opamp to increase signal from LONG cable

Discussion in 'General Electronics Chat' started by mxabeles, Apr 11, 2010.

  1. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    hey there
    i have a distance sensor hooked up to 25' cable, 3 conducters - gnd, 5v+ and analog output. There is some added noise and a weaker signal than when I plug the sensor directly into the microcontroller (arduino). Can I use an opamp (single sided preferably, i just bought an LM324) to cancel noise/bump signal?
    Thanks, I know this INSANELY basic.
    Best,
    M
    PS does opamp circuit go between sensor and cable or between cable and microcontroller?
     
  2. Bychon

    Member

    Mar 12, 2010
    469
    41
    There are "rail to rail" op-amps that can work in a 5 volt envelope. It must be placed at the sensor end. After the cable has gone 25 feet, the damage is done. No sense amplifying the noisy signal at the wrong end of the cable.
     
  3. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    You should boost the sensor BEFORE the cable, else you will be amplifying the noise also.
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Show us the circuitry on your sensor end, particularly all bypass capacitors that you are using.

    What is the cable that you are using? Wire gauge? Is it (hopefully) coaxial cable?

    Once noise gets on an analog signal, you will have a dickens of a time trying to get rid of it.

    Much better if you can prevent it from getting in there from the start.
     
  5. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    bypass cap 47uf electrolytic btwn power ground. didnt seem to have much of an effect on the signal itself.
    OK so i will try with op amp. I found some schematics online that should work, but if anyone wants to post some schematics I'm not gonna cry about it ::))).
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The 47uF electrolytic is a start. Also use a 0.1uF metal poly or ceramic cap; at least one for each IC.

    An LM324's common mode range goes to ground, but not Vcc; limit is Vcc-1.5v. Out range is roughly 20mV to Vcc-1.5v.

    If you try running it from a single 5v supply, don't plan on seeing anything higher than 3.5v or lower than 20mV out from it.

    You'll also add a 3mV to 7mV offset when wired as a voltage follower/buffer.
     
    mxabeles likes this.
  7. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    to be honest, the signal is not any weaker with long cable, its just alot noisier. I tried hooking up an impedence buffer circuit with opamp (analog out from sensor to +, - tied to output) but nothing better in terms of signal stability. Is this the best type of opamp circuit to be using in this situation?
    Thanks,
    M
     
  8. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Since you are talking of using an old LM324 quad opamp that has very poor performance at audio frequencies then your signal must be low frequencies or DC. 25' of cable is nothing. 1000' of cable might cause your problems.
    Maybe the receiving end of the cable has a resistance that is much too low? What is the sensor and what is the receiver? What level?
     
  9. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    the sensor is an sharp IR distance sensor with bypass cap btwn power and ground. It outputs apx 0.2 - 2.7 Volts DC based on distance. It the analog output is being fed into a arduino microcontroller to control audio data in max/msp.
     
  10. Bychon

    Member

    Mar 12, 2010
    469
    41
    I've done this with a j-fet and it worked fabulously, except the DC offset of the j-fet was not a consideration...and it won't work in this application. Still, using a current buffer should work. I suggest that a terminating resistance be added at the Arduino end. That will lower the impedance of the line. I used 1500 ohms to terminate my application and the 324 will handle that load.

    If this not work, the noise is coming in at the sensor...before the buffer amp, not in the wire acting as an antenna. Whole 'nuther problem.
     
    mxabeles likes this.
  11. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Using coax or properly shielded cable will offer you a better signal to noise ratio IF the cable you are using is the problem. Actually it can only help.
     
    mxabeles likes this.
  12. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    thanks guys!
    To answer a couple of questions:
    The noise is not coming from sensor, ive test run this with very short jumper cables from sensor to microcontroller and it performs A-ok.
    Second, the cable could be an issue. What I have is a three conductor cable. The three conducters are each in plastic covering and there is a metal foil wrapped around all three.
    There IS a "drain" i think, a bare metal solid wire that is touching the foil. Should I be using that as ground? That way it is kind of like a coaxial cable, i think. Right now of course I am using the plastic insulated wires.
    I would just go ahead and try this myself but it would require de-soldering my op amp circuit so I thought I might ask ahead of time.
    Thanks M
     
  13. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    You do not want to use the wire you referred to as the drain as a signal carrying wire. Ground at best.
     
  14. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    bychon: can u please explain in layman terms how to hook up the terminating resistor?
    Thanks.......ie. what is the resistor between, 5v, gnd, sensor output, something in the opamp circuit?????? your talkin to an artist here :)))
     
  15. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    yes retched, I was thinking as ground! Do u think using the aforementioned wire would be better than an insulated wire??? tnx
     
  16. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Of course the drain wire should be connected to ground. The drain wire should be used instead of the ordinary wire that you now use as the ground wire and it should be connected to the sensor if it needs a ground connection. Then the foil becomes the shield.
     
    mxabeles likes this.
  17. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    I love the "of course" as I said, ur talking to an artist! lol. Thanks man!
     
  18. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Better an artist than a politician. Just to give a try, do you have 25 feet of standard stranded wire? I would do a test, even with a 25' extension cord to see if that clears it up. I really think the problem lies in the cabling.

    [ed]
    Audioguru, just saw your post.
    He is correct....as usual... :)
    [/ed]
     
    Last edited: Apr 11, 2010
  19. Bychon

    Member

    Mar 12, 2010
    469
    41
    The terminating resistor goes at the computer end of the cable between the signal wire and the 0 volt insulated wire. In that position, it requires the amplifier to send current (.001 amps) through the signal wire to the 0 volt wire. Stray noise leaking into the signal wire also has to drive that resistor, and since the noise is so much weaker than the amplifier signal, it gets squashed.
     
    mxabeles likes this.
  20. mxabeles

    Thread Starter Active Member

    Apr 25, 2009
    182
    0
    by 0 volt insulated wire do you mean ground? cause by the way I tried using the "drain" wire as ground and it didn't really alleviate the problem :((
     
Loading...