¿ªÔÆÌåÓý

  1. Tinysa
  2. Topics

Locked #bug_report #bug_report

 

Found something odd during a Phase Noise measurement, only when you enter more then 6 digits behind the dot for example 10.000 000 000 MHz then the Center Frequency becomes 349.996 MHz instead of 10.
 

Behaviour here (tinySA_v1.3-338-g42ce69a):

Enter 10.000 000? ? ? ? --> 10.000 000 MHz? ? - OK
Enter 10.000 005? ? ? ? --> 10.000 005 MHz? ? - OK
Enter 10.000 000 0? ? ?--> 100.000 000 MHz? - WRONG (interpreted as 100 MHz)
Enter 10.000 000 5? ? ?--> 100.000 000 MHz? - WRONG (interpreted as 100 MHz)
Enter 10.000 000 00? ?--> 350.000 000 MHz? - WRONG (interpreted as 1 000 MHz)
Enter 10.000 000 000 --> 350.000 000 MHz? - WRONG (interpreted as 10 000 MHz)

So it looks like the decimal point gets moved towards the right by the number of digits behind the 6th decimal place. (1 000 MHz, 10 000 MHz ( >350 MHz) are cut to 350MHz)?
 

Exactly. It not only in the Measurement menu but also the Signal Generator Frequency setting.
(I also checked the NanoVNA which doesn't have this bug.)

Found this playing with a gpsdo from the HBTE group.
 

Retested with?tinySA_v1.3-347-g31eb02e

It's almost fixed but there is still a problem when entering more than 8 decimal points.
(entered values are with unit "MHz")

Enter 10.000 000? ? ? ? --> 10.000 000 MHz? ? - OK
Enter 10.000 005? ? ? ? --> 10.000 005 MHz? ? - OK
Enter 10.000 000 0? ? ?--> 10.000 000 MHz? ? - OK (a maximum of 6 decimal places are displayed on the screen)
Enter 10.000 000 5? ? ?--> 10.000 000 MHz? ? - OK (a maximum of 6 decimal places are displayed?on the screen)
Enter 10.000 000 00? ?--> 10.000 000 MHz? ? - OK?(a maximum of 6 decimal places are displayed?on the screen)
Enter 10.000 000 000 --> 1.410065 MHz? ? ? ?- WRONG

Another finding:
START frequency unit is always fixed to MHz. (Guess just a typo - its fine for STOP and CENTER and SPAN)
- entering some value in Hz shows MHz
- entering some value in kHz shows kMHz
- entering some value in MHz shows MMHz

Example:
 

It is nearly April 1st so here is a really dumb, stupid question.
Why do you need to enter a frequency to 8 decimal places. 6 decimal places is 1Hz, why do you need one hundredth of a Hz.
I did say it is a stupid question.
 

Your question is absolutely justified. There is indeed no practical use (at least for me) of entering a hundredth of a Hz. ;)
That said, I always feel that errors that occur in one place where they are not a problem can be potentially serious in another place.
That's the reason I post things like this, even if they don't affect me at all. Please let me know if sharing the serious things is enough...
 

That is so common....
"More decimals = More accurate"
In my last job, people were entering GPS locations of field based equupment to ridiculous lengths. I challenged them, pointing out that their coordinates defined locations to the thickness of the paint coating , whilst their GPS equipment was only accurate to 3m. But it happens.
Just my 2.000000 cents worth..
John
 

I feel with both John and @SleepingCat.

John is right, especially when it concerns measurements.
People can be so seemingly pretentious, although in most cases it's ignorance I'm afraid.

But I think @SleepingCat has also a point, another point.
If something goes wrong in an otherwise useless case (but you don't know why), what else can happen?

It becomes almost philosophical.
Are you from the school "If it ain't broke, don't fix it" OR from "If it ain't broke, you're not trying"? ? ;)?

Just my 2¡À1 cents,
BenS
 

On Fri, Mar 25, 2022 at 12:15 AM, @SleepingCat wrote:
There is indeed no practical use (at least for me) of entering a hundredth of a Hz. ;)

¡°It is the mark of an instructed mind to rest satisfied with that degree of precision which the nature of the subject admits, and not to seek exactness where only an approximation of the truth is possible.¡± - Aristotle



That said, I always feel that errors that occur in one place where they are not a problem can be potentially serious in another place.
I'm with SleepingCat.? Had I come across that bug/quirk, my mind would immediately jump to: "Hmm,? what else does that affect?"
 

I'm happy with the bug report as this revealed a bug in all fractional input that could have caused weird things.
If you calculate a frequency in MHz you often get more digits then needed and it's simpler to enter all instead of counting if you reached the tinySA resolution.

--
For more info on the tinySA go to https://tinysa.org/wiki/

1 - 10 of 10