Strange output with RC one shot and comparator

[aiq25]

I’m working on a simple RC based one shot that goes into a comparator to provide a 40-50 ms reset pulse during board start up for a digital IC. I’m using a 4.7uF and 10kOhm resistor that goes into a comparator with a 2.5V threshold. The circuit is powered from a 5V voltage supply. Sorry I don’t have the schematic handy right now but will try to upload it later.

In simulation everything works fine. The comparators outputs a nice stable digital signal. When I prototyped the circuit I used a “fast” push-pull comparator but my output has a slope to it, almost 20-30 ms to raise to 5V. This is strange behavior since the output should just be a square wave. Did anyone else have this issue? I tried two different comparators.

I also tried to use a dual op-amp, with one acting as a buffer amplifier but the results were the same.

[T3sl4co1l]

Waiting for circuit.

And, why not use a regular reset generator like MCP102?

Tim

[aiq25]

Waiting for circuit.

And, why not use a regular reset generator like MCP102?

Tim

I was trying to use the parts we already had in the lab. Plus this circuit apparantly is used in a production design, so it should've worked but I'm not sure what is happening.

Attached is the schematic. I don't have the oscilloscope screenshot, so here is a crude drawing of the output.   

I used a MCP6541 in the prototype.

[uer166]

Do you expect the comparator to work as a comparator at all voltages, including 0 and as it rises to the nominal levels?

[aiq25]

Do you expect the comparator to work as a comparator at all voltages, including 0 and as it rises to the nominal levels?

The reference voltage has a time delay and I will try removing the 100nF. I didn't expect the output wave to have a raise, since it is a comparator I expected the output to reach 5V once it hits the reference voltage of 2.5V, otherwise the output should be 0V. What I see at the output of the comparator is a raise in voltage, over 20-30ms. Is this assumption wrong?

[T3sl4co1l]

What is the "5V" from, how long does it take to reach 5V?

Tim

[Zero999]

Waiting for circuit.

And, why not use a regular reset generator like MCP102?

Tim

I was trying to use the parts we already had in the lab. Plus this circuit apparantly is used in a production design, so it should've worked but I'm not sure what is happening.

Attached is the schematic. I don't have the oscilloscope screenshot, so here is a crude drawing of the output.   

I used a MCP6541 in the prototype.

Get rid of C1 and put a 1M resistor from U1's output, to the +input. A diode across R4 is also a good idea to protect the comparator's input, when the power is removed and enable the circuit to reset quickly, by providing a discharge path for C3. I'd also increase R4 and decrease C3: timing capacitors should generally be as small as possible.

[aiq25]

What is the "5V" from, how long does it take to reach 5V?

Tim

In this case the 5V is coming from a lab power supply. I haven't timed it but it has a fairly fast response (no where near the ramp up of the output).

[aiq25]

Waiting for circuit.

And, why not use a regular reset generator like MCP102?

Tim

I was trying to use the parts we already had in the lab. Plus this circuit apparantly is used in a production design, so it should've worked but I'm not sure what is happening.

Attached is the schematic. I don't have the oscilloscope screenshot, so here is a crude drawing of the output.   

I used a MCP6541 in the prototype.

Get rid of C1 and put a 1M resistor from U1's output, to the +input. A diode across R4 is also a good idea to protect the comparator's input, when the power is removed and enable the circuit to reset quickly, by providing a discharge path for C3. I'd also increase R4 and decrease C3: timing capacitors should generally be as small as possible.

Thanks! Will try that.

[aiq25]

Turns out the circuit design was fine, it was the power supply ramp up. Not sure how I missed that.