• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 41 subscribers
  • Views 219 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.

TIDA-00366: shall we use OPA2350AU after AMC1301 for Differential-to-Single Ended Signal Conversion?

Dave Hunson
Expert 1720 points
Part Number: TIDA-00366
Other Parts Discussed in Thread: OPA320, , AMC1301, TLC372, LM393, OPA2350

HI everyone!

I would be happy if you take a time and answer these questions :

Q1: In "TIDA-00366 => TIDUBX1B.PDF => page 17 => Figure 13. Differential-to-Single Ended Signal Conversion" instead of OPA320 ( U5 ) shall we use OPA2350AU?

Q2: the NTC thermistor in IGBT module is electrically separated from all high voltage pins and it is somehow Isolated , so why we need to use AMC1301 or any other kind of Isolation?

Q3:in TIDA-00366 => TIDUBX1B.PDF => page 20 =>4.7.1 Selection of Comparators says:
"
A typical industrial drive needs to have overload protection to operate within 300 to 500 ns. The TLC372
helps reach this goal. The TLC372 is very low cost and a dual-channel comparator. It operates from a
supply range from 3 to 16 V and has a typical response time of 200 ns."

as far as I know most in IGBTs Short circuit capability is 10us , so why we need fast reaction op amp for overload protection around 300 to 500 ns, why we can not use LM393 with response of 1.5us?

thanks for your attention!

regards.

dave.

  • +1
    TI__Expert 3450 points

    Hi Dave,

    thank you for your interest in our reference design.

    Q1: Please allow us a few days to look into the different parameters between OPA320 and OPA2350.

    Q2: The isolation is added for the NTC to handle e.g. a fault on the IGBT module generally, not necessarily related to the specific IGBT used in that design. In case of a fault, e.g. due to thermal or electrical overstress the NTC might be connected to the high-voltage. In that case the isolated amplifier prevents the high-voltage from the low-voltage side. We leave it to our customers to decide whether they need isolation or not. 

    Q3: We see requests of the total time to protect the IGBT from around 4us to 10us. The comparator is only a portion of the signal chain and additionally delay is added for example due to gate driver propagation delay, the MCU which can turn-off the PWM in case of an OC event, and additional glitch filters after the comparator to prevent short transients turn-off the IGBT. The design wanted to outline what's possible using the TLC372 comparators. No need to use that part, when the LM393 (1.5us) meets your overall requirements.  

    Regards,
    Martin Staebler

  • 0 in reply to Martin Staebler
    Expert 1720 points

    thanks dear 

    your answer was perfect!

    I am waiting for answering my first question because I really need to see if I can use OP2350 instead of OPA320.

    I forgot to mention about one more question.

    Q4:In "TIDA-00366 => TIDUBX1B.PDF => page 20 =>Figure 17. Overload Protection Circuit , we have a series resistor [R20, R23, R21 ] for reference voltage , we have the same for  other  channels (V-Phase and W-Phase) as well. which in total 9 resistor .  shall we use just 3  resistor [R20, R23, R21 ]  and connect the reference voltage to two other channels to reduce  series resistor?

    in order to know what I am talking about , In HVkit DMC we have one reference voltage that entering to 3 comparator (U,V,W) for trip fault

    thanks for your time best wishes.
    dave.

  • +1
    TI__Intellectual 970 points

    Hi Dave,

    For the first question, I think OPA2350UA can work well in this circuit.

    I compared the two OPA devices' datasheets. It seems that OPA320 has much better CMRR level than OPA2350UA (114dB / 90dB). This may lead to a little more error in OPA2350 as the AMC1301's outputs have a typical 1.44V CM bias; but it could be too small to notice in this application as 90dB (or 74dB in worst case) is already rather well. Moreover, SW caliberation in current sampling ADC can also help to eliminate this error.

    For the 4th question, sharing one set of voltage devider to all three phases are applicable in engineering. In this design we seperated them for easier layout work and better noise rejection (to keep them close to the comparator chips for every phase).

    Best regards,

    Jerome Shan

  • 0 in reply to Jerome Shan
    Expert 1720 points

    Thanks a lot dear Jerome Shan

    Your answer was perfect !

    Best wishes. 

    Dave.