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.

TPS3701 positive and negative voltage window detector

Other Parts Discussed in Thread: TPS3701, TPS3700, TPS3838, TPS3890, OPA1622, TPS386000

Hello dear TI community.

Im using TPS3701 as a positive and negative voltage windows detector to enable or disable my LDO's.

But I've got some questions regarding my design.
I saw that negative voltate sensing could be done with TPS3700 (via Ti design note), I assume it's still valid for TPS3701? 

Another question is my bigger concern, I've got two independent power lines (TPS7A4701 and TPS7A3301) which are powering some delicate opamp circuit.
Both comparators are measuring voltage and send enable singnal totally independent - which can make circuit unstable or even damaged due to power supply failure (only one regulator will work ATM).

I don't want to use only one comparator to feed enable singnal to both LDO's because it's totally pointless in my opinion (it will not solve my problem anyway).
So my question is - is there any way to combine two independent outputs from those comparator into one useful signal?

That signal will be later used to feed both EN pins of my regulators.
I just want to avoid situation when one voltage will be outside comparator range and LDO will be turned off and another one will be running anyway.


I'm looking to achieve mutual enable ONLY when both voltage from two comparators will be within range, if one voltage will be outside the window range then both regulators will be turned out.

If possible I want to stay with TPS3701 because solution size and cost and precision are really good with those chips.
And calculating everything is pretty easy too, I've got everything already calculated.

My circut is using two regs adjusted to +/- 18V, and thery feeded by +/- ~20-25V so both comparators are adjusted to monitor voltage within that range.
BTW I know that there some chips in the market which could be used to sensing both VCC and VEE voltages in one package, but they can be configured only as a UV or OV detector which not suit me in this case.

Thank you for help and time :)

  • Mateusz,

    Yes, TPS3701 can do negative sensing just as the TPS3700 device. You can follow the same TI design note. Also, because the TPS3700 and TPS3701 have Open-drain outputs, you can tie the outputs together so that there is one single output for power good or power fault detection. If you need help calculating the resistor values, please use the tool on the project folder page under the "tools & software" tab.

    I hope this answers your questions. Please let me know if you have any more questions or need more support. Thanks!

    -Michael

  • Thank you for answer.

    But I assume that my pull up resistor need to be connected to all outputs? And here's the deal is that possible and safe?
    I'm asking because one comparator is powered up by VEE and another from VCC.

    So if it's possible to tie all outputs into one (wired-or/and) without any problem due to diffrent power supply it will be perfect :)
    I will connect all outputs to single pull up resistor connected to VCC then.

    BTW TPS7A3301 positive voltage EN range is somehow limited to few volts.
    So may I use voltage divider or zener diode to feed safe voltage into EN pin?

    Pull up resistor will be connected to voltage between 20-25V which is far above TPS7A3301 enable range.

    Another question, if I connect a LED to comparator output (4 outputs as wired-or/and) will it change output voltage?
    Because I can connect LED in two flavours as a OK diode when voltage is within range, or NOT OK when voltage is below/above range..
    When output is high (all of them) LED will shine - OK state, when output is low NOT OK LED will shine.

    Which option is more safe and will not false trigger voltage regulator? This question is more related to NOT OK LED.

  • Each output needs a pull-up resistor so if you tie outputs together, a single pull-up resistor will work so long as the output current is in spec. The pull-up voltage needs to be the same for all tied outputs, but this voltage can be different than the voltage to power the device.

    The LED will have a voltage drop across is when ON and you can find the voltage drop in the datasheet for the LED. I recommend signaling NOT OK when voltage is out of range, because ideally, most of the time the circuit will be OK so you don't want LEDs ON and burning power the majority of the time. That's just my opinion but either way will work.

    You can use a voltage divider or Zener or just set your pull-up voltage accordingly.

    I hope I answered your questions. Let me know if anything is unanswered or if you have any more questions. Thanks!

    -Michael
  • Thank you for clariification.
    I was little missguided and confused, because I know how wired-or/and work but most designs are powered from VCC, and I've got VEE too.
    I was worried about failure if one comparator is power from totally diffrent power supply than it's pull up resistor.

    I've got two another question regarding my design :

    1) is there any chance to use two LEDs for NON OK operation? Let's assume I want to stay with two LEDs one per power line.
    Then it will be much easier to know which power line is faulty.
    But I still want to use one out signal to feed both LDOs (four outputs will tied together).

    So there will be VEE FAULT and VCC FAULT LEDs.

    2) is there exist any chip in your catalog that could be used as a RST signal delay?
    I just want to add simple delay for output before it reach another IC.
    I'm not looking for any other function like a phase invertion etc.
    Simple delay with control via R or C (or both) will be fine.
  • 1) When you tie the outputs together, they become the same node so I am not sure two LEDs will be any different than one LED that signals fault when the outputs are tied together. The outputs must be independent for separate LEDs to work.

    2) Yes we have fixed delay supervisor / reset ICs and programmable delay. For low current consumption, I recommend TPS3838. For high accuracy and performance, I recommend TPS3890. The delay is set according to the equation in their datasheet by changing the capacitor value on the cap delay pin (CT)

    Please let me know if you have any more questions! Thanks.

    -Michael

  • 1) I was worried about it, all outputs will work as a one, so only one LED could be used.
    I was thinking that there might be some way to separate it to achieve better visual information about failure.

    2) how about using those chips to delay signal from combined outputs of two TPS3701?
    I don't need anything else than a simple delay which will be used to feed OPA1622 input or other chip with enable input when I will see reason to :)

    Is there exist simple solution to delay that signal (with option to customize that delay time with simple capacitor/resistor/network) I'm looking forward to know what are you sugessting.
    i want to avoid any glitches and false triggering etc.

    Some competetive comparators/voltage monitors already got RST output (besides of OV/UV out) with option to delay RST signal.
    But those solutions are rather expensive.
  • Mateusz,

    You can add capacitors to the INA, INB pins at the input to delay the output RESET. You will need a trial and error to see what delay you want since you are simply delaying the input from crossing the threshold.

    Devices with positive and negative voltage monitoring with delay tend to be more expensive. We have a 4 channel device that does this called TPS386000. This will be overkill for your circuit however. Let me know if the additional capacitors on TPS3701 will work for you.

    -Michael