• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 49 subscribers
  • Views 448 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.

TLV3691: Is there any performance degradation when operating the inputs right at VCC?

Andrew McBain
Prodigy 85 points
Part Number: TLV3691
Other Parts Discussed in Thread: TLV7031, TLV7081

I am making a circuit using the TLV3691 nano-power comparator using VCC = 3.3V.  I want to connect one of the inputs directly to VCC.  The other input is connected trough a 2 MEG resistor.

The 2 MEG resistor should prevent any damage to the part if the second input exceeds VCC.

But is there any performance degradation in input bias current, propagation delay, or input offset voltage when operating the inputs right at VCC?

  • 0
    Guru 140200 points

    Hi Andrew,

    the specified common mode input voltage range of TLV3691 is from (V–) – 0.1V to (V+) + 0.1V. It's not wise to exceed it because the performance will suffer then.

    Why not decreasing the input voltages by using resistive voltage dividers? Eventually, additional diode clamping could be used.

    Kai

  • 0 in reply to kai klaas69
    TI__Mastermind 24260 points

    Andrew

    In general Kai’s comments are quite valid, however, what you have proposed is not uncommon. My only concern is the size of the 2M resistor that you proposed to protect the pin that is being actively driven. You didn’t mention how high the voltage can get. I propose you consider our tlv7031 if you can accept the higher IQ if 335nA. I prefer this device for this application because this device has a unique input protection circuit that does not suffer when driven beyond the supply voltage. As long as the active input stays below 6.5v, you won’t have any performance degradation. That assumes the other input is at 3.3v (same as supply voltage).

    If you can accept an open drain output and leadless package, our tlv7081 is designed to operate above the supply. I’ll leave it to you to explore that device. One input is actually connected to the supply internally, just as you are proposing.

    Whichever solution you choose, there’s no problem setting the switching threshold as the supply. Input bias current will be slightly higher but you need to be more concerned with changes in bias current at high temp when connecting large value resistance to cmos inputs.

    Chuck

  • 0 in reply to Chuck Sins
    Prodigy 85 points
    The comparator is used as part of a protection circuit for a 5V DAC output.  The comparator detects when the output goes too far outside the nominal operating range (0V to 5V) and then an SSR is opened to isolate the output from the fault. 

    The DAC output voltage is connected to the comparators using some resistor dividers with an output impedance of a few mega-ohms.  The large resistances will keep the input current into the comparator below a few micro amps during a fault.
    The expected fault voltages are as high as 36V DC with intermittent (<1s) spikes to ±100V.  The output stage of the DAC output driver circuit won't be harmed by continuous shorts to voltages within ±15V.  The output stage is current limited in such a way that it could withstand brief shorts to ±100V (so that the SSR has time to open).
    I think that whatever comparator I use I will be putting a few micro amps of current into the input stage during a fault.  I looked at the TLV7031 as you suggested.  I like that the inputs remain high impedance when taken beyond the supplies.  That could be a nice feature in some other designs.  In the DAC protection circuit I will be relying on the input clamping diodes so I won't benefit in this case.
    I picked the TLV3691 specifically because it says in the datasheet that the input stage has diode clamps that can tolerate up to 10mA.  The package size of any input protection diodes I would have to add would be as large (if not larger) than the comparator itself, so I would rather have a device that specifically mentions they are present internally.
    The maximum bias stated in the TLV3691 datasheet is 20nA at 125C.  The large divider resistors and bias current together shouldn't create more than 100mV of error in the output measurement.  To account for any errors I set the compare thresholds 2.5V outside of the 0V to 5V range.
    Though I could set the divider ratios so that the thresholds fall well within the supply voltage of the comparator, I can save a couple of resistors in the design if I set the compare thresholds right at the comparator supply voltages.  That's what I was hoping to do here.  I am trying to make the design as small as possible so any diodes or resistors that I can eliminate are a big plus.
  • 0 in reply to Andrew McBain
    TI__Mastermind 24260 points
    You’ve definitely put some thought into your design. I don’t see anything fundamentally wrong with your plan and now I like your idea of limiting the input current to the uA range since over voltage occurrences are expected. In general it’s not a good idea to think of ESD protection diodes as being comparable to external discrete diodes. They can handle mA’s of current but they are not designed to handle that amount of current for any extended amount of time. So limiting the current to uAs is a good decision. Best of luck on your design.
    Chuck