Post Go back to editing

High voltage fully differential amplifier ( supply voltage >= +/- 30V )

Hello,

Do analog devices have fully differential amplifier which has supply voltage >+/-30V ? 

I have differential signal having range +/-24V. 

Thanks,

Rahul. 

Parents
  • Hi Rahul,

    I don't see any ADI diff amps which can handle +/-30V supplies. There may be solutions for you, but need more details.

    Do you plan to provide "gain" to this large input, or "attenuation"?

    What would be the diff amp you're looking for drive?

    Regards,

    Hooman

  • Hello Hooman, 

    My differential signal range would be +/-10V to +/-24V. And output swing would also be between +/-10V to +/-24V. Meaning I need to gain in the range of 0.4 to 2.4. 

    Fully differential amplifier need to have sufficient supply range for me to have output swing till +/-24V. 

    Do you suggest any solution? 

    Thanks. 

  • Hi Rahul,

    I cannot find a fully differential amplifier (FDA) that can handle the large supply voltages you need. The highest supply voltage FDA is AD8476 unity gain (+/-9V) or LT6411 (+/-6.3V).

    A more complicated way to get there would be to use bootstrapping on your device supply voltages. This would dynamically vary the V+, and V- pin voltages to accommodate the output signal swing without ever exceeding the total supply (V+ - V- or Vs) voltage. This is highlighted here:

    https://how-to.fandom.com/wiki/How_to_boost_the_output_voltage_swing_of_an_operational_amplifier

    However, to use +/-30V supplies with the bootstrapped configuration, you'd need to use a device that can handle about 30V and FDA's listed above won't get you there. Also, this technique only works with relatively low frequency signals or the dynamic supply swinging won't be able to keep up with the signal.

    You could use a single ended output instrumentation amplifier like LT1920 (+/-18V) and then use a technique like what's shown below to create the differential output pair:

    Of course the LT1920 shown above has to have the supply bootstrapping applied to it to handle the +/-30V. A good candidate for the OPA shown would be ADA4700-1 HV op amp (+/-50 supplies).

    So, I don't really have a simple answer for your problem but you may be able to apply the ideas above to get there.

    Regards,

    Hooman

  • Hello Homan,

    Thanks a lot for the solution. I'll certainly try out the second circuit which looks easy with minimal components. 

    Could you please take a look at the attached datasheet of LTC6244, page 21? 

    It explains FDA from single ended opamps. I'm not sure how it works. Could you suggest any note which explains working of the circuit?

    PDF

  • Hi Rahul,

    Regarding the LTC6244 Figure 9a schematic (Fully differential buffer / amplifier) operation:

    I don't have any app note that describes how the circuit operates. However, I've looked at it a little and I can offer this explanation as to the circuit's low frequency response (CIN and CF open):

    Each output (e.g. VOUT+) is the sum of the inverted output (e.g. -VOUT-) and 2x of what appears at the RIN / RG junction. The signal at this junction in-turn is influenced by the corresponding input (e.g. VIN+ on the top side) and an attenuated version of the opposite side output (e.g. VOUT-/Gain). By virtue of the symmetry of the circuit, the differential output (VOUT+ - VOUT-) becomes a gained up version of the difference input (VIN+ - VIN-) only. However, I only know that this works because I used the attached derivation to prove it to myself. I don't have a good intuitive reasoning to explain its simple and effective results.

    Here is a more rigorous analysis for your reference:

    LTC6244 diff amp derivation 7_8_19.zip

    Regards,

    Hooman

Reply
  • Hi Rahul,

    Regarding the LTC6244 Figure 9a schematic (Fully differential buffer / amplifier) operation:

    I don't have any app note that describes how the circuit operates. However, I've looked at it a little and I can offer this explanation as to the circuit's low frequency response (CIN and CF open):

    Each output (e.g. VOUT+) is the sum of the inverted output (e.g. -VOUT-) and 2x of what appears at the RIN / RG junction. The signal at this junction in-turn is influenced by the corresponding input (e.g. VIN+ on the top side) and an attenuated version of the opposite side output (e.g. VOUT-/Gain). By virtue of the symmetry of the circuit, the differential output (VOUT+ - VOUT-) becomes a gained up version of the difference input (VIN+ - VIN-) only. However, I only know that this works because I used the attached derivation to prove it to myself. I don't have a good intuitive reasoning to explain its simple and effective results.

    Here is a more rigorous analysis for your reference:

    LTC6244 diff amp derivation 7_8_19.zip

    Regards,

    Hooman

Children
No Data