• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 3 answers
  • Subscribers 16 subscribers
  • Views 370 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.

TS5A21366: vs TS5A23166 when Switch OFF (IN=L) status, CON2 waveform difference

Roy Chou1
Genius 14355 points
Part Number: TS5A21366
Other Parts Discussed in Thread: TS5A23166

Hi Team,

Please see below measurement. We found that TS5A21366 has some undershoot but TPS5A23166 is ok when Switch OFF (IN=L) .

Below is the measurement of TS5A21366. And TS5A23166 CON2 waveform is smooth so that I don't attach it. May you help let me know which spec may point out the issue?

We think that the spec is OFF isolation, but the datasheet list the same value. May we let me know your comments? Thanks.

Regards.

Roy

  • 0
    TI__Mastermind 46797 points

    Hi Roy,

    Where is the Grey Scope Shot taken - I am a bit confused where the shot is being taken.

    That being said - 

    Off Isolation would cause the ripple you see in the pink (COM2) line when the NO2 line is switching. The off isolation at 1MHz with a 50Ohm load is -65dB - but its tested using a sine wave - not a square wave. So higher frequency content of the square wave can leak through which is what looks like what is happening. 

    But I am not entirely sure where the grey line is being measured - if you could verify that for me as well so I can check what spec you should be looking at.

    Please let me know!

    Best,

    Parker Dodson

  • 0 in reply to Parker Dodson
    TI__Genius 14355 points

    Hi Parker,

    Measured point was the COM2. 

    Below is our measurement setup. First we measured the TS5A21366 and stored the waveform as grey.( Both of NO2 and COM2)

    And then we swapped the chip from TS5A21366 to TS5A23166 and also stored the waveform. Both of NO2 waveforms were overlapping and so you only see yellow NO2 waveform) But we separated the COM2 waveform. As you see, the TS5A21366 wasn't smooth.

    Regards,

    Roy

  • 0 in reply to Roy Chou1
    TI__Mastermind 46797 points

    Hi Roy,

    Thanks for the clarification!

    Please let me know if I am wrong, but based on the setup the only thing that is changing is the IC - everything else including layout are exactly the same - correct?

    However I believe the issue is off isolation

    TS5A21366 Off Isolation:

    Of Isolation TS5A23166

    The TS5A21366 has a steeper slope for off isolation - which means that higher frequency energy is going to make more of an impact at the output for the device. This is important as you are running square waves through the device - which have higher frequency content - especially when the input signal goes from HI -> LO or LO -> HI. This seems to be the issue that you are seeing. One when the square wave is going from LO -> HI the square wave is curved - which means less higher frequency content on the rise - which you can see in your results as the peak is lower on the LO -> HI transition. The HI->LO transition is steeper and doesn't have the curvature the rising waveform has - this means more high frequency content. HI->LO transitions also have the worst leak through.

    With all of this mind though has the customer  been able to recreate the results for both IC's with different IC's or have these tests only been ran on one pair of parts. While the one performing slightly worse isn't very surprising I am not sure if this is the only cause of the ripple - if possible trying running the tests with new IC's and see if the same result is coming up. If it is then most likely it is just how the part operates.

    Please let me know if you have any further questions.

    Best,

    Parker Dodson

  • 0 in reply to Parker Dodson
    TI__Genius 14355 points

    Hi Parker,

    Thanks for information. Our condition is 3.3V, 100kHz and 3.3V, 400kHz. It's hard to point out the value in chart. Do you have RAW data to check the more accurate spec?

    We understand that we used sinewave to do the measurement. I think the reason is that sinewave is easy to see the variation. Is it correct? But we think square wave could use the OFF isolation spec to be the reference. Do you agree with us?

    Regards,

    Roy 

  • 0 in reply to Roy Chou1
    TI__Mastermind 46797 points

    Hi Roy,

    We typically don't share the raw data and one thing to note is that our test is using a VNA with a 50 Ohm load other loading conditions could possibly create worse conditions so the results you see in application may deviate due to difference in testing - but for your application a good estimate according to the charts is as follows. 

    TS5A23166:

    100KHz ~ -85dB

    400KHz ~ -73dB

    TS5A21366:

    100KHz ~  < -75dB (chart doesn't go that low at 100KHz)

     

    400KHz ~ -70dB 

    With that being said odd harmonics are going to also be present which are going to have lower amplitudes but will have less off isolation to affect them. I do think the TS5A23166 is going to have worse off isolation performance than the TS5A21366 which you have seemed to prove.

    However the off leakage seems a little high so have you been able to to test a different pair of fresh IC's to see if the results are the same? 

    Best,

    Parker Dodson

  • 0 in reply to Parker Dodson
    TI__Genius 14355 points

    Hi Parker,

    Thanks for detailed.

    Do we have OFF isolation chart when V+ = 3.3V? I only see 5V spec in datasheet.

    Regards,

    Roy

  • 0 in reply to Roy Chou1
    TI__Mastermind 46797 points

    Hi Roy,

    We don't have the 3.3V data - but the 5V data and 3.3V data curves aren't going to be much different. A good example of this is shown in the 23166 datasheet - the green and blue lines are essentially drawn over each other with a few areas of better performance from the 5V supply but the deviation is about 1 to 2 dB so it isn't a huge difference between the supply voltages. This is partially due to the Source to Drain capacitance not being heavily affected by supply voltage.

    Best,

    Parker Dodson