¿ªÔÆÌåÓý

Hello Erik
Good idea (at least I think so..)
I want to be notified.
Hopefully some more of the Windows program too...
TIA?????????????????????????? de Joe.

Hi Erik

Would Like to be informed about beta releases please.

Regards
Jeff

I can imagine an attack and decay time to be optionally set in zero span mode. This can will help spot very short events before you decide to set up a trigger
With attack and decay at default of zero you get the normal zero span mode behaviour.
Would that be useful?
Or is this quasi-peak feature only relevant to people doing EMC pre-compliance testing and thus must be implemented according to the formal spec (which is impossible)?
--

Now the FW has become more stable I'd like to set up a beta test channel where I can announce early releases of new functionality without running the risk of upsetting new users.
The beta FW will be made available in the beta folder inside the regular FW folder at:?
Reply to this message if you want to be notified when new beta releases are available and are willing to try these beta releases.

You should be concerned about tests 7 and 8 failing
Can you open the case and post a picture of the component side of the PCB?
If possible, after removing the shields, they can easily be lifted
--

¿ªÔÆÌåÓý

Hello Linuxers,

I have problems with TinySA-Saver in my Debian 10 LXDE installation. Please look at following dump, you possibly discover what I missed.

TNX, 73

Martin, OK1RR

-------------------------------------------------

martin@martin:~/temp/tinysa-saver-master$ python3.7 -m pip install .

Processing /home/martin/temp/tinysa-saver-master
Collecting PyQt5 (from TinySASaver==0.2.2)
? Downloading (3.3MB)
??? 100% |¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€| 3.3MB 604kB/s
? Installing build dependencies ... done
??? Complete output from command python setup.py egg_info:
??? Traceback (most recent call last):
????? File "<string>", line 1, in <module>
????? File "/usr/lib/python3.7/tokenize.py", line 447, in open
??????? buffer = _builtin_open(filename, 'rb')
??? FileNotFoundError: [Errno 2] No such file or directory: '/tmp/pip-install-vlzc0cdh/PyQt5/setup.py'

??? ----------------------------------------
Command "python setup.py egg_info" failed with error code 1 in /tmp/pip-install-vlzc0cdh/PyQt5/
martin@martin:~/temp/tinysa-saver-master$ python3.7 tinysa-saver.py
Traceback (most recent call last):
? File "tinysa-saver.py", line 17, in <module>
??? from TinySASaver.__main__ import main
? File "/home/martin/temp/tinysa-saver-master/TinySASaver/__main__.py", line 23, in <module>
??? from .TinySASaver import TinySASaver
? File "/home/martin/temp/tinysa-saver-master/TinySASaver/TinySASaver.py", line 24, in <module>
??? import scipy.signal as signal
ModuleNotFoundError: No module named 'scipy'
martin@martin:~/temp/tinysa-saver-master$

Modern EMI receiver (e.g. ~100k€) do the quasi-peak detection inside an FFT scan but it also costs more scan time than normal average/peak detection. The problem with quasi-peak is, that it should be similar to the sensitivity of the human ear. So it has a filter curve similar the sensitivity of the ear in the audio range.?

Without FFT, each reception bandwidth (9kHz/100kHz for distortion measurements) has? to be monitored long enough to filter the audio curve. Even with FFT each FFT segment minimum has to stay that time. I do not know, if any DSP algorithms are public for quasi-peak measurements, but for sure it will slow down the scanning and will need the broadcast bandwidth settings.?

But quasi-peak only makes sense to evaluate the distortion on broadcast channels. Normally the scans are done with average and peek, und only if the level is over the peek limit, the quasi-peak is remeasured, as peek if every time above.? ?

¿ªÔÆÌåÓý

Ok I have noted it. I was just about to throw it away so gave it the old faithful.?

On Sat, Dec 5, 2020 at 12:41 PM, Erik Kaashoek wrote:

So quick attack and slow decay per frequency?
Or only in zero span mode?
Which attack and decay times?

? ?The quasi-peak bandwidths and sweep times are specialized resulting in very slow scan times.? Usual procedure is to first do a peak scan and if all emissions are below the limit no quasi-peak scan is necessary.? If some emissions are above the peak limit then on go back and perform a quasi-peak scan in those particular ranges.? Following the that test procedure can shave hours off your test time versus performing a quasi-peak scan over the entire test range.

? I'm not aware of any sub $1000 spectrum analyzers that offer quasi-peak detection or quasi-peak bandwidths.? Again, I've been out of the EMC field for a number of years and have not kept up with the latest and greatest.

? - Herb

On Sat, Dec 5, 2020 at 05:47 AM, pete waters wrote:

I used plain old 3 in one

Dirt and dust sticks to 3 in one, and it gets thicker and less "lubey" over time.? There are two products I highly recommend for that sort of thing, both from Caig: Deoxit D5 and MCL Moving Contact Lubricant.? I, and a lot of others, won't use anything else.

So quick attack and slow decay per frequency?
Or only in zero span mode?
Which attack and decay times?
--

Or another USB cable?
--

On Sat, Dec 5, 2020 at 11:55 AM, Erik Kaashoek wrote:

The aver16 is a 1/16 running average so it will be a bit like a qpd

? Quasi-peak detection works more like pulse-width modulation (PWM), the higher the repetition rate the greater the output.? A high repetition rate emission will be close to the value of a peak detector, while a low?repetition rate?emission will be below an average detector.

? - Herb

On Sat, Dec 5, 2020 at 09:37 AM, Mat Breton wrote:

Has their been any discussion on adding a Quasi-Peak Detector (QPD) algorithm in addition to the averaging/min/max detection algorithms?

If so, would it likely go into the tinySA app?

? ?I worked in EMC for a number of years and don't remember there being a DSP algorithm for quasi-peak detection.? All our receivers used physical quasi-peak detectors.? That was a number of years ago and maybe the mathematics have advanced since then.

? Best strategy in lieu of not having a quasi-peak detector is to ensure all your emissions are below the peak limit.? The peak emissions should always be higher than the same emissions measured with a quasi-peak detector.? Of course that means that you may spend some time reducing emissions below the peak limit that already meet the quasi-peak limit, but its better than arriving at a test house and finding your device fails the quasi-peak limit.

?- Herb

And that was before he updated to the new FW
--

The aver16 is a 1/16 running average so it will be a bit like a qpd
--

Has their been any discussion on adding a Quasi-Peak Detector (QPD) algorithm in addition to the averaging/min/max detection algorithms?

If so, would it likely go into the tinySA app?