5
\$\begingroup\$

I am trying to make a simple Schmitt trigger circuit with the TLV9302. Input voltage is 5V. On the non-inverting input, I have a voltage divider from the 5V supply using two 10k resistors, with a 1M hysteresis resistor. The inverting input has a signal that measures a temperature using a 1k thermistor. The unused op-amp in the dual package is set up as a voltage follower with the non-inverting input connected to GND.

Schematic snippet showing a Schmitt trigger circuit - the non-inverting input has a voltage divider from 5V using 10k resistors, the output connects to the non-inverting pin through a 1M resistor for hysteresis, and the inverting input is a voltage divider with the low side connected to a 1k thermistor

According to simulations and calculations, this should give a window of ~25mV centered at 2.5V. However, the midpoint reads ~2.4V. This does not seem like a huge difference, but 100mV will make a difference in my application.

I removed the hysteresis resistor, and just measured the voltage of the divider. The top resistor (VDD5 to input pin) reads 2.6V, the bottom resistor (input pin to GND) reads 2.4V. I confirmed the divider resistors are a solid 10k by removing them from the board, and also using brand new resistors for a second attempt (they are 0.1% tolerance, so I am not surprised they read properly)

There are no other connections to the node, except the non-inverting input of the op-amp. By my calculations, this means that ~20uA is apparently flowing into the non-inverting input of the op-amp??

As far as I can tell, the board is free of shorts. The op-amp inputs are not clamped to each other internally (one reason I chose this part). I'm a bit at a loss of what's going on here. The only thing I can think of at this point (the last bastion of a troubled designer) is the op-amp is damaged somehow; I may try replacing it.

Anyone have any ideas?? Thanks!

Datasheet: https://www.ti.com/lit/ds/symlink/tlv9302.pdf?ts=1750967442834&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FTLV9302

Edit: here's a screenshot of the PCB design in Altium. The standard trace thickness is 25 mils. I ordered boards from OSH Park. C2 and C3 are not populated (I removed them from the schematic screenshot for clarity), they were included for potential noise filtering.

Screenshot of the circuit board design in Altium

\$\endgroup\$
3
  • \$\begingroup\$ --"I removed the hysteresis resistor, and just measured the voltage of the divider. The top resistor reads 2.6V, the bottom reads 2.4V." -- That makes no sense. How can you have two different voltages from the two divider resistors which are supposedly connected at one point? \$\endgroup\$ Commented Jul 3 at 13:36
  • \$\begingroup\$ Sorry that was poorly worded - the voltage across the top resistor (VDD5 to input pin) measures 2.6V, the voltage across the bottom resistor (input pin to GND) reads 2.4V. Since I have the hysteresis resistor disconnected, and both R9 and R10 are 10k, that means 260uA is flowing from VDD through R9, and 240uA is flowing from the divider node through R10. Which indicates 20uA must be flowing into the op-amp. \$\endgroup\$ Commented Jul 3 at 14:59
  • \$\begingroup\$ Then you have a bad component or an incorrect connection. \$\endgroup\$ Commented Jul 3 at 15:05

3 Answers 3

Reset to default
4
\$\begingroup\$

The typical input current for that op-amp is +/-10pA with the shown configuration.

You have an apparent unexplained 20uA which is 2 million times higher. Assuming your '5V' is actually exactly 5.00V, of course.

Possible explanations (more or less in order of my guess as to likelihood from your limited information):

  • U1 pin 8 power supply connection missing (eg. not soldered) on op-amp (causing the op-amp to be powered at ~1.8V through the input protection network).

  • Damaged op-amp

  • Damaged or incorrect resistors

  • Severe contamination of the PCB

Suggest a good visual inspection under a microscope or magnifier, and measure voltage on pin 8, and if that doesn't turn up something (and you don't mind butchering your PCB a bit) cut the trace to pin 3. If you do mind, then remove the IC.

Visual inspection should look at the solder joints, at the traces (for breaks or hairline shorts) and at the components (for mechanical damage such as a scratched resistor).

\$\endgroup\$
1
  • \$\begingroup\$ Thanks for the comment! The 5V is exactly 5V, measured directly at the pins of the op-amp. The output of the op-amp, measured both at the output pin and on the edge of the feedback resistor reads 0V or 5V properly, indicating it is soldered on. I tried replacing the resistors, and made sure to measure them before/after soldering on the board, and they read the correct 10k resistance. The PCB looks fine, I assembled it with a hot plate and solder paste, the passives are 1206 packages and the op-amp is an SOIC, so pretty easy to hand solder. I may try changing the op-amp today though. \$\endgroup\$ Commented Jul 3 at 12:49
3
\$\begingroup\$

My guess would be that you have a thermistor of 10K. This leads to a nominal voltage at the inverting input of 4.54V, at 22-25 degrees.

Then you are exceeding the common range for the opamp. Notice in the datasheet that the maximum allowed common mode input voltage is: Vcommon = Vcc-2V, so in your case 3V. The device can start to behave odd if not damaged.

enter image description here

Source: TLV9302 datasheet

\$\endgroup\$
7
  • \$\begingroup\$ The OP is powering their op-amp from 5V, not 3V. \$\endgroup\$ Commented Jul 3 at 0:40
  • \$\begingroup\$ @TimWescott I don’t think that’s what was said. The way I understood it is VCC-2V=3V which is the case since VCC is 5V as you say. \$\endgroup\$ Commented Jul 3 at 3:53
  • 1
    \$\begingroup\$ @TimWescott Yes I intended to say that the common mode voltage is maximum 3V. \$\endgroup\$ Commented Jul 3 at 5:36
  • \$\begingroup\$ Exceeding the common-mode input range of a CMOS op-amp won't do that unless you also go outside the power supply rails. \$\endgroup\$ Commented Jul 3 at 6:49
  • \$\begingroup\$ The thermistor is definitely 1k, and I can confirm this as well with the voltage measured at the voltage divider network there. The 47k and 1.1k in parallel is ~equal to the resistance of ~room temperature, so I see the expected ~2.5V there. \$\endgroup\$ Commented Jul 3 at 12:52
0
\$\begingroup\$

Thanks for your suggestions everyone. Over the past few days I've been checking things over on and off, and finally just removed the op-amp today. Taking it off the board, I did not see any solder bridges or damaged traces underneath. Replacing it with a brand new component fixed the issues, I get all the proper voltages now, and the circuit works! I've never had an op-amp fail like this before, perhaps I accidentally fried it during test set up and didn't realize it. I will keep a lookout for any latent issues, but a Schmitt trigger isn't exactly the most complex design out there :)

Thanks again!

\$\endgroup\$

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.