Connect with us

dropping a variable voltage by 1v

Discussion in 'Electronic Basics' started by Ray Manning, Nov 25, 2005.

  1. Ray Manning

    Ray Manning Guest

    I'd like to offset a variable voltage by 1 volt. What I have is an LM34
    temperature sensor that outputs a linear voltage based on temperature
    (+10mV/F). I'd like to change that to (+10mV/F - 1V). What's the easiest way
    to drop the LM34's output by 1V?

    I tried a potentiometer but my understanding is that a pot is a divider so
    it doesn't give me the result I'm looking for. I also tried a diode that was
    supposed to have a 1V drop but that didn't seem to work either (not sure
    why). Is there a simple solution?

    Thanks,
    Ray
     
  2. An opamp with 4 resistors configured as a subtracter (difference
    amplifier) outputs the difference of two voltages plus a third
    voltage. A 2 opamp version has two of the inputs with high impedance,
    so you don't have to worry about loading high impedance inputs.

    http://www.national.com/an/AN/AN-20.pdf
    This opamp might be a good choice:
    http://www.national.com/an/AN/AN-116.pdf
     
  3. Chris

    Chris Guest

    Hi, Ray. Get an op amp and four equal value resistors (match values
    with your ohmmeter -- the closer they are, the better), and make a
    difference amplifier like this (view in fixed font or M$ Notepad):
    ___
    .-|___|-.
    | 22K |
    | |
    ___ | |
    Vin o--|___|---o VCC |
    22K | |\| |
    '-|-\ |Vout =
    ___ | >--o----o
    1VDCo--|___|-o---|+/ |
    22K | |/| |
    .-. GND .-.
    22K| | 22K| |(not matched)
    | | | |
    '-' '-'
    | |
    | |
    === ===
    GND GND
    (created by AACircuit v1.28.5 beta 02/06/05 www.tech-chat.de)

    This circuit will do the job for you. If you don't need anything below
    about +50mV DC output from the op amp, you can use a single supply and
    half of an LM358, or 1/4 of an LM324 (available at Radio Shack and many
    other places). If you need a negative voltage output, you'll need a
    dual supply (say, +/-5V) and you can use an LM741 or any other standard
    op amp.

    Good luck
    Chris
     
  4. Bob Masta

    Bob Masta Guest

    Sure, you just need an op-amp difference circuit. One input
    is your signal, the other is 1V, and the output is the difference.
    You will need a known voltage source, but it doesn't have to be
    1.00 V or even close, since you can scale it with the input resistors
    or a separate divider. You can even use a negative reference and
    sum it with the LM34 to get a difference.

    Best regards,



    Bob Masta
    dqatechATdaqartaDOTcom

    D A Q A R T A
    Data AcQuisition And Real-Time Analysis
    www.daqarta.com
    Home of DaqGen, the FREEWARE signal generator
     
  5. John Fields

    John Fields Guest

     
  6. Chris

    Chris Guest

    Oops. Wrong labels -- try this:

    ___
    .-|___|-.
    | 22K |
    | |
    ___ | |
    1VDCo--|___|---o VCC |
    22K | |\| |
    '-|-\ |Vout =
    ___ | >--o----o
    Vin o--|___|-o---|+/ |
    22K | |/| |
    .-. GND .-.
    22K| | 22K| |(not matched)
    | | | |
    '-' '-'
    | |
    | |
    === ===
    GND GND
    (created by AACircuit v1.28.5 beta 02/06/05 www.tech-chat.de)

    Inputs backwards -- sorry.

    Chris
     
  7. Ray Manning

    Ray Manning Guest

    John,

    Are you suggesting that if I applied -1v to the GND terminal of the LM34,
    the output would be offset by -1 volt? If I understand correctly, this would
    work because the output would still be referenced against the original
    ground. If thats the case, can I produce the -1V using the input +Vs (12v in
    this case) and the orginal ground? Wouldn't that solution (offseting the
    input) be the same as offseting the output or am I missing something?

    Thanks,
    Ray
     
  8. John Fields

    John Fields Guest

    ---
    If by that you mean that the thing connected to the LM34 has as it's
    ground, 0V, then yes.
    ---
    ---
    You're missing that the LM34 doesn't care what the supply voltage is
    (within limits) since it only uses that to run the chip. Its
    output, however, is referenced to the "holy" ground pin. That means
    that at 0°F the output will be at whatever voltage the ground
    terminal is at and will go more positive by 10 millivolts for every
    degree F change in ambient temperature. That is, if the ground
    terminal is at 0V, the output will be at 0V and will go to 1V for
    the change from 0°F to 100°F, but if the ground terminal is at -1V
    when the temperature is 0° the output will be at -1V and will rise
    to 0V as the temperature goes to 100°F. _If_ you measure that
    voltage with a voltmeter referenced to the junction of the two
    supplies.

    Look at it like this: (in Courier)


    +------------------+
    | |+
    | [12V]
    | |
    [LM34]--[VOLTMETER]--+
    | |+
    | [1V]
    | |
    +------------------+
     
  9. Ray Manning

    Ray Manning Guest

    I understand what you are saying but still don't see how to accomplish it.
    The curcuit that is driving the LM34 hands me the ground reference (0V) and
    a +12V source. How do I produce the negative voltage to connect to the LM34
    to change its reference point?

    It would seem to me that I would have to use the same difference amplifier
    others have suggested for dropping the output by 1V to create the negative
    1V for the "holy" ground pin. Are you saying there is an easier way?

