1
\$\begingroup\$

I am using a open-collector output comparator (ROHM BA8391G) to enable a voltage regulator when Vin > 12V is detected (comparator output will be HIGH when Vin > 12VDC). Vin can be in the range of 12-28V. (Note that I am aware that in reality, the output will not go HIGH until Vin is ~13.4VDC)

I am using a voltage divider circuit to feed Vin to the non-inverting (+) input of the comparator and connecting the 3.3V supply directly to the inverting (-) input. I do not plan on using feedback since the Vin voltages will not be very granular.

schematic

simulate this circuit – Schematic created using CircuitLab

Does the 3.3V supply connection to the comparator need a series resistor to limit current? And more generically, when do the inputs need current limiting resistors?

I have seen some examples online that do not have a series resistor. Such as this TI design guide

However, the datasheet for the component says "Excessive input current will flow if a differential input voltage in excess of approximately 0.6V is applied between the input unless some limiting resistance is used." My interpretation of that is that if the difference between the inputs is >0.6V than there can be excessive current flow...? The largest possible difference between the two inputs is when Vin=28V; Vdiff = 3.7V

Here is the simplified schematic from the comparator's datasheet that I am using: enter image description here

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Please change your schematic to indicate the PN of the comparator you're using, rather than the TL081 op-amp. \$\endgroup\$
    – The Photon
    Jun 13, 2019 at 15:15
  • 1
    \$\begingroup\$ Your schematics and your text do not comply with respect to the used device, so please edit adequately! Anyway, some comparators and operational amplifiers have a very limited range for the differential input voltage, so a series resistor could help; additionally, a pair of antiparallel diodes between the inputs might be used to limit the differential voltage... \$\endgroup\$
    – aschipfl
    Jun 13, 2019 at 15:18
  • \$\begingroup\$ The very first sentence specifies the part I am using. The PN I am using is not available in the circuit simulator. This detail should not impede a competent engineers ability to answer the question. \$\endgroup\$
    – zme
    Jun 13, 2019 at 15:29

3 Answers 3

1
\$\begingroup\$

Your question needs some attention because you draw a schematic with a TL081 which has JFET inputs meaning that there is an almost zero current flowing into these inputs.

But you are actually using the BA8391G which has this internal schematic:

enter image description here

I see no JFETs, I see PNP inputs.

That means that there will be a small current (typically 50 nA) flowing out of the inputs. This 50 nA will cause a small voltage drop across any resistors at the input. In your schematic the resistors R1 and R2 are of such a low value that this voltage can be ignored as the 50 nA will create a voltage drop of less than 1 mV.

If you would be using higher values for R1 and R2 then that 50 nA could be of significance and then you would want a similar voltage drop at the 3.3 V input of the comparator so that both inputs have the same voltage drop.

This is why you sometimes see a resistor added in series with an input, not so much to limit the current but to make the DC offset the same. Dave from the EEVBlog made a video about this, watch it here.

For your situation this does not really apply as:

  • R1 and R2 have quite low values
  • your application does not require to have the lowest DC offset possible

If you do not apply a higher voltage than the comparator's supply voltage and not lower than ground level then no damaging current can flow.

You could add a resistor (it does not harm to add a 100 k resistor) but it would not make the circuit better in any way.

So: No, it makes no sense to add a resistor.

\$\endgroup\$
3
  • \$\begingroup\$ Thank you so much!! Very informative! \$\endgroup\$
    – zme
    Jun 13, 2019 at 15:44
  • \$\begingroup\$ @SunnyskyguyEE75 can you explain how the 0.6 differential would create a situation of excessive current? \$\endgroup\$
    – zme
    Jun 13, 2019 at 15:46
  • \$\begingroup\$ I can only surmise. The datasheet is not clear about this. It says Vid=+36Vmax applied to Vin- and Ii= - 10mA max at Vid>0.6 \$\endgroup\$ Jun 13, 2019 at 16:00
2
\$\begingroup\$

TI parts have 1k series internal inputs to limit current to 10mA with 20V.

enter image description here

Rohm parts chose not to include this.

But your 30k series will limit the current to < 1mA.

\$\endgroup\$
5
  • \$\begingroup\$ Thanks for the informative response. So the current limiting is for the protection diodes. That makes sense. But if the datasheets says that current limiting is needed when the differential between the inputs exceeds 0.6V, doesn't that imply that there is some diode connection between the inputs? Not just the rails? The voltages at the inputs will never be very close to the rails. I added the "simplified schematic" from the comparator datasheet to the original post for reference. \$\endgroup\$
    – zme
    Jun 13, 2019 at 15:34
  • \$\begingroup\$ YES but ESD diodes used for protection are rated for only 5mA dc because of their small size and speed.Just like CMOS \$\endgroup\$ Jun 13, 2019 at 15:36
  • \$\begingroup\$ Sorry but can you clarify what the "YES" was referring to? The "simplified schematic" does not show diodes connecting the two inputs together so i don't understand why 0.6V differential between the inputs means anything. \$\endgroup\$
    – zme
    Jun 13, 2019 at 15:41
  • 1
    \$\begingroup\$ Yes it is current limiting for negative differential voltage and current. Yet Positive differential voltage MAX applied to Vin- = Vee (-) + 36V \$\endgroup\$ Jun 13, 2019 at 16:21
  • 1
    \$\begingroup\$ Otherwise (very large) series Rin * Iin (small) = creates input offset voltage . \$\endgroup\$ Jun 13, 2019 at 16:28
1
\$\begingroup\$

Comparators and opamps have high impedance and low current leakage on the input. You can check this by looking at the input bias current in the datasheet. You can think of the input to most comparators and opamps as a resistor having higher than 1GΩ of resistance.

If the opamp/comparator has voltage protection diodes, then the current can increase if the voltage gets close to the power rails, in this case you would need to protect the diodes that are inside of the opamp and current limit the input with a resistor.

\$\endgroup\$
2
  • \$\begingroup\$ So if the voltage at the inverting input will always be 3.3V, does that mean I could omit the series resistor since it is far away from the rail voltages? \$\endgroup\$
    – zme
    Jun 13, 2019 at 15:38
  • 1
    \$\begingroup\$ Yes, if it is always below the rails no series resistor needed. \$\endgroup\$
    – Voltage Spike
    Jun 13, 2019 at 15:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.