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.

ESDS452-Q1: IO to IO voltage? Constructed differently to drawing?

Part Number: ESDS452-Q1
Other Parts Discussed in Thread: ESDS552, ADS1261, ESDS452

Tool/software:

This TVS diode array is drawn as two separate bidirectional TVS diodes in the datasheet:

The working voltage is +-5.5v, so I'd expect IO channel 1 to be able to handle +-5.5v, and channel 2 to handle +-5.5v.

But if I put +5v on the first channel, and -5v on the second channel, it conducts as if it was internally constructed like this:

Which is absolutely not what the diagram in the datasheet implies, and also isn't how other dual channel TVS I have used before have behaved like.

For example, this dual bidirectional TVS from Diodes Incorporated does not have this unwanted behavior:
https://www.diodes.com/assets/Datasheets/DESD3V3L2BTQ-DESD24VL2BTQ.pdf

  1. Is there a part of the datasheet that implies this limitation?
  2. Is there an alternative part I can use?
  • Hi Alexander, 

    Yes, our bidirectional devices are internally constructed like the image you have drawn. I understand how we have it shown in our datasheet can be misleading, but it technically is meant to show the functionality of the device and not the internals. 

    In regards to the testing you have done, are you seeing the device conduct between the two channels? If so, then this is something we can check in our lab and see if we see the same issue. 

    Also, is there a specific application you are needing +5 on one IO and -5V on another? It might be better to use our ESDS552 which has a higher voltage so there isn't a potential for conducting between IOs.

    Best,

    McKenzie

  • I am protecting a divided down reference input to a ADS1261 with a 0.2x divider after the TVS:

    So, according to the datasheet - the max analog input on any pin is AVDD/AVSS +- ~0.2, so 14v cutoff means 14v*0.2 = 2.8v.

    So, the goal is to have each individual input capped to +-2.8v, but allow differential voltages up to ~10v. The problem is with +5v on one and -5v on the other IO pin, this shorts out my input as the internal construction is NOT as pictured as you mentioned. I'd argue the functionality of the device is not depicted correctly.

    Choosing the suggested ESDS552 seems like it'd clamp at 19v, which would allow the analog inputs to see 3.8v.

    Having the datasheet be drawn this way directly affects these kinds of applications of TI dual channel TVSses, whereas the part from Diodes Incorporated does not exhibit this behavior. If all of the TI TVS do this, I am forced to not use a TI part.

  • Hi Alexander, 

    I have already discussed about this with my manager and we will work on updating the datasheet to reflect this. Because I understand this can be confusing and misleading. 

    In regards to what the analog input sees, would 3.8V be too high? It looks like the abs max on those pins is ~7V. 

    Best,

    McKenzie

  • These ratings here - I want to clamp ESD to it. AVSS/AVDD = +-2.5v - so, +-2.8, but * 1/0.2 as it is divided => ~14v.

  • Hi Alexander,

    Understood. Since most of our 2-channel bidirectional devices are similarly designed, the only solution I can think of it using 2 1-channel bidirectional devices. We have ESD451  which is still a 5.5V working voltage device with a clamping voltage better than the ESDS452. 

    Best,

    McKenzie