Afonso Teles

> how do I get the starting and ending x-values for the fourier transform data?

If you query WFMOutpre? you'll get the y and x scale. You can assume it starts at DC (unless you're using the advanced spectral analysis, in which case you'll have to query it there) and then multiply the scale by 10 to get your total span.

> When I get the waveform data in my python script, only incorrect y values are included. What am I doing wrong to get the wrong y-values from the math channel?

Two things come to mind:

1) The data you're getting is linear data, you need to convert it to dB.

2) Make sure the endianess of your selected encoding and your read function match.

Hi "Big" Jim,

Indeed, that's an error in the manual. I'm submitting a report to have it fixed, thank you for bringing it to our attention!

Hi Pablo,

There's several layers of corrections, that's why your question is hard to answer. There's also not much public documentation on this, so I'd just suggest trying it, that's the fastest way.

And yes, that makes sense. The only internal memory for saving waveforms are the 4 REF slots, which you could also try using.

Hi "Big" Jim,

SRFBINARY is not a valid data encoding, perhaps you meant SFPbinary. Also note that you don't need to set byt_nr as it's automatically set by data:encdg.
Might also be worth trying FPBinary (just changes the endianess).

For future reference, I suggest querying *ESR? after every write to help pick exactly this sort of problem. I have wrapper functions that do this automatically.
For production code, after proper testing, you can remove the *ESR? queries.

Hi Jonathan,

How's your RS-232 connection setup on the scope and in VISA? Make sure they share the same settings (baud rate, flagging, parity, etc).
Also make sure that your VISA is actually searching for RS-232 devices (TekVISA doesn't by default).

Hi Pablo,

There are several steps of "calibration and correction" on all of our modern scopes, I would need more details to know what exactly you're referring to.

In terms of saving the IQ data through SCPI: you can use the *SAV command to save a .tiq file on the scope and then transfer it with FILESystem:READFile .

Hi "Big" Jim,

Would you be able to share a TekVisa or NIVisa trace of the relevant VISA calls? That would be the best way for me to see what's going on.

Hi "Big" Jim,

That's a very complex question to answer as you're essentially comparing the speed of running FFT + the max measurement on the scope + transferring the result of that measurement to your PC versus the speed of transferring the waveform to the PC and doing the FFT + measurement there.

My guess is that running the FFT and max measurement on the scope and then transferring just the result of max would be significantly faster than the other option, but I don't know the speed of your connection or computer.

If you have the time to test both, that would be ideal.

Hi Pablo,

Correct, the IQ data is the only way to get the full spectrum data (followed by proper windowing, FFT, etc).

The IQ data in the .mat file is an array of complex single precision values (matlab should automatically recognize it when you try to open it), plus a few more values of metadata.

Which corrections are you referring to?

Hi Denis,

I think your if statement might be the problem, I would suggest trying a simpler one.

Also, make sure your signal is toggling high to low and back.