Hi
Checking an tube amp with FFT at 500 mV comes out from 403
Starting at 1 kHz with the intention to go up step by step.
Until 15 kHz the value is fine, next step at 19 kHz the value shown on little windows drop to few mV but on screen the tone is at the same level of the previous kHz.
Then I did a check with a loop out-in of 403 and happen the same results.
Something wrong on my side?
Bye
Walter
Which fft window do you use?
Hi
I use a 64k or 128 k normally
I select the tab Time; easy to set the voltage range but not thr time base
Walter
In addition I can’t find the way to set time base for scope with a QA40x
While with 403 Plot it is easy!!!
Walter
I confirm the same happens with me: RMS Volt windows shows a wrong voltage, once over 19k frequency.
Walter, about time base for scope, when in TIME Display, just zoom in selecting a portion of the scope view.
Or use the (experimental) Oscilloscope Visualizer, which let you choose X-Y values.
Hi
many thx!
About zoom you mean ctl+ and ctl- pointing on screen ??
It is strange that on manual this info are not listed
One more aspect is the signal comes out from QA403 sw.
While with Plot the signal if continuos so the ouput of a device is stable, with 403 sw, with RUN, it is not continuos.
I am wrong?
In case there is a solution?
Walter
Just hold the left mouse buttom to draw a virtual window in the graph. You can repeat the procedure till you get the desired time view. It’s like what you do in “Windows PAINT” program to select an area of a picture.
To go back, click on the bottom right corner “arrow”.
@walge we could provide more help if you would explain what exactly you are trying to measure rather than how you are measuring.
In any case, if you right-click the RMS button on the left a dialog will pop up with minimum and maximum range frequencies for the RMS value. Since the default maximum is 20KHz this means a rolloff as you approach 20KHz which is what you are seeing, I guess. Increase the max frequency to 40KHz and make sure your sampling rate is at least 96KHz for those high frequencies.
Mark
About the signal coming out of the QA403, the QA40x hardware uses bursted outputs for measurements. Try a 96k S/R with a 64k FFT: with me it works pretty stable.
Using the IDLE mode let you have a continuos wave out generator only: Output unstable? - #4 by Gerrittube
Where do we find the (experimental) Oscilloscope Visualizer?
Hi
The strange thing is when I switch to Plot everything works fine, the signal is stable!!!
Differ for 403 sw
Which is the difference?
The Plot , even is limited, is much more easy to use in each test available.
In addition I send a screenshot of a test of THD/Freq
It is strange that end at 30kHz when the setting are 50 kHz with 192k of sampling.
And the curve on top is a level in output that is a perfect stright line and differ from a curve obtained with Freq. response test.
Hi
I will give more infos.
Personally I think the 403 is fine but the related documents are limited as explanation
And seems that are differences in term of performance between Plot and 403 sw.
I will try also Phonoanalyzer
Walter
Walter, I never used Plot program, sorry can’t help you on that.
Anyway, do you know Bob Cordell tutorial on Qa403? I learned a lot reading it as well as the Matt’s blog.
Hi
I read some parts of Cordell, it helps but not in deep
Walter
Hi @Walge, can you give a concise explanation of what you are seeing? I’ll go first so you can try to reproduce.
Connect QA403 in unbalanced loopback (L+ OUT to L+ IN, L- IN shorted).
File->New
Gen1 to 21k and -6 dBV, and enable
Switch to Time Domain and use Left Mouse Button to zoom in.
You should see the following or similar:
Note this is running, and every acquisition appears to be the same. This is because the bursts are coherent (sine always starts at same phase for the acquisition). There is no jumping. The yellow trace appears steady (no movement) even though acquisitions are happening).
Can you duplicate that? And is it rock steady?
Now, you might wonder why the waveform doesn’t appear to be a sine. and that’s because the QA403 time domain will show raw samples. At 48k sampling, a 21k tone has 2.29 samples per cycle . There are tricks you can do to make it look more like a sine (sinc interperpolation, resampling, etc), but my preference generally is to see the raw samples.
Tried the experimental oscilloscope today in the latest October 2026 version. It’s very rudimental and lacks easy settings for timebase and amplitude. With CTRL-Scroll I see some changes AND I get an error. This is not yet the solution I was looking for. At least we need to have an easy way to get a single sinewave filling the window properly. To do this now is a lot of work and difficult to achieve.
