I’m trying to measure the Frequency Response (FR) of a moving coil SUT (step-up transformer) and think I’m hitting some limitations in the QA40x app’s Frequency Response tool.
Even though I’ve attempted to extend the app’s frequency range by setting the sampling rate to 192k and setting the End Freq of the ExpoChirp to 96k, I suspect the app is still limited somewhere/.
Hopefully the following 3 screenshots helps to visualize the limitation in the QA40x app
QA40x FR
QA40xPlot FR
my Oscilloscope’s built-in FR tool
I’d love if they all captured the exact same info, but the QA40x’s FR doesn’t capture enough of the HF response to be useful.
An MC transformer primary is a very low impedance, even a typical transistor/FET based MC stage is circa 100R- enough to cause significant loading of the generator with consequent distortion, voltage drop and frequency dependent non-linear voltage drop. The generator output z is halved for SE vs Differential. My QA403 is ~100R per leg (SE), so 200R bal.
You’re using the R channel for reference in the QA40x S/W where neither of the other pieces of S/W or hardware have that facility. But why would you expect a SUT to have a flat response towards DC? It can’t. I’d be taking some spot measurements with a low impedance drive source via an attenuator for sanity checks and I think you’ll find the 40x S/W is showing a more accurate response if the SUT is indeed a MC one.
I’m NOT expecting a flat response to DC. I’m trying to determine how sensitive this Moving Coil SUT is to secondary loading.
I was hoping to see that the response from the QA would be similar to what I’m seeing on the oscilloscope with the understanding that the input and output impedances for the 2 devices are slightly different. (100k QA input vs 1Meg oscilloscope input in this instance).
What’s really interesting, IMHO, is that the QA40xPlot software shows a much more accurate FR plot than the native QA40x software. I assume this is a better apples to apples comparison since they are using the exact same hardware (and I tried making the stimulus in the 2 apps as close as possible).
To my untrained eyes, it seems there’s a limitation in the QA40x app.
I didn’t realize that C (or A) weighting could be turned off. After turning it off, the FR measurement looks much closer to what I would expect.
This is one of the reasons why I posted this note in the first place. I’m glad the problem was user-related and not software related. I’m going to edit the title of the original post.
Can you be more specific in describing your problem?
”can’t run it” and “doesn’t work” could mean a variety of different things.
One possibility: Do you have the Right (input/display) channel turned on? If you’re copying my “settings”, you’ll see that I only have the Left Display channel enabled.
Hi @walge, the power measurement is really only going to make sense when you are measuring tones. In Freq Response mode, you are emitting a sweep, and the measurement won’t make sense. This is a bug. The Peak Power Watts measurement handles it correctly, the RMS power measurement doesn’t. This will be fixed for the next release (the correct result here is you’ll see NaN displayed, which means Not a Number.
And just as a refresher, the difference between the peak power measurement is that it will look for the peak in volts and convert to power based on your indicated load.
RMS power measurement will take the RMS measurement across the start/stop frequencies you’ve indicatd, and report power based on that reading.
I point this out because sometimes folks correctly note that RMS power isn’t a thing (the correct term is average power). And they are right, but the RMS measurement has RMS in the name to distinguish it from the peak-derived measurement.
