aioscilloscope / Learn / Waiting for trigger

Triggering: why the scope says "waiting for trigger"

The scope is on, the probe is on the signal, and the screen says it is waiting for a trigger. The trigger decides when the scope takes a picture, and it is waiting for a moment that is not coming. Here is what each setting means, the real case that sent us looking, and the settings for a serial line.

Single shot, rising edge at 2.96 V, on a serial line resting at 3.41 V with nothing being sent: no edge comes, and the scope says so.

CH1What the trigger does

The scope samples all the time, but a screen of samples taken at a random moment shows a different slice of the signal every time. The trigger picks the moment: when the source channel crosses the level in the chosen direction, the slope (rising: upward through the level; falling: downward). Every capture lines up on that moment, so a repeating signal stands still on the screen.

CH2Auto, Normal and Single

SweepWhat it doesUse it to
AutoWaits briefly for a trigger, then captures anyway. There is always a trace, still when triggered, drifting when not.find a signal and see roughly what it is
NormalCaptures only on a trigger and keeps the last capture on screen while it waits.watch a repeating signal, or one that comes now and then
SingleCaptures once, on the next trigger, then stops.catch one burst, a power-up, a glitch

Only Normal and Single can wait forever. On NovaOS (the trigger sheet's Sweep row has Auto and Normal; Single is its own key), when two seconds pass without a trigger in Normal or Single, a banner over the plot says so, for example "No trigger at 2.96 V on CH1", with a Level 50 % button, and the run status reads NO TRIGGER. In Auto the scope keeps capturing and marks the trace UNTRIGGERED. Before the first capture the screen says "Waiting for a trigger" and what is armed, such as "single shot armed on CH1". If the source channel is off, the banner says "Trigger source CH2 is off" and offers the first channel that is on.

CH3Slope and level

The level must lie inside the signal's swing: a 0 to 3.3 V signal never crosses 4 V, and never crosses −1 V either. In the middle is safest: it is furthest from noise at either end and catches slow edges at a well-defined point. Level 50 % on NovaOS puts the level halfway between the lowest and highest points of the trace on screen. On a running signal that is the middle of its swing. On a flat line it puts the level on the line itself, where noise alone may trigger it: it cannot find a signal that is not there.

CH4Set the trigger, step by step

  1. Choose the trigger source. The channel the signal is on: CH1 here. A trigger on a channel that is switched off never fires.
  2. Choose the slope. Falling for a serial line (its start bit), Rising for most pulses, Either when you do not know.
  3. Put the level inside the swing. Between the signal's lowest and highest voltage, ideally in the middle: 1.6 V for 3.3 V logic, 2.5 V for 5 V logic.
  4. Choose the sweep. Auto to find a signal, Normal to watch a repeating one, Single to catch one burst.
  5. Make the signal happen. Arm first, then make the device send. An idle line has no edges, whatever the level.

Our real case: a single shot that never came

We armed Single with a rising edge at 2.96 V on an XBee's serial output, and nothing happened. The line sat at 3.41 V, its idle level.

The level was not the problem: 2.96 V lies inside the line's 0 to 3.41 V swing, and every time the line came back up after a low bit it would cross 2.96 V upward. The problem was that the line was idle: nothing was being sent while the scope was armed, so there was no edge at any level. Sending a command to the radio made the scope trigger at once.

Two things would still have been better. A falling edge catches the start bit, so the capture begins at the start of a character rather than somewhere inside one. And 1.6 V, the middle of the swing, leaves room for noise and for a line whose high level sags; 2.96 V is only 0.45 V below the idle level.

Trigger settings for a serial (UART) line

Falling edge at 1.6 V: the scope stops on the first start bit, so the capture begins with the first character.

Once a decoder is set up, NovaOS can also trigger on the decoded bus itself (the decode sheet's "Trigger on B1"), for RS232, I2C, SPI, CAN and LIN.

Still waiting? Check these

CheckWhat to do
Is the signal there at all?Switch to Auto: a flat line means nothing is happening, or the probe is on the wrong pin.
Is the source the right channel, and on?The trigger source must be the channel with the signal.
Is the level inside the swing?Compare it with the trace in Auto; put it in the middle.
Is the slope right?A line that only ever falls once (a reset going low) never makes a rising edge.
Is it a one-off?Arm Single first, then make it happen.

Questions

Why is my waveform rolling or unstable?

The scope is not triggering: in Auto it captures anyway, at random moments. Put the trigger source on the signal's channel and the level inside its swing, and it stands still.

What is the difference between Auto and Normal trigger?

Auto captures even without a trigger, so the screen never goes blank. Normal captures only on a trigger and holds the last capture while it waits.

What is trigger holdoff?

A time after each trigger during which the scope ignores new ones. On a burst of pulses it makes the scope trigger on the first pulse of each burst rather than on any pulse.

How do I capture a one-time event?

Set the trigger for it, press Single, then make the event happen. The scope stops on it and keeps it.

More to try

Every guide