Total Pageviews

Thursday, 9 February 2017

VMA Simple Spectrum Analyser: Important bug fix

Hi,

Shame on me!

When I first started playing with my first SMA Simple Spectrum Analyser device, I noticed that all individual measurements are sent from the device as 4 Bytes.

On the original NWT series, the first two Bytes form a 10 Bit value, which is repeated on the second two Bytes. With all my tests using my ADF4350/ADF4351 based device, the second and fourth Byte was always zero.

This led me to believe that the ADF4350/ADF4351 based device could only produce an 8 Bit measurement.

And so I programmed the software accordingly...

Only yesterday a new user (thanks!) complained about a strange clipping on Signals with more than  -40dBm. And he told me that with WinNWT4 those Signals are displayed correctly.

My first reaction: impossible.

Today, after some discussion with said user I realized a huge mistake I made: the ADF4350/ADF4351 can of course do 10 Bit resolution. I just never fed a signal with more than -50 dBm!

Arghhh! How stupid of me. Always playing with TV, CATV and SAT signals, I never tried to feed a signal from a signal generator with -30 dBm, -20 dBm. heck 0 dBm (it is safe to go up to +5 dBm).

This is how it looks like:

Picture 1: Clipping due to my software only evaluating the first 8 Bit instead of the whole 10 Bit

Picture 2: WinNWT4 showing the signal correctly

As a result, I sat down and corrected all routines, adding the missing two bits.

It does work now:

Picture 3: Same signal shown correctly using 10 Bit resolution of the ADF4350/ADF4351

 Please download the latest version!

Regards,
Vitor

No comments:

Post a Comment