TLC372: Some questions about TLC372 application circuit.

Most likely it keeps the comparator input st ground when there’s no input or the system is powering up.

Chuck

Hi Lai,

looks like R26 shall present a low impedance to the comparator input in order to make the input line more immune against EMI (capacitive stray coupling, etc.). Remember that TLC372 is a CMOS chip with very high input impedance.

R26 can also form a voltage divider in combination with a source resistance of input signal driver.

Or, R26 might be needed to convert a current input signal into a voltage signal.

There can be many reasons for R26.

What I don't like so much in the circuit is C26 and C30. Never put a cap directly at the output of a comparator. When going low a very high current pulse can be the consequence, possibly eroding the signal ground and causing ground bounce within the comparator chip. Even oscillation of the comparator can occur. Caps directly at the output of a comparator is a no go. If you want to filter the output signal of a comparator by the help of a cap, then always insert a suited current limiting resistor.

Another disadvantage of the circuit is the lack of an added hysteresis. This can result in heavy oscillations.

Kai

Chuck Sins said:

Most likely it keeps the comparator input st ground when there’s no input or the system is powering up.

Chuck

Thanks for your answer!

Hai Lai

kai klaas69 said:

Hi Lai,

looks like R26 shall present a low impedance to the comparator input in order to make the input line more immune against EMI (capacitive stray coupling, etc.). Remember that TLC372 is a CMOS chip with very high input impedance.

R26 can also form a voltage divider in combination with a source resistance of input signal driver.

Or, R26 might be needed to convert a current input signal into a voltage signal.

There can be many reasons for R26.

What I don't like in the circuit is C26 and C30. Never put a cap directly at the output of a comparator. When going low a very high current pulse can be the consequence, possibly eroding the signal ground and causing ground bounce within the comparator chip. Even oscillation of the comparator can occur. Caps directly at the output of a comparator is a no go. If you want to filter the output signal of a comparator by the help of a cap, then always insert a suited current limiting resistor.

Another disadvantage of the circuit is the lack of an added hysteresis. This can result in heavy oscillations.

Kai

Dear Kai,

Thanks for your detailed answer.

Here is the circuit diagram of comparator opamp input signal.

And, here is a statistical chart of signals with different amplitudes (desired results).

And this is the actual measured results.

From picture 3:

1.Can see that it can not distinguish the two peaks, as shown in Figure 2

2. The vertical axis is much higher than the desired value in Figure 2.

How to increase the ability to distinguish pulse amplitude, I mean the resolution of opam comparator?

Thanks,

Hai Lai

Dear everyone,

Here are some pictures of the signal.

Figure 1: blue is the signal at pin 1 of TLC372, green is the same signal as pins 2, 6 of TLC372.

Figure 2. blue is the signal at pin 7 of TLC372, green is the signal at pins 2, 6 of TLC372

Figure 3. blue is the signal at gate 0.

Hi Lai Hai,

I'm not sure whether I understand you correctly. Can you please explain in a bit more detail?

Kai

kai klaas69 said:

Hi Lai Hai,

I'm not sure whether I understand you correctly. Can you please explain in a bit more detail?

Kai

Dear Kai,

What additional information do you need me to provide?

Hai

Hi Lai Hai,

what do the two statistical charts mean in detail? What have they to do with the comparator circuit? Or by other words: Where do you see from that the comparator circuit is not able to distinguish pulse amplitudes good enough?

Kai

kai klaas69 said:

Hi Lai Hai,

what do the two statistical charts mean in detail? What have they to do with the comparator circuit? Or by other words: Where do you see from that the comparator circuit is not able to distinguish pulse amplitudes good enough?

Kai

Dear Kai,

Two statistical graphs with horizontal axis representing pulse amplitude.

And the vertical axis represents the number of pulses.

As in the first chart, it can be seen that pulses with amplitudes are clearly expressed according to the characteristics of the signal source.
And the second chart seems to have overlapped.

Hai

Hai

Looking at the waveforms, and not the statistical charts, the only potential issue that I see is the loading on the output (cap load that Kai mentioned previously).  This is causing the rising edge to be very slow.  Is it possible that is causing your system to not capture pulses that are close in time?  You are using a relatively fast response comparator but then you are slew rate limiting the outputs by connecting a cap load directly on the output.  At the least, I would re-run your experiments with the cap loads removed.

Chuck

Hi Lai Hai,

nice to hear that you could solve the problem. Congrats!

Kai

kai klaas69 said:

Hi Lai Hai,

nice to hear that you could solve the problem. Congrats!

Kai

Dear Kai, 

Thanks for your support.

Hai