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.

TLV3605: Negative Supply for TLV3605

Part Number: TLV3605
Other Parts Discussed in Thread: TLV3801


I am connecting the TLV3605 to a negative voltage (For example -1.7V) at its VEE pin. I know that the LVDS output will refer to the VEE pin as its ground reference, in this case, connecting a -1.7V to the VEE pin, the common mode voltage will become -0.5V.

However, I am DC biasing the output common-mode of TLV3605 back to +1.25V as below.

The comparator input will go up to 200MHz. I would like to know that with this practice,

1) In long term, will this degrade the comparator?

2) Will this cause long term reliability issue?


  • Ooi 

    Thanks for your post. I’m going to have to review your plan more tomorrow but I think our tlv3801 seems to be a better fit since you can maintain ground reference on the output even though the input is powered from split supply. Please review and see if you agree. 

  • Hi Chuck Sins,

    I agree that tlv3801 fit the split supply on the input and maintain ground reference on the output requirement for me but I also need the adjustable hysteresis control which is not available in tlv3801. As the adjustable hysteresis control is more prioritize here, no choice that I am going for the tlv3605 with the DC biasing practice to fit my requirement.


  • Understood and thanks for the follow up. I’m just concerned this won’t work. I’ve done this re-centuring technique after AC-coupling but not without the caps. We use this technique when using an oscilloscope to 50 ohm terminate to system ground. We can get away with this since we test with 50% duty cycle waveforms. I’ll get back to you tomorrow after I consult with our design team. 

  • Hi Chuck Sin,

    Just FYI that our application don't allow AC coupling on the circuitry.


  • Ooi

    I pretty much figured that ac-coupling was not acceptable but thanks for verifying.  Unfortunately, I don't have good news for you.  As a result of our LVDS output stage architecture, your proposal to split the 100 ohm load into two 50's where the common point resets the VCM to proper LVDS protocol levels will not work.  I can imagine that you are assuming our output stage can be treated like an ideal 3.5mA current source but that is not the case.  The only suggestion I have for you is to consider using two comparators.  The first stage could be the TLV3605 that provides the adjustable hysteresis pin feature while the second stage is our TLV3801 which can level shift the LVDS signal.  I realize this solution comes at quite a high cost since it requires two devices but this is the only way I can see this working.  


  • Ooi, I am closing this post. Feel free to reply back or open a new thread if you have any more questions.