not always clash with another unit.
    It should be crystal controlled to set the frequency accurately. Crystals do not take much power but the transmitter will, depending on the range.

    With a 300mm aerial, it will be operating at about 400MHz and the aerial can be any conductor. Some handheld telephones wind the aerial in a coil and coat with a flexible sleeve (known as rubber ducks). These are less efficient than a straight wire.

    Water will absorb the radio waves and conductive sea water will hide the transmitter so the transmitter should be kept dry until required.
    A metal boat will also hide the transmitter (see Faraday Cage)

    Receiver
    This is where your micro will be useful. You will need some circuitry to decode the tone signals and reset that particular channel, then switch on a warning device if the tone has not been received in the alloted time.

    The receiver should be crystal controlled to match the transmitters.
     
  3. isis91

    isis91

    9
    0
    Nov 23, 2012
    Thank you for your answer. I think I understand most of what your saying. That i wont need a uC for the modules are great news, as that will cut the cost of the project by almost 20%. Im not sure how i would go about modulating the signal nor setting the time between transmisssions randomly without using a uC though. what component should i use to control the transmitter?

    The signal doesnt have to go further than at maximum fifteen meters, so hopefully an antenna cioled around the wrist will do. The boat is made froom steel though, do you think that would completely block the signal? i could mount the receiver outside, but it would be annoying to have it sound the alarm every time somebody goes below deck. Maybe if i put to antennas on the receiever, one outside and one inside?

    I want to make the transmitter modules waterproof so that they dont get destroyed by spray or rain, hopefully this wont block the signal?
     
  4. duke37

    duke37

    5,364
    772
    Jan 9, 2011
    I have gone about as far as my limited knowledge will take me.
    The steel boat will block the signals completely but signal may go through the windows. Windows with electrical wires for heating will block the signal.

    You may not need a random generator, consider one transmitter sending a tone every 5 sec, another at every 6 sec, another every 7 sec, will this work?
    A tone and transmitter trigger can be made from a single chip e.g. 556 or 4093.

    You may be better off with two receivers, one fed from outside and one from inside. The uC could look at both. Two antennas to one receiver can give signal null depending on signal strength and phase.

    Plastic waterproofing should not affect the signal much.

    I have never used a uC so cannot help here.
     
  5. BobK

    BobK

    7,682
    1,688
    Jan 5, 2010
    Despite what others have told you, these modules do need something like a microcontroller. There are basiscally a wireless data link that sends serial data, not an analog transmitter that can be used to send different frequency tones.

    Bob
     
  6. isis91

    isis91

    9
    0
    Nov 23, 2012
    Hmm, im a little bit at a loss here now. On the upside, while googling on the 555 chip, i found a transceiver which i think will work better for my project. its also cheaper.

    http://imall.iteadstudio.com/wireless/im120606002.html

    Using a transceiver instead of a transmitter doesnt make much difference if i have to use a uC anyhow, but it does save a lot of noise as i can have them transmitting one at the time, at the order of the main.

    Are there any microcontrollers except for the AVRs that can be pogrammed using a 15 dollar programmer?

    If i have the base asking for responses i should be able to have the uCs sleeping until they are activated by the transceiver to save power, no?

    So were im at now is a module consisting of a transceiver, a microcontroller and a battery. No need for crystals, voltage regulators or anything else?
     
  7. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    I see nothing but a design that will cry wolf more then detect actual incidents, and thus it the end making it nearly useless as the alarms will constantly be ignored as most will be false...

    I'm sure with enough money and planing a decent system could be permanently installed ship/boat wide with an array of antennas that blanket the entire ship to eliminate dead zones but this explodes a simple idea into a massive project...
     
  8. isis91

    isis91

    9
    0
    Nov 23, 2012
    would you care to elaborate? What you are describing would be useless ofcourse. Do you mean that the transmitters wont be able to send through the boat? wireless networks work very well in the boat, and that works on 2,4 ghz does it not? why would these be so much worse? there are not metal walls inside the boat, and there are plenty of windows aswell as a main hatch where i plan to put the base unit.
     
  9. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    For this system to be effective there has to be zero to no false alarms, with each false alarm response time and how serious people take the alarm will diminish... There is a ton of issues that will attenuate a radio signal that have to all be taken into consideration that could cause a false alarm...

    Almost all substance will attenuate the signal, glass is great for looking through that doesn't mean that radio waves pass right through... There are many formulas to glass windows some attenuate the heck out of radio waves, in fact most current 'energy efficient' window glasses do a dandy job of attenuation... Foil insulation in walls or construction can also wreak havoc with radio waves... As well as almost all electronic devices causing attenuation... The list could go on and on, and last but not least even environmental things like humidity and water have effects as does the position of human bodies between the transmitter and receiver...

    In theory all this can be overcome with the proper design but it's likely to take some trial and error and the complexity of the design will increase with each compensation...

    You linked to a 32 mW @ 434MHz radio... That is apples to oranges to a 2.4GHz radio that is generally anywhere from 50-4000mW of power...

