• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 12 subscribers
  • Views 577 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

SN6505B: sn6505b glitch on primary

SeattleEE
Intellectual 480 points
Part Number: SN6505B


Hi! I've prototyped a half wave bipolar circuit to generate +/-13V from 5V. Total primary current in is about 700mA. The transformer is Wurth 1.7:1 760030013  (rated for 350 mA). So, the transformer is being pushed way to hard at this point and I'll switch to a 1A transformers later.

On both primary legs there's a glitch you can see below. Any idea what this might be? Perhaps a byproduct of magnetic saturation?

The topology is as shown below. This is similar to Fig 50 in the SN6505B spec with the ground moved to center of cap stack. I understand there's a requirement that the loads be very close in magnitude otherwise a flux imbalance occurs. Any thoughts on how much disparity there can be before trouble arises and what to look for? From what I've been able to understand, a 10% current imbalance will manifest as a 10% change in V-t product, and if that drops below your minimum, then all bets are off. So, one leg will see V-t product go up, another see V-t go down. Unfortunately, there's not a lot I can find written about this.

  • 0
    Intellectual 480 points

    Pictures didn't upload. Retry.

  • 0
    Intellectual 480 points

    Retry again on pictures. Looks like pasting pictures into post doesn't work?

  • 0 in reply to SeattleEE
    TI__Genius 17100 points

    Hi SeattleEE,

    Thank you for posting to E2E! Small peaks or "glitches" in the D1/D2 switching waveforms are expected when a transformer saturates. For example, a transformer designed for SN6505B shows the waveform below when an external clock is used on the SN6505 to achieve a ~210kHz switching frequency on a >1A load:



    Transformers each behave differently, so the "glitch" style may change or not be present at all when a transformer saturates.

    In some transformers, small glitches can also be observed in normal operation as an effect of the transformer's primary and parasitic inductances, with the peaks growing or shrinking along with the amount of current flowing through the transformer, and transformer 760390013 has a relatively high primary inductance. Can you share these same waveforms with different amounts of current flowing through the transformer?

    Regarding load balancing, TI does not provide recommendations other than operating the bipolar rails as closely balanced as possible. We cannot comment on how the V-t product changes for specific parts in different conditions, but a marginally unbalanced load could cause a transformer to saturate due to a magnetic flux build-up over time.

    In our observations, transformers with higher V-t product margin (like a transformer with 24V-us V-t product when the requirement is 8V-us) have shown better resistance to saturation conditions and recovery from them.


    Thank you.
    Manuel Chavez

  • 0 in reply to Manuel Chavez
    Intellectual 480 points

    > but a marginally unbalanced load could cause a transformer to saturate due to a magnetic flux build-up over time.

    Thanks very much. Do you have a feel for how long "over time" might be? In other words, if the imbalance is present but doesn't result in saturation after X hours, including some time in a chamber high/low temps, and high/low input voltage, can I be confident it won't get worse over time? This would be for a load current that is 30-40% below the point where the glitch occurs.

  • 0 in reply to SeattleEE
    TI__Genius 17100 points

    Hi SeattleEE,

    There are cases where imbalanced loads do not result in transformer saturation, and if no imbalance is present after extended testing for process, voltage, and temperature variations, that can be indicative that saturation will not occur.

    What are different current consumption levels for R_L+ and R_L-? Do you expect this system is unbalanced because of its topology?


    Thank you,
    Manuel Chavez