Power Amp for Heated Sensor

Discussion in 'General Electronics Chat' started by DangerousBill, Sep 12, 2012.

  1. DangerousBill

    Thread Starter Active Member

    Jul 21, 2010
    30
    1
    Greetings from a long-term lurker. I want to humbly seek advice on a problem that's defeated my best attempts at solution.

    History: I previously designed a controlled-resistance power supply for an associate. It had to deliver 6 V at 0.6 to 0.8 A. I split a 24 VDC supply with an LM675T to power the entire circuit. The device under test was a heated sensor with an attached platinum RTD. The bridge error adjusted the output to a second LM675T to keep the resistance constant. It has worked successfully for a year.

    Now, he wants a circuit that will deliver 15-18 V at 0.6 to 1.0 A. The heater is platinum and also serves as the RTD. My first thought was 'scale everything up' and run it all from a split 48 V supply, but some important components, like instrument amps, don't seem to be available that will run on 48 V. Not to mention the heat sinking issues. So I'm shooting for a single-ended solution, run from a 24V, 2.5A supply. (See the attached figure.)

    Experiments convinced me that the LM675 couldn't operate very close to the negative or positive rails without breaking into song (abt 80 kHz, even with compensation up the wazoo), even at gains of 20 or more. So it seems I have to default to another way to control the current, either a Darlington or a MOSFET.

    My question: which would be the best solution in this case? I have little experience with MOSFETs in medium power applications. All suggestions welcome, please.

    Thank you in advance,
    Dangerous Bill

    Restrictions:
    - it needs to be DC; no PWM allowed, therefore, operation in the linear range
    - one side of the sensor has to be grounded
     
  2. crutschow

    Expert

    Mar 14, 2008
    13,003
    3,232
    You can use a standard single supply, rail-rail type op amp to drive an N-MOSFET power transistor or an NPN Darlington power transistor with a follower output (from transistor source/emitter). The feedback for the op amp is taken from the transistor follower output.

    The op amp needs to have a max voltage rating of at least 30V (±15V).

    If you use a MOSFET, a logic-level type with current and power rating to handle your load would work best.

    Make sure the transistor is mounted on an adequate heat sink.
     
  3. DangerousBill

    Thread Starter Active Member

    Jul 21, 2010
    30
    1
    Thanks. That pretty much answers my question: Of Darlington or MOSFET, either device would do the job adequately.
     
  4. crutschow

    Expert

    Mar 14, 2008
    13,003
    3,232
    Yes. You just need the Darlington to have sufficient gain at the maximum output current so that the op amp can provide the required base current.
     
  5. Dodgydave

    Distinguished Member

    Jun 22, 2012
    4,980
    744
  6. DangerousBill

    Thread Starter Active Member

    Jul 21, 2010
    30
    1
    That's the sort of circuit I was looking for.

    For what it's worth, I just bought the last available copy of that book in the observable universe.
     
  7. DangerousBill

    Thread Starter Active Member

    Jul 21, 2010
    30
    1
    This got me over the hump. I've got the circuit blocked and I'm doing the detailed schematic for the prototype now. Thanks again.

    As for your sig, it is now the case that 'Invention is the Mother of Necessity'. Did we know we needed the cell phone or the Internet before they were invented?

    Tx, Dangerous Bill
     
Loading...