    Comparing this proposed system to wi-fi signal performance is flawed at the core on protocol as well... Wi-fi doesn't toss up an man overboard alarm if a packet is dropped it simply sends it again, this all happens behind the scenes and it happens all the time... A wi-fi network dropping 10, 20, 30 or even 40% can easily be the norm in a house/board or whatever... Now consider dropping 1 to 4 out of 10 pings of your main overboard system...

    You can of course compensate with faster ping times, error correction and the want not but that again increases the complexity of the system with each step of correction...
     
    Last edited: Nov 24, 2012
  10. isis91

    isis91

    9
    0
    Nov 23, 2012
    Im sorry for the confusion. I actually wrote to replys, one before the one that has been posted. for some reason dod the first reply dissapear. It went something along these lines:

    I found this 2,4ghz transceiver which is both cheaper and i think will work better for my purposes. I should be able to get less noise aswell as save energy on the module by having them sleep until they are activated by the base. Does any of you know of a microcontroller except for the AVRs that can be programmed using a 15 dollar programmer?

    as far as i can tell, i wont be needing anythin but the uC, the transceiver and a battery for the modules?

    I'll try to paint a picture of the boat in paint to give you guys as clear an idea as possible of the setting. We are on a very limited budget and i dont want to get into a project that wont work. As you say, false alarms will destroy the whole point of the alarm.

    The windows of the boat are plexi, and the main hatch will almost allways be open. This alarm is not meant for stormy days, but rather the calm days when your on low alert.

    As i see it, from a very uneducated point of view, even 90% lost packages would be okay if you accept a response time of 30 seceonds. it cold be implemented so that when 15 seconds passed since the base spoke to module A, it will start pinging it. If it has had no response for 15 sec, it sounds the alarm. Or am i way off track here?

    Could one long antenna along the whole length of the boat do the job?

    Thanks for all your help guys!
     

    Attached Files:

  11. BobK

    BobK

    7,682
    1,688
    Jan 5, 2010
    I wouldn't worry about sleeping your uController when the receiver is drawing 12ma.

    Bob
     
  12. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    You have now turned the transmitter into a transceiver, and this means it can't go into sleep it has to be up full time waiting for incoming 'pings'...

    As Bob hinted you just destroyed run time, your battery is now measured in hours...
     
  13. wingnut

    wingnut

    242
    9
    Aug 9, 2012
    I once saw a man-overboad system on Dragons Den. It worked on a hydrophone and each person wore a unit activated by immersion in water. This unit then broadcast through the water to the hydrophone on the boat.
     
  14. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,497
    2,838
    Jan 21, 2010
    Yeah, that and a personal EPIRB.
     
  15. isis91

    isis91

    9
    0
    Nov 23, 2012
    Ok, so you think i'm better of sticking with plain transmitting units? Will the tranceiver be using that kind of power even when its not receiving any data?

    What would happen if i used the original 400ish Mhz receiver with an antenna that went along the length of the boat? Or the 2,4Ghz tranceiver for that matter, but only using the sendfunction for the modules.

    Battery time will be an important factor, as having to remove the modules every couple of days will be very inpractical. As long as i get over three weeks i will be fine, because i doubt we'll do any longer runs than that. While ashore the modules can be left by themselves recharging.

    Thanks
     
  16. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    Not really as the two way communication provides the ability for better error correction and detection...

    Read the datasheets, they contain a wealth of information... There are different power drains for different states... A transceiver can not be shut off as it has to be ready to receive at all times, if it's off it isn't going to receive... The power when acting as a receiver is less then the power as a transmitter, you have to factor it all in...

    Can't say, you would need to test real world to get real world data...

    Same as above...

    In that case you really need to reconsider your battery options... You need to finalize a design and then measure it's overall current drain... Use this as a base for your battery mAh rating... I suggest you choose a battery that is 2-3x the mAh rating that you factor to compensate for real world loses...

    If that is the case design it to last 6-8 weeks or longer...

    You bring in another factor, rechargeable batteries will degrade over time thus shortening run times... You also now have to factor in that possibility that a unit will be deployed without a full charge or a weak charge that will trigger false alarms...

    IMO when you are designing a system like this you need to take in all the WORST case scenarios and build the device to function at an acceptable level at worst case scenario... This means A LOT of thought/planning and trial and error testing...

    You also stated earlier that this is a fair weather device, now you state deployment might be three weeks... There is no way to even hazard a guess of fair weather over a 3 week period at sea, so designing it for fair weather is a huge omission of the potential issues faced real world...
     
  17. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,497
    2,838
    Jan 21, 2010
    If I were you, I would have each person have a personal epirb onboard. Then I would have a receiver *always* turned on monitoring the epirb broadcast frequency.

    If you miss the bloke going overboard, or can't find them, other people will be able to help.

    I think that falling overboard in mid-ocean is a good enough reason to set of a personal epirb. The fact that you can be alerted and do your own rescue is an added bonus.

    edit: the problem with personal epirbs is that apparently they don't come in versions that are automatically activated on imersion.
     
    Last edited: Nov 27, 2012
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-