Feedback for version 0.9a here

Information and discussions on Spectrum Analysis
SteveAndrew
Posts: 119
Joined: Thu Nov 16, 2017 11:06 pm
Location: Gold Coast, QLD, Australia

Re: Feedback for version 0.9a here

Postby SteveAndrew » Wed Jun 13, 2018 9:53 pm

vsonnier wrote:Definitely a problem of the current locale reading numbers, where "," is supposed to be the French Locale convention for decimal separator, opposed to "." in English.
As for all, Stackoverflow may have an answer for this :
https://stackoverflow.com/questions/834 ... ut#8348390

I have found where the problem is. At startup, I check the registry for the location of the SDRplay API DLL. The API version number is returned as a string. The error occurs when trying to convert the string to a double. Anybody whoe's locale uses a "," for the decimal separator is going to experience the same problem. At the moment, the quick fix is to set their locale to English/USA or a similar variant.

Thanks again for the traceback, that was a great help in identifying and locating the problem area.

Reason: No reason

rune sm3mzy
Posts: 3
Joined: Thu Dec 15, 2016 12:17 pm

Re: Feedback for version 0.9a here

Postby rune sm3mzy » Wed Jun 13, 2018 10:17 pm

i had to change the loocale to great britain from swedish now it starts up,earlier it did not do anything via the desktop shortcut or start menu.

best regards Rune
Last edited by rune sm3mzy on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Paul
Posts: 242
Joined: Fri Feb 17, 2017 9:52 am

Re: Feedback for version 0.9a here

Postby Paul » Thu Jun 14, 2018 6:37 am

Sorry about sluggish response Steve - A small forum problem (sorted I believe).
Re: Trace visibility on waterfall:- The attached is the best that I have been able to achieve using contrast and brightness controls - should I expect better?
Everything else is so impressive, it seems churlish to go on about it!
With regard to measurement of harmonics, perhaps this is nly part of what I was looking for, which is the ability to "fingerprint" the source of transmission by measuring frequncies within it - for instance, I believe that if one had a recording of animal (e.g. bat) vocalisations, it might be possible to identify individuals.
The same would presumably apply to radio transmission sources, which could be interesting.
Of course, this may be pie in the sky....but thanks for your response.
Attachments
2018-06-13.gif
2018-06-13.gif (135.2 KiB) Viewed 994 times

Reason: No reason

sdjack
Posts: 11
Joined: Mon Apr 18, 2016 10:13 pm

Re: Feedback for version 0.9a here

Postby sdjack » Thu Jun 14, 2018 7:37 am

SteveAndrew wrote:
sdjack wrote:Tnx for the reply.Yes the buttons are enabled, when i push ie. the 100kz or 10khz button it is still remaining in 1Mhz bandwide, 200khz or 20khz stays at 2Mhz, 500khz or 50khz is fixed at 5Mhz.

I cannot reproduce the problem here. Just to clarify your answer: You say for example, if you select 100KHz, or 10KHz, that the analyser stays at 1MHZ bandwidth, do the 100KHz or 10KHz buttons become selected, but the displayed bandwidth stays the same, or are the 100KHz and 10KHz buttons not being selected at all. If possible can you tell me what happens with the Start/End numbers under the centre freq control on the sweep control panel, and the sweep width figures on the on-screen display when you select the different sweep width buttons?

Also, can you tell me what your region setting is on your machine ? - There have been some reported problems with systems set to French/France. In that instance, the analyser will not start at all, but there may something else connected to region settings that are having an affect. It might be worth changing your settings to English/USA or English/UK if they are not already set to that, and seeing if that makes any difference.


Hi Steve,

thats it. I change my region settings to english uk, it has been german, and now evrything works, brilliant.
Last edited by sdjack on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

SteveAndrew
Posts: 119
Joined: Thu Nov 16, 2017 11:06 pm
Location: Gold Coast, QLD, Australia

Re: Feedback for version 0.9a here

Postby SteveAndrew » Thu Jun 14, 2018 7:38 am

Paul wrote:Re: Trace visibility on waterfall:- The attached is the best that I have been able to achieve using contrast and brightness controls - should I expect better?

A couple of points: Firstly, there is a problem at narrow sweep widths due to the DC spike. This is being addressed and was one of the reasons sweep widths of less than 10KHz have been disabled. The next release will use a non-zero IF which will eliminate the DC problem.
Secondly, at such a narrow sweep width, you might get a cleaner result by increasing the number of FFT bins - I see you are currently set at 16384. I would suggest 32768, or even 65536. This will increase the spectral resolution, but of course, with an added speed penalty - no free lunch here.
Thirdly, it might be worth enabling averaging in the Display panel, try an initial setting of 4 and go from there. The downside would be a reduction in magnitude response speed, once again, no free lunch.

None these really address your question regarding waterfall image quality though, they were just mentioned in passing. It looks as if there is still some range left on the brightness control, I would try reducing the brightness a bit. From what I can see, it looks like you have reached full saturation on the white level, reducing contrast might help as well, although reading your post, it seems you might have tried this already. From my own experience I would have thought you would get a better display than that which you have posted.

The waterfall display is being overhauled, along with a lot of other code. I'll be offering the user an optional colour display instead of fixed monochrome. The waterfall's dynamic range will also be increased which should result in a better display as well.

If none of the above gives you a better display, please let me know, it might be that the waterfall needs some attention.

Reason: No reason

g1hbe
Posts: 354
Joined: Sat Jan 17, 2015 3:28 pm
Location: Cheshire, UK

Re: Feedback for version 0.9a here

Postby g1hbe » Thu Jun 14, 2018 11:47 am

Thanks for a very interesting piece of software. On a rather cursory look at it yesterday I noticed some thing odd. I normally use a Signal Hound SA-44b analyser, so I did a comparison with that.
With the CF set to 3.0 MHz and a signal on 3.01 MHz at -40dBm, the SH shows a peak as expected at 3.01 MHz and -40dBm as in the shot below.

Transferring the same input to the SAS with default settings and set up for the same CF etc, I got a most peculiar display. See the 'picket fence' photo below. Shoving the LNA STATE setting from 0 to 8 corrected it and we end up with the final photo. I'm assuming the central spike is because of the zero IF.
What concerns me is that novices with no other equipment with which to compare may not realise there is something odd going on and assume they have unwanted peaks in their Device Under Test.
BTW, all tests were with my HP8648b sig gen and RSP2 on antenna B using your SAS software.
I hope this info is useful (and that I haven't done something stupid).
301 test SH.jpg
301 test SH.jpg (202.01 KiB) Viewed 975 times
Picket fence.jpg
Picket fence.jpg (183.43 KiB) Viewed 975 times
LNA corrected.jpg
LNA corrected.jpg (176.87 KiB) Viewed 975 times

Reason: No reason
Andy

Tech_Support
Posts: 350
Joined: Mon Jun 01, 2015 7:00 pm

Re: Feedback for version 0.9a here

Postby Tech_Support » Thu Jun 14, 2018 1:12 