THD measurementThis would be a huge effort when e.g. the table could present at the end the THD.
Actually there´s no alternative from counting the shds together.
When one of our tested inverter shows a suspicious THD on the analyzer, we connect a scope at the output and watching the shds on the FFT screen.
Table function allows you to display the shds in numeric form, as mentioned above it would be cool if the scope would display the thd also.
EDIT:
http://www.sengpielaudio.com/calculator-thd.htm
THD table, calculator
I played some more:
It was easy to
verify the calculation of the HP35665A THD value displayed;
I did it this way:
- Define the number of peaks to take into calculation at the 'Harmonic' menu, e.g. 10.
- Grab the 11 peak values by marker and put them into Excel.
- Do the calculation, e.g. following the paper Rob found:
Just linked this in another thread for Peter:
https://www.researchgate.net/publication/362542224_Total_Harmonic_Distortion_THD_analysis_utilizing_the_FFT_capabilities_of_modern_digital_storage_oscilloscopes
Having read it now, it seems to be kind of a seminar paper on bachelor level, but it does give the basics you need.
Result: HP35665A shows 0,199% THD, calulation gives: 0,19931% So HP do it by rulebook of course.
On the Siglent:1st Step: by hand- I fiddled with the reference level to get the maximum count of harmonics into the table, which was 8 in my case.
- Typing the values into XLS
- calculate
The result is 0,2288% which is in the same ballpark.
EDIT:
2nd step: automatedAfter exporting the FFT as CSV and adding some quick and dirty formulas, we get 0,2151%, which is feasible, because the formulas do take 10 peaks beside the base frequency:
![](https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/?action=dlattach;attach=1664548)
The xls ist attached too. The export was some 500.000 records, I shortened it to 5000*, because all the action is in the first 5000 records.
- Open your csv export in XLS, convert csv into columns (make sure to address the decimal delimiter right) and paste the columns 1:1 into the left part of the attached XLS.
- Fill in the base frequency (yellow), and there you go.
![Cool 8)](https://www.eevblog.com/forum/Smileys/default/cool.gif)
*) In case of regular need one should think about a better config of frequency range and number of points.
![Face Palm :palm:](https://www.eevblog.com/forum/Smileys/default/facepalm.gif)
And finally,
to do the HP3310A justice: After some more adjusting it now shows a THD of 0,118% at 5kHz and 1V RMS.