    Thanks,
    Ray
     
  10. John Fields

    John Fields Guest

     
  11. Ray Manning

    Ray Manning Guest

    The LM34 is driven by a circuit that converts the LM34's voltage to a
    frequency that is read by a remote computer. The computer supplies the
    circuit with a ground and +12V and receives a frequency on a third wire. The
    circuit in turn supplies the LM34 with the same ground and 12v supply and
    monitors the output voltage of the LM34.

    - Ray
     
  12. Jasen Betts

    Jasen Betts Guest

    drop it's the supply and "ground" by one volt.

    or something with an op-amp


    in ----[100k]--.
    | +|\
    0V-----[100k]--+--| \
    | >----+--- out
    1V-----[100k]--+--| / |
    | -|/ |
    | |
    `--[100k]--'


    a silicon diode ansd two germanium diodes in series will
    give approx 1V drop but it's terperature dependant...
     
  13. John Fields

    John Fields Guest

    ---
    OK. So what you have is a voltage-to-frequency converter and an
    LM34 wired like this:


    CPU+12>---------+-------------------+
    | |
    +-----+-----+ +-----+-----+
    | +12 | | +12 |
    CPU fIN<--|fOUT Vin|<------|Vout TEMP| <~~~
    | GND | | GND |
    +-----+-----+ +-----+-----+
    | VFC | LM34
    CPU GND>--------+-------------------+


    Where the LM34 outputs 10mV/F° starting with 0V at 0F, and what you
    want is to offset the output of the LM34 to start at -1V for 0F and
    rise at a rate of 10mV/F°?

    If that's the situation, then you can either lower the ground of the
    LM34 to -1V or raise the VFC's ground to +1V. Raising the VFC's
    ground to 1V would be easier, but since its output would also be
    referenced to the 1V "ground" it would probably be offset as well,
    so that may not be a good idea. Can you tell us what you're using
    for a VFC or post its spec's?

    If you have -5V or -12V available at your CPU and you've got an
    extra wire in the cable, you could generate the -1V like this:

    CPU+12>---------+-------------------------+
    | |
    +-----+-----+ +-----+-----+
    | +12 | | +12 |
    CPU fIN<--|fOUT Vin|<------------|Vout TEMP| <~~~
    | GND | | GND |
    +-----+-----+ +-----+-----+
    | VFC | LM34
    | |
    CPU GND>--+-----+------+-------+-[0.1µF]--+
    | | | |
    | [0.1µF] [R1] |
    | | | |
    [0.33µF] +-------+ |
    | | | |
    | | [R2] |
    | +---+---+ | |
    CPU-12>---+--------|IN OUT|---+----------+
    +-------+
    79L05


    If you don't, then you could use something like a 7555 to build a
    charge pump followed up with a regulator, like this:

    CPU+12>------------+------------+------------+--+12
    | | |
    +-----+-----+ | +-----+-----+
    | +12 | | | +12 |
    CPU fIN<-----|fOUT Vin|<-----|------|Vout TEMP| <~~~
    | GND | | | GND |
    +-----+-----+ | +-----+-----+
    | VFC | | LM34
    CPU GND------------+ | |
    | | |
    GND | |
    | |
    +---[C1]---+------|------+ |
    |K | | | |
    [1N4001] [Rt] +---+---+ | |
    | | | Vcc | | |
    +----+ +--|TH OUT|--+ |
    |- | | |___ _| |
    [C2] | +-O|DIS R|O--+12 |
    | | | | GND | |
    GND | [Ct] +---+---+ |
    | | |7555 |
    | GND GND |
    | |
    | |
    | 79L05 |
    | +-------+ |
    +-----|IN OUT|---+-------+
    | +---+---+ | |
    | | [R1] |
    | | | |
    [0.33µF] +-------+ [0.1µF]
    | | | |
    | [0.1µF] [R2] |
    | | | |
    +---------+-------+-------+
    |
    GND

    If you're interested in either of these approaches I'll be happy to
    post the component values.

    Hmmm... After looking at the spec's for the 79L05 I think the lowest
    output available from it is 1.2V, and I couldn't find any 1V shunt
    regulators or adjustable negative regulators with outputs which went
    below 1.2V, (In all fairness, I didn't look that hard ;)) so in
    order to keep from having to use a Zener and a voltage divider
    (which might be OK depending on how tight your -1V spec is) it looks
    like the charge pump and an opamp with a reference and its output
    forced to -1V with proper biasing will work. Depending on the
    current requirement for the LM34's output it might be possible to
    use a chip with an opamp, a comparator, and a reference in it to
    build the whole thing.

    How much more can you tell us about your application?

    In particular, what are you using for your voltage-to-frequency
    converter?
     
  14. Jamie

    Jamie Guest

    Hmm, you could use a dual osc, one that generates your freq to the
    computer and the other to generate a small low power - source for the
    Vee (common of the LM34).
    using a couple of caps and 2 diodes and a divider for the common to
    have the common of the LM34 connect to this -1 output.
    since there is very little current involved here i think you could get
    away with it.
    i don't know if your using a mpu or what but, either an extra osc or
    a clocked output can generate the constant you need for the - voltage
    source.
     
  15. Jasen Betts

    Jasen Betts Guest

    it may be possible to modify the VCO (voltage to frequency circuit)
    by raising its ground by one volt

    OTOH can you account for the 1V offset by modifying the software to read the
    frequencies differently?
     
  16. Jamie

    Jamie Guest

    not to criticize because i know your a smart guy and all but shouldn't
    the Diode be going to ground on the K side and A of the diode to the
    C1 and maybe another Diode in series to C2 to isolate the 0.7 average
    ripple ? or have i over looked something?
     
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

-