Reset via USB

Using Arduino IDE 1.x
1

I'm trying to understand how arduino uno (current version R3) resets via USB,
which is fine for uploads, but bad for connecting the Serial Monitor from the IDE.
I found http://arduino.cc/en/uploads/Main/Arduino_Uno_Rev3-schematic.pdf and extracted these relevant elements into a LTSpice circuit.

Comparing simulation

with reality, it's easy to identify the pullup resistance of 10k, and to see that a reset is triggered when the RESET signal is pulled down to < 3V, e.g. with a resistor of 4.7k between RESET and GND.
Bigger resistors, e.g. 47k giving about 4V @ RESET, do not trigger a reset.

Now I have problems with the "DTR" ( rather _CTS) signal coming from the 16U2. Is it completely different to my assumption of 5V dropping to 0 to trigger a reset ? Or is there a shortcut across the 100nF capacitor ?
Did I misread the schema? The LTSpice simulation is definitely different from what I experience:
According to LTSpice, reset should rather require a bigger capacitor and should be easy to inhibit with any pullup resistor of 1k to 5k.
The resistor of exaxctly 110 Ohm, which is described in the forum, might rather "work" by overloading the power supply by pulling the max 40 mA ? For me, neither a 200 Ohm nor a 100 Ohm resistor worked to inhibit reset when Serial Monitor connects. Im rather fine with a big capacitor between GND and RESET, but that does not fit to my "shortcut instead of 100nF" theory.

Does anybody know more in that direction, or has an oscilloscope and is willing to share what he sees during a "Reset via USB", please ? ( I don't have a second arduino to build my own oscilloscope -- is 100 µs enough resolution ? )

2

Just in case someone is interested - Now I know more about LTSpice -

Default voltage switching has a very slow slope. Specifing nothing (default) or 0 means the same, and is by far not the possible minimum for rise and fall time.
Selecting 1.e-7 sec ( 0.1 µs if I calculate right), is probably already more idealistic than reality, and shows nicely the theoretical behavior:

I(R1) represents DTR going off at 3 ms and on again at 8 ms

Looking at the data sheet, i learned that the minimal reset pulse time is about 2.5 µs, much less than the resulting around 1 ms DTR pulse.

And, regarding Arduino as a scope , ADC times of >100 µs and pure digital loop times of > 11 µs ( my 16 MHz Arduino Uno R3 ) would show something, but not everything, in that speed range...

3