Post Go back to editing

AD8400 Digital Potentiometer Voltage Range

I'm using an AD8400 potentiometer in the feedback path of an op-amp powered with +/-12V supplies.  The DC bias is 2.5V.  The circuit is oscillating.  I think the problem is a transient negative voltage on the op-amp output forward biases the protection diodes and causes excessive current flow into the VCC pin.  This pulls down my resistor divider 5V supply.  If the pot kept working during the transient, the negative voltage from the output would make its way to the inverting node of the op-amp and pull the output back to 2.5V.

What is the best approach?

  1. Use a 5V supply to the AD8400 that's capable of handling more current.
  2. Use a pot that supports higher voltage and/or reduce the rails on the op-amp so they're equal to the rails of the pot?

Is there ever a way I can use this part in a circuit that has transients beyond the pot's supplies?

  • Hi Charles,

    Could you try to connect a 10nF capacitor between inverting input and op-amp output in order to cancel oscillation at high frequency?

    Let me know if this fix the problem.

    Best Regards,

    Miguel

  • Miguel:

    It turns out there were two things happening in this circuit;

    1. Noise from this circuit was coupling via my rework wires into the previous stage, whose gain is very high. --> Eliminating this stopped the oscillation.

    2. Depending on supply sequence, sometimes this the op-amp's output would start off negative.  This would cause high current to flow through the AD8400's protection diodes.  The current tripped my bench supplies current limit, and between that and pot not functioning properly, the feedback didn't find it's way back to the interverting node.  Therefore it would lock up in this failure mode.  I corrected this by adding 1k of resistance to op-amp output and adding a Schottky diode to keep the output from going negative.  Now it comes up regardless of sequencing.

    Considering it went into the high-current-through-pot-protection-diode mode several times, I'm glad I didn't smoke the pot -- a phrase you don't hear that often.

    Is it not recommended to use this part in an op-amp circuit whose rails go well beyond the pot's limits?  Is my resistor and diode solution sound or just a band-aid to avoid the real solution of lowering the op-amp's rails to be the same as the pot's?

    Thank you for your help.

    Regards,

    Charles

  • Hi Charles,

    I guess if you are using this digipot is mainly bc the BW...

    I would recommend the AD5290 but in this case is 10k so lower frequency.

    If you will really want to use this digipot, I would recommend to completely protect the digipot terminal by adding to diodes.

    Best Regards,

    Miguel

  • Hi Charles,

    The critical issue that I think you're getting to is the need for the ends of the pot or rheostat needs to be at worse within the Digital pots power supply levels, as defined in the absolute max ratings, and the spec tables. Depending on the circuit configuration, and voltages applied, that doesn't "necessarily" mean that just because the amp may have wider supply voltages than the digital pot, that you shouldn't use it. Ultimately it's an application issue and assessing what voltages are possible to be seen at the digital pot verses what the digital pot will tolerate.

  • This question has been closed by the EZ team and is assumed answered.