• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 49 subscribers
  • Views 301 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

OPA192: opa192 on a high voltage application

S.G.
Prodigy 40 points
Part Number: OPA192
Other Parts Discussed in Thread: ADS8861, ISO7841,

Hello TI Team,

We have created a 4-Ch isolated design for differential measurement applications based on the ADS8861 reference design. The input to the differential amplifier is handled by a unity buffer configuration of a OPA192 with the addition of a PTC and Shottky protection. Powering is managed with an isolated dc/dc and communication through an ISO7841 digital isolator.

We want to measure the individual voltage of a capacitor bank comprising of four capacitors in a square configuration. The bank maximum voltage is 1kV, so each of the capacitors is charged at 500V.  Each capacitor has a resistor network attached to the terminals to balance the charging and also act as a divider for the the voltage measurement. The idea was to use the already build 4 Channel board for this measurement. The divider is configured so the input of each channel is maximum 10V.

What is observed is that upon connection of all channels, the voltage on the dividers shifts and looks like also the bank becomes unbalanced. Capacitors (1 and 2) on the positive branch increase voltage a bit while Capacitors (3 and 4) on the negative decrease. This does not happen when only 2 channels are connected at the either 1,2 or 3,4 Capacitors. Simulation shows that the most positive input and the most negative input have higher leakage current through the shottky diodes. I also suspect that since all channels clamp inputs at the level of the OpAmp power supply this changes the current flowing on the balancing resistors.

The idea was to remove the shottky diodes so the inputs go directly to opa192. The power supply of all OPA192 is common and it is  +-12V for all channels. The common mode voltage is not exceeded in each of the channel but the voltage difference between channels will be 500V. Is this safe for the amplifier and the circuit?

Kind regards

  • 0
    Guru 140200 points

    Hi,

    hard to answer without a schematic...

    Since the imbalance is very probably caused by the capacitor leakage currents, you could check the performance with removed caps. Is there still an imbalance then?

    Have you tried to replace the Schottky diodes by diodes showing smaller leakage currents? Removing the protection diodes at all may not be a good idea. Think of connection problems at the cap or resistor terminals or a catastrophic cap failure.

    Another remedy could be to decrease the balancing resistors and by this to minimize the impact of leakage currents on the balancing.

    Kai

  • 0 in reply to kai klaas69
    Prodigy 40 points

    Hi Kai,

    Below you can see a Simulation screenshot form TINA :

    This is the roughly the schematic. The simulation is running on the charged state of the bank without the analog front end connected. 

  • 0 in reply to S.G.
    Guru 140200 points

    Hi,

    I think the balancing resistances are too big.

    What sort of caps are we talking about? Aluminium electrolytics?

    Kai

  • 0 in reply to kai klaas69
    Prodigy 40 points

    Hi Kai,

    The capacitors are metallized polypropylene film capacitors of 40mF/1kV. The bank is always charged at 1kV and connected to a charger.

    Kind regards

  • 0 in reply to S.G.
    TI__Mastermind 19697 points

    S.G.

    My concern with eliminating the Schottky diodes would be related to the transient start-up of the circuit.  When power is first applied can the amplifier common mode go beyond the supplies?  As Kai suggested perhaps you could transition to a lower leakage diode.  For protection circuits, Schottky diodes are often used because of their low voltage drop (thus they clamp near the supply compared to other options).  However, in your case if the diode clamp is not as close to the supply rail that is ok as the abs max on opa192 is +/-20V.

    Best regards,

    Art