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.

TPD2E001: current leakage into power rail (when power switched off)

Part Number: TPD2E001
Other Parts Discussed in Thread: TPD2E2U06

Hello,

in a recent design I'm using the TPD2E001 to provide ESD protection to two LVC-devices connected to external BNC connectors, see here:

However, I figured that for the input signal ("IN") the TPD2E001 through path "A" provides current to the +5 V rail when the power supply is switched off. 

With the power supply switched off and a 1kHz signal provided to the "IN" connector, the +5 V rail (green trace) looks like this:

This current leakage into power rail (with the ESD bypass capacitor C1 even smoothing the input signal considerably) prevented a different component (also connected to +5V rail) from triggering its power-on-reset when the power was turned on. Consequently this component failed due to improper initialization.

I fixed this by breaking the connection of the Vcc pin to +5 V rail (see arrow at "A" in schematic above). The TPD2E001 explicitly describes this kind of application. Furhter, I could not see any signal degradation so far.

Still I wonder what ramifications there are due to not using a Vcc connection. As TPD2E001's diodes are not biased by Vcc any more,
a) their parasitic capacitance might increase (as detailed here, p. 24). and

b) there might be a new path "B" for AC signals to ground.

Do I have to worry about them? 

After the TPD2E001 gave me so much trouble, I am looking for a replacement device that needs no VCC connection. Would TPD2E2U06DRLR be a good choice? At least the datasheets recommends it.

Thanks.

Dan

  • Hi Dan,

    I'm a little confused on the waveforms you sent. The way I understand it, the yellow trace is the 1 kHz signal measured from the "IN" connector, and the green trace is the measurement of point "A" at the same time with the power supply turned off. Is that correct? If so, where is the 1.3 V bias on the green trace coming from? I figured it should be close to zero with the power supply turned off. 

    Regarding your concerns (a and b), you are correct on both. Without a Vcc bias, the parasitic capacitance of the diode will increase which will lead to higher signal attenuation. It will also result in a higher breakdown voltage of 10 V for the diode. Please refer to this E2E post for more details: https://e2e.ti.com/support/interface-group/interface/f/interface-forum/959387/tpd2e001-vcc-pin-usage 

    Like the app note you referenced earlier states, it is better to use a device with no Vcc pin. TPD2E2U06 is a much better device and will provide the protection you need.

    Regards,

    Matt Smith 

  • Hi Matt,

    Yes, your understanding is correct. The green trace was measured with a probe tip attached to the +5 V rail, e.g., at point "A". It was done with the power supply off. 

    I concluded that the positive bias stems from the current leaking from the higher-voltage input signal at the "IN" connector along the path drawn in red (see schematic) through the hiding diode into the lower-level +5 V power rail (and all other devices connected to it). That's evil.

    So I'll need to update the design with ESD-protection diodes that don’t require a Vcc connection, e.g., the TPD2E2U06.

    Thanks.

  • Interesting. Well, like I said, TPD2E2U06 should definitely fix this issue. Glad to hear you can update the design. Please let me know if you have any other questions.

    Regards,

    Matt Smith

  • Thanks Matt.  I read the E2E post that you referred to. Seems to me that it describes a use case (signal at an IO pin of the TPD2E001 and Vcc pin connected to a supply that is switched off) very similar to mine. I just wanted to know why you where "confused" and asked where the 1.3 V bias is coming from? It seemed obvious that the leakage to Vcc occurs through the hiding diode. I hope you agree with my findings.

    That said, my main intention for posting was to give a note to others about the issues/challenges/pitfalls that might exist when using this device for ESD protection.

  • Hi Dan,

    I agree that the leakage would go through the hiding diode. The confusion came from the green trace staying at a constant 1.3 V with a few mV spike when the yellow trace goes high since the yellow trace is the only voltage source on the system at that time. I would assume that the green trace would look very similar to the yellow trace unless there is an additional DC bias that I'm not seeing.

    Clearly I'm wrong in my assumptions since the data shows otherwise. Would you mind explaining this for my own understanding as well as anyone else who might have this same confusion?

    Regards,

    Matt 

  • Hi Matt,

    Oh, now I get it. So I misunderstood your statement ("I figured it should be close to zero with the power supply turned off. ") and the reason for you confusion completely. I thought you would not agree about attributing the leakage to the TPD2E001. Sorry for being obtuse.

    Regarding the form/level of the green curve (signal on Vcc), I admit that I've not pondered it extensively.  I was happy when I figured out the reason why it exists and how to get rid of it. As I mentioned earlier, I think it's smoothed out by the various capacitors connected to it.

    Best regards,
    Dan