Zener Series Resistor Value?

Discussion in 'General Electronics Chat' started by tracecom, Jan 21, 2012.

  1. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    In the circuit below, if the Zener is a 1N4733 which has a maximum current rating of 178mA, I think I understand that the absolute minimum value of R1 must be 39Ω, but how does one calculate the maximum value of R1 if the Zener is simply supplying a reference voltage for an op-amp?

    [​IMG]
     
  2. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    Op-amps inputs draw current in uA so you can ignore it.

    Just design it for 10mA (or whatever the zener datasheet specifies the zener voltage) current through the resistor just to keep the zener in good regulation.
     
  3. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    Thanks. I don't find that parameter on the datasheet. Some references on the WWW specify as little as 1mA while others say 5mA. The datasheets give a "Test Current", which I suppose is a safe value to use, but in the case of the 1N4733A, that's 49mA. Surely there is a better answer, but I haven't found it.
     
  4. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    If you need a good stable reference you may want to consider a voltage reference device rather than a simple zener diode. Such as LM 136
     
    Last edited: Jan 21, 2012
  5. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    Thanks. I'm sure you are right, but I am working with an existing circuit trying to minimize its current requirements without compromising performance.

    Additonal Googling has turned up a recommendation to use 10% of the maximum current and another to use the knee current (which I also don't find on the datasheet.) Still another says that the safe values to use are those between the test current and the maximum.
     
  6. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    The test current IZT is 49mA, and this current should give a zener voltage within the specified tolerance limits. A much lower "knee" current of 1mA is given. At this current the zener is beginning to function, with a specified slope resistance. The zener voltage will probably be out of tolerance though.
    http://www.futurlec.com/Diodes/1N4733.shtml

    We
    see then that 49mA is the optimum current for best accuracy with this 1N4733A device. This is a fairly big diode though, rated at 1W. If the requirement is only for a reference to an op-amp, which requires little current, it would be usual to use a smaller diode, say one rated at 400mW or 250mW. The test current for a lower rated diode will be lower. For example, these BZX84 diodes are rated at 250mW, and the test current for a 5.1V device is 5mA.

    http://www.nxp.com/documents/data_sheet/BZX84_SERIES.pdf

    Edit: The point made by others about band-gap references is a good one. These can be very frugal, down to tens of micro-amps for some devices.

     
    tracecom likes this.
  7. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    R1 in the circuit as it currently is built is 1k, which calculates to about 7mA of current, but the voltage is sagging to 5.05V. When I make R1 150Ω (about right for 49mA) the voltage is 5.27V. And that means that a 1/4 watt resistor is no longer big enough since it is now dropping 6.71 volts at 49mA. :(
     
  8. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    So try a few hundred ohms, 330Ω say - That should give about 20mA, and should dissipate about 140mW.

    How critical is this thing anyhow?
     
  9. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    If you supply to the op-amp is stable then you can use a resistive voltage divider with a 100nF filter capacitor across the input of the op-amp and 0V.
     
  10. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    The Zener is supplying a reference voltage to an op-amp in a voltage follower configuration, which, in turn, is connected to a thermistor in a voltage divider. Then, that thermistor voltage divider output is connected to another op-amp in a comparator configuration with an adjustable voltage divider on the non-inverting input. I have the comparator output controlling a MOSFET which is switching a heat source. So the exact voltage of the Zener is not critical, but the stability of the reference voltage directly inpacts the repeatablity of the comparator output.

    As usual, I am just experimenting to try to learn about components - how they perform and what their limits are. In this case, I would like to minimize the current required by the entire circuit while maximizing the accuracy of the comparator switch point.

    Thanks for your input.
     
  11. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    I will try that. The circuit already includes a 100 nF cap from the input of the op-amp to ground. Thanks.
     
  12. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    It might help to know what this is for. As an example, taking a fraction of the supply as a reference is OK for some things, but power-on or power-fail detectors are not among them.
     
  13. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    It's a thermostat; see post 10 for more info.
     
  14. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    If the thermistor were used in a potential divider, and the reference voltage were taken from a similar divider, and both were driven from the same supply, then the supply voltage stability may become less of a concern.

    What you would then have would essentially be a bridge circuit: providing that the thermistor self-heating was small, the bridge balance condition would be essentially independent of the supply voltage.

    If the only function of the zener is as a reference to stabilise the supply to such a bridge, and power consumption is important, the use of a large 1W device would not appear to be the optimum.

    Finally, since verbal descriptions of circuitry are difficult and subject to error, it might be useful if you could post your circuit.
     
Loading...