Erratic behaviour in OP07

Discussion in 'The Projects Forum' started by franno, Nov 20, 2013.

  1. franno

    Thread Starter New Member

    Nov 20, 2013
    3
    0
    Hi everybody, I'm building a circuit to amplify a thermocouple signal with an OP07. I've built about 20 PCBs and some of them have encountered a strange problem.
    When I repeatdely turn on and off the device, the amplifier output suddenly goes from 0v to 0,5v. It is as if the offset seting has changed. Once the output has gone to 0,5v once, every time I turn it on afterwards, it stays at 0,5v. If I wait about 10 seconds with the device off an turn it on again, the ouput goes back to 0v.The problem is that I can´t solve this by constantly adjusting the offset, as the PCBs are enclosed in a case.
    Testing, I've noticed that if I heat a little bit the voltage regulators, the operational output rises, but when they cool down the output doesn´t go below 0,5v. The problem must definitley be in the regulation stage.
    The PSU is composed of a 7812, a 7805 connected to the 12v output and a 7905. The problem seems to appear only if I heat the positive regulators. In normal work conditions, these don´t go above 115°F. I've already tried with heatsinks, but they don't seem to work. Should I change the regulators? Maybe a 2940 and a 2940-12 could solve this?
    I would really apreciate any help you guys could give me.

    PS: I attached an image of the amplifier part of the circuit. It is used as a differential amplifier.
     
  2. #12

    Expert

    Nov 30, 2010
    16,346
    6,833
    and yet, you do not present the regulation stage to us.

    Guessing from no information about the circuit, did you include the capacitors that are necessary to keep the regulators from oscillating?
     
  3. franno

    Thread Starter New Member

    Nov 20, 2013
    3
    0
    Sorry for that. Both electrolytic capacitors are 470uF, and either the input of the 7812 and the input of the 7905 are connected to the diode bridge, which supplies from a center tap transformer.
     
  4. #12

    Expert

    Nov 30, 2010
    16,346
    6,833
    All of that looks good. You seem to have used the right sizes, so I assume you read the datasheet. Another important aspect is the TYPE of capacitor. An aluminum electrolytic will not substitute for a ceramic capacitor because this is about high frequency oscillations, and aluminum caps aren't very good at high frequencies. Be sure you read the footnotes and used the right type of capacitors.

    I had a thought while I was away. A problem I had with my own design. The idea that the problem starts when you switch the supplies on and off. If one regulator starts before the other it will drive it's opposing polarity regulator backwards and it won't start. One cure for this is to add a Shottky diode on the output of each regulator, oriented so it seems to be doing nothing at all, and that will stop the positive regulator from driving the negative regulator backwards (and vise-versa). I didn't have any Shottky diodes that day so I added 22 uf of capacitance to each regulator output to slow them down a bit. That fixed the problem at MY house.

    Of course, switching quickly on and off will defeat this because the storage capacitors will not have had time to discharge. Punchline for an old, stupid joke: If it hurts to do that, don't do that. If switching the power quickly on and off causes glitches, don't switch the power on and off quickly.

    That's all I have right now.
     
  5. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    Are all those capacitors physically close to the regulator pins? If not that could cause oscillations, the onset of which could be heat sensitive.

    Do you have an oscilloscope to look for oscillations?
     
  6. Dodgydave

    AAC Fanatic!

    Jun 22, 2012
    5,003
    745
    Why do you need the +12v regulator, just use the +5V on its own?
     
  7. RamaD

    Active Member

    Dec 4, 2009
    254
    33
    Most of the parameters for the voltage regulator are specified for a minimum load of 5mA. While this is normally not a problem in +ve supplies in the circuit, the -ve supply may not have enough load on it. Can you try a min. load of 5mA with a resistor across the power supply output? I have come across the output oscillating in -5V with meagre load!
     
  8. franno

    Thread Starter New Member

    Nov 20, 2013
    3
    0
    Crutschow, the capacitors are really close to the regulators, no more than an inch. Unfortunately, I don´t have an oscilloscope.

    Dodgydave, the +12v regulator supplies another part of the circuit.

    RamaD, I'm not sure if I got you well, but you´re suggesting to put a resistor between the -5v output and GND in order to raise a little bit the output current (above 5mA) of the regulator?
     
  9. #12

    Expert

    Nov 30, 2010
    16,346
    6,833
    Yes. Regulator chips have a minimum load.
     
  10. RamaD

    Active Member

    Dec 4, 2009
    254
    33
    Yes, a 1K across the -5V and Gnd. I vaguely remember reading a blog of someone whose ADC kept blowing up, where the output voltage was higher, pretty close to the input voltage due to lack of minimum current. The ADC, which was using this higher voltage kept blowing up! He had compared a few regulators of different makes and had published the results.
     
  11. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    No more than an inch is not "as close as possible". I´d say they should be no further than 1/4" inch from the regulator.

    Also, what purpose does C13 serve? It seems to me that it only allows RF noise to straight enter the input terminal of the opamp.
     
  12. #12

    Expert

    Nov 30, 2010
    16,346
    6,833
    Good question.
     
Loading...