Could anyone please let me know how large output impedance the uno has? Especially the pin 11 and pin 5. I'm measuring the output voltage from the pins. I connected my probe directly to the GND and pin 55 or GND and pin 11. The signals are supposed to be 8Mhz PWM by program. But the measured waveform is weird, overshoot and undershoot. Now I'm going to match the impedances between uno and oscilloscope. But I have no idea how to make it. Could anyone please let me know how to match them?
Ringing is an artifact of stray inductance in the oscilloscope probe ground-lead. To see accurate waveforms from logic signals you need a low impedance logic probe - normal oscilloscope probes are high impedance optimized for investigating analog circuitry without loading it.
The output transistors in the ATmega chips are roughly 30 to 40 ohms resistive, given the information in the datasheet.
You can make a low-impedance divide-by-10 probe by using 50 ohm coax, with a 50 ohm resistor to ground at the 'scope end and 470 ohm series resistor to the signal being tested. You have to keep the ground connection as close to the signal wire as possible and not let a loop form, such loops will always ring.
Impedance matching is used for maximizing output power. In this usage though the output is used to drive other circuitry for power control. It is not used to extract power.
Here all you care about is fidelity, which means a low-impedance logic probe. Most people
don’t bother, just ignore the ringing (unless its actually in your circuit and not the 'scope
probe!). Keeping the ground clip close to what you are measuring and using a very short
ground clip wire will help keep stray inductance under control. Ground planes also help.
nownuri: The signals are supposed to be 8Mhz PWM by program.
?? could you explain how to make a 8Mhz PWM signal with a 16Mhz micro. Leo..