Measurement

The D43 program only uses one set of cursors. If you have a dual trace oscilloscope (as I do) then you can only make measurements on one channel at a time.

The cursors have three lines. Two intense lines on the outside, and a lighter line in the center. Measurements are made on the center lines.

You must manually set the "Timebase" and "Vertical deflection" settings in the D43 software to match the values selected on the switches of the oscilloscope - the software cannot read the switches. Additionally, you must set the "Multiplier" selector. My oscilloscope has the usual 1X and 10X, as well as an AC 100X that can be combined with the 1X and 10X settings. Use only the multipliers that your oscilloscope really supports.

The "Time," "Frequency," and "Vpp" readings are based off of the cursor positions only, and work at all times. Here's an example of a typical cursor measurement:

That's a 1kHz square wave signal. The amplitude is 3.9V.

The "VRMS" and "VRMS(AC)" measurements are only made when the digitizing display mode is active. The digitizing mode depends on proper setup, and depends on the trace properly filling the grid. If these conditions aren't met then the RMS measurements will give completely wrong results. The only way to tell if the conditions are met is to look at the digitized signal.

Use the gray cursor to set the zero line for the RMS measurement. This will properly include any DC offset in the RMS calculation.

RMS(AC) will remove the DC offset and calculate only the AC RMS value.

The RMS measurements are totally independent from the blue cursors.

Here's an example of the RMS measurements:

This is an example of an improper RMS measurement:

The trace doesn't fill the digitizing area (the bounds of the original grid) resulting in random data on the left side of the screen. The RMS measurements are completely incorrect.

For proper RMS measurements, you need multiple cycles of the waveform on screen. Ideally, you would have ten or more cycles. Try to use a sweep speed such that you get about two cycles per division. For a 1kHz signal, that would be a sweep speed of 2 milliseconds per division. Also make sure that the trace is sufficiently bright. A wiggling red digitizer trace will deliver bad results. Get as many cycles as can be reasonably managed, and don't expect to really get those 5 significant digits that the display shows. You'll probably get about two useful digits. Anything beyond that is wishful thinking and a figment of the computer's imagination.