Announcement

Collapse
No announcement yet.

Transducer driver/receiver

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Transducer driver/receiver

    Now that we get somewhere with transducers, I was thinking about the electronics driving them.


    Before I go to the drawing board I would like to check some basic principles.


    The general setup is as follows:


    1) Echo data is digitized in fish and ONLY the characteristics I really need is send topside.


    2) The heart of this is around a PIC controller with 2 A/D ports.


    Using PIC's will simplify electronics a great deal.


    Might need 2x PIC incase I get into overruns but at the cost of $15,- per PIC that is no problem


    3) Transducer gain is controlable from boat.


    I.e. I can readjust the amplification of the


    receive amplifier in steps if need be.


    Amplifier is build within transducer array to keep noise down.


    4) Communication to and from fish is done via dual lightpipe for full duplex operation.


    I am planning for lightpipe because:


    Plus:


    -----


    a) No amplification needed when length of lightpipe is 300 meters or less (For coax that would be about 50 meters)


    b) Much higher data speed possible.


    c) No interference from other signals.


    d) Very thin cables.


    c) Very good for digital datatransfer.


    Minus:


    ------


    a) More expensive


    Well... the whole project is expensive.. so..


    ---------------------------------------------------------


    The questions I have are:


    1) Is it enough to send just 1 pulse or must it be several pulses ?


    2) How long before the transducer is settled again


    after the send pulse


    3) Instead of sending complete echo signal to computer


    I want to send 3 digital frames


    Frame 1: 8 bits for meters distance


    Frame 2: 7 bits for cm distance


    Frame 3: 4 bits signal strength


    Frame 1 & 2 will give me actual distance of target with a resolution of 1 cm.


    Frame 3 will give me the signalstrenght in 16 steps.


    These frames are only sent when an actual echo was received, to minimize data traffic on lightpipes


    Is this enough data to make a reasonable picture ?


    4) How important is it to have a echosounder in fish to measuring hight above bottom ?


    Or would a depth indication be sufficient.


    (difference is that echosounder will give actual height and depth gauge will give average height)


    I hope Sture can give the answers, but of course everyone's input will be very welcome.


    Regards, Jan

  • #2
    Re: Transducer driver/receiver

    Jan,


    The optics sound a little esoteric. Do we really need all that much bandwith? If we do, great, but as we have not yet figured out the bandwith requirements (or at least they have not been posted yet) this seems a little premature.


    One idea I had would be to use regular ethernet on the coax, and then add some matching network so the DC power could share the cable.


    Perhaps ethernet is overdoing it as well, because there would be more electronics on the fish, and getting the analog network right might be tricky too. Perhaps a simpler, slower homemade protocol might work.


    It's been a while since I looked at a PIC, and my manuals are at another site. How do they communicate?


    -Bob

    Comment


    • #3
      Re: Transducer driver/receiver

      Bob,


      I want to use optics because it is not sensitive to external noise.


      Since it is my plan to digitize the data as soon as possible I prefer the optics.


      PICS can be programmed to do all kinds of protocols because one can write the protocol needed.


      Found several programs on internet that could be used.


      I am looking into the possibility to buffer received data in a RAM and then send data from RAM in bursts topside.


      The general idea (at the moment) is to buffer enough data for 1 line of pixels (= 1 ping of sonar)


      I was thinking of either using twin pipe to send/receive data 1 pipe data & 1 pipe clocking


      This is easiest and extremely fast.


      Another possibility is to send data in bursts of 20 bits (= 1 echo)


      The receive oscillator will be started by first bit and will freerun for the next 19 bits.


      This short time the oscillator will stay in sync with send oscillator.


      Any Protocol will slow things down an I think it will not be needed.


      During Ping, data is send from topside to fish. (but that is a feature only my fish will have)


      But first I want to know exactly what characteristics of echo is important and what not.


      That will also decide the amount of data to be sent.


      Regards, Jan

      Comment


      • #4
        Re: Transducer driver/receiver

        Jan,


        I think that using a buffer is a very good idea. This way you spread out the transmit time over the cable so bandwith is reduced. Also, digitizing early seems like a good idea. Digitizing in itself will help reduce errors by making the system less sensitive to noise.


        My only concern is that the more electronics that are on the fish, the more upset I'll be if the cable snaps or the fish gets flooded!


        One idea I had was to put a whole industrial PC down there. This would make life simple, because all kinds of values could be tweaked from the surface, using a remote control program, but suppose things went wrong?! I quickly scrapped that idea!


        I guess what I am trying to say is that I feel the cheaper the fish electronics the better.


        Cheaper doesn't mean less funcional though. That is why I thought of the Ethernet idea. I figured that using a standard part would give us more functionality for less.


        I know virutally nothing about using fiber optic transmissin lines, but I suspect it may require expensive electronics in the fish, and/or an expensive cable.


        If I am wrong then please let me know! My only point is that I want to avoid putting expensive parts where they might get damaged.


        -Bob

        Comment


        • #5
          Re: Transducer driver/receiver

          Bob,


          The electronics in the fish is not very expensive.


          It extensive use of PIC's will reduce the amount of parts.


          Sending signal over lightpipe is nearly as easy as sending it through a copper wire.


          The main difference is that lightpipe is connected to xmit/receive diode via special connector.


          You cannot just put lightpipe in connector but need special tools for it. (Glassfiber end must be polished and very wel fitted)


          Normally when one orders a length of pipe one can also let them put the connectors on it.


          The major problem I have at the momement is to know what I need to send topside.


          I hope that time between sending and echo + amplitude is enough, but am not sure about it.


          Hope Sture will give answer to this.


          If that is all, I will got for standard RS232 setup.


          There are inexpensive chips using a stripped RS232 protocol which will do the trick.


          I believe that the whole communication electronic will be below $50


          This is for both sides, but without lightpipe and connectors


          At the moment the transducer electronics do not seem to be very expensive (a PIC + some odd bits)


          But this can change depending on what Sture says what we need from the echo.


          I hope I can start the process flow diagram for the electronics next week.


          Then I will start on transducer and start writing the PIC program.


          This will be difficult at first because eventhough I am a programmer I never programmed a PIC before.


          I do not suppose its difficult otherwise they will not sell these things in the quantity they do.


          Regards, Jan

          Comment


          • #6
            Re: Transducer driver/receiver

            Hi Jan,


            If you use a lightpipe are you also going to have power going down the cable? The transducer will want a fair amount of power or at least lots of volts (100 to 1000). You need a tank circuit or something to drive the volts as high as possible. The volts are what make the piezo ceramics expand. The more volts the stronger the wave. I like designing around PICs. It makes the design very flexible. I like every thing I hear so far, I think we’re making progress.


            Tod

            Comment


            • #7
              Re: Transducer driver/receiver

              Tod,


              I have a seperate powersupply in fish.


              I was not planning to have an electrical connection between boat ancd fish.


              All is done via fiber optics.


              I have no idea what voltage I am going to need for transducer, am still gathering information about them.


              If they need several hundred volts than that is no problem (assuming there is no big current involved)


              I will then charge a capacitor with high voltage and discharge it over transducer.


              Regards, Jan

              Comment

              Working...
              X