• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 62 subscribers
  • Views 356 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.

UCC21550-Q1: UCC21550 How to calculate DT when Rdt=1K?

Guohai Lin
Intellectual 2771 points
Part Number: UCC21550-Q1
Other Parts Discussed in Thread: UCC21550, UCC21551

Dear teams,

Rdt=1k is not within the range of the datasheet. Is there any other way to calculate DT at this time?

Thanks！

  • 0
    TI__Genius 17486 points

    As stated in datasheet section 8.4.2.2,

    tDT = 8.6 × RDT + 13

    For a 1k DT resistor, the deadtime will be 8.6*1+13, or about 21.6ns.

    The usable deadtime range for the UCC21550 is 27-870ns, so you may have trouble getting 21.6ns accurately.

    Best regards,

    Sean

  • 0 in reply to Sean Cashin
    TI__Intellectual 2771 points

    Hi Sean，

    Then why is the usage range of RDT marked as: 1K<RDT<100K?

    Thanks！

  • 0 in reply to Guohai Lin
    TI__Genius 17486 points

    Hi Guohai, 

    Your highlighted text indicates that "1.7k" is the lower bound. I will try to test the deadtime with a 1k DT resistor and get back to you with the time value. It probably works fine, just the deadtime duration starts to deviate from this equation by more than the stated 7% tolerance.

    What is your customer's deadtime target? Also, what is their secret to switching so quickly and efficiently? When I try to use a deadtime that low I start to have massive shoot through current. They must have a very small, low capacitance switch and an excellent gate drive circuit!

    Best regards,

    Sean 

  • 0 in reply to Sean Cashin
    TI__Intellectual 2771 points

    Hi Sean,

    This will be applied at an operating frequency of 800KHz. The customer is still in the early design stage, so he wants to know the DT when RDT=1K as a reference, which is not officially used. So is this a resistor value that can be calculated?

    Please allow me to ask a new question:
    The working isolation voltage of UCC21550 is VDC<2121V. What will happen if it is applied in an environment exceeding this voltage value? Is it just a reduction in service life? Or will it cause damage to some parts of the IC? Please provide specific failure symptoms.
    Thanks!

  • 0 in reply to Guohai Lin
    TI__Genius 17486 points

    Hi Guohai,

    I did not yet get the chance to measure the deadtime. The designer has replied to me that 

    "Dead-time less than 1.7k falls into the minimum dead-time region.  The tolerance is extremely wide and there can be a large deviation between rising and falling dead-times.  I have data for only 3 parts and RDT=1k results in between -0.6 to 5 ns of dead-time." 

    Best regards,

    Sean

  • 0 in reply to Sean Cashin
    TI__Genius 17486 points

    The working isolation voltage is a DC spec. Section 6.6 of the datasheet indicates that the transient voltage can be much higher (10kV in oil).

     

    The cause of failure is arching and dielectric breakdown. Inside the device, there is a silicon dioxide dielectric that has a very high permittivity. Often the weaker link is the breakdown of the air around the device, and the isolation is limited by "creepage" and "clearance" distance. There are different standards for isolation. The UCC21551 is has a double barrier and is considered "reinforced". Here is an article to learn more: https://e2e.ti.com/blogs_/b/analogwire/posts/reinforced-isolation-when-1-is-greater-than-2?keyMatch=REINFORCED%20ISOLATION  

    Other types are basic and functional isolation.

  • 0 in reply to Sean Cashin
    TI__Intellectual 2771 points

    Hi Sean,

    I found this picture in the datasheet. Does it mean that the increase in isolation voltage just reduces the life of the chip? For example, when Vrms=3200V, the IC can be used for about 1.E+07 seconds?

    Thanks！

  • +1 in reply to Guohai Lin
    TI__Genius 17486 points

    Hi Guohai,

    Yes, that is exactly the theory behind this graph. It is based on industry recognized standards. 

    This graph is based on a lot of assumptions about "accelerated life". It is difficult to have an accurate measured test for 54 years! High voltage does lead to "hot carrier injection degradation" in semiconductors, which destroys the chip dielectric over time.

    3200V is still well below the 5000Viso for this device. What is the application? The high voltage is likely not going to be applied across this barrier continuously, 100% of the time. 

    Here are some links with more information:

     https://www.ti.com/lit/wp/slyy063/slyy063.pdf

    https://www.ti.com/lit/an/slla353a/slla353a.pdf

    https://e2e.ti.com/blogs_/b/analogwire/posts/what-the-new-din-vde-v-0884-11-2017-01-standard-means-for-your-isolated-designs

    Best regards,

    Sean