Slower Rise time from digital pin with 50 Ohm Scope Input

Official Hardware UNO Family UNO R3
1

The rise-time from a square wave from one of the digital pins I'm measuring is around 10nsec with a 50 Ohm input impedance on the Oscilloscope.

I working on a project where I need a faster rise-time so I tried combining 6 of the digital pins, all outputting a square wave with the PORTb command so they are synchronous. Now I'm measuring 3.6nsec.

It looks like as the current requirement decreases from each pin, the rise-time out of the pins gets faster. Why could this be. I don't think this is a limitation of the scope that I'm using so its something within the arduino that's slowing down the rise-time as it outputs more current.

Would anyone know why this is happening?

EDIT: I am connecting the output of the digital pin to a scope with a coaxial cable. I am not using a 10x probe. I know this is not standard practise but my experiment requires the use of a coaxial cable.

2

Probe capacitance and rather high output impedance of the port pins. Those pins are not designed to deliver more than about 20 mA safely.

Use a 10X probe.

3

Those outputs have their own rise time characteristics but they are also charging some external capacitance. As you connect more outputs together they can collectively charge that external capacitance faster.

Also you can use things like an external Schmidt trigger or even just a logic gate to decrease the rise time.

And why do you think things are slowing down? 3.6ns is faster than 10ns

4

The ATmega chip (Arduino Uno, etc.) is rated for an "absolute maximum" of 40mA which works out to 125 Ohms minimum at 5V.

I don't know if the lower impedance affects the rise time but you stressing the chip beyond its specs.

2 Likes
5

Can you switch the 'scope input from 50 Ohm to the typical 1 megohm input impedance? If so, we usually use the 1 meg setting and a 'scope probe for most measurements.

6

Hi @ronnniely!

Just for curiosity: Have you compensated the scope probe?

7

Again, the port pins are not designed to work into a 50 Ohm load. You MUST use a buffer amplifier if you want to drive such loads.

8

Even if its charging an external capacitance which in this case is 20pF i.e the input capacitance of the oscilloscope, the rise time should not exceed ~2ns

9

I get that but I want to understand why

10

my experiment setup needs the scope to be at 50Ohm. But you're right, the rise-time does not degrade if I keep the scope input at 1Mohm

11

The capacitance is charged and discharged by CURRENT, not voltage! So current limits of the pin electronics limits the current maximum, resulting in slower pulse changes.

2 Likes
12

For a capacitor, the basic defining equation is

Q = CV

The instantaneous charging current is

I = C dV/dt.

Plugging in C = 20 pF, dV = 4V, dt = 10 ns, get I = 10 mA.

I suspect the total capacitance being charged is higher than 20 pF.

13

Hi, @ronnniely
Welcome to the forum.

What scope are you using, and does it have 50 Ohm input impedance.

Tom.... :smiley: :+1: :coffee: :australia:

14

So use a 50 ohm termination resistor in your circuit and monitor it with a high impedance scope probe.

15

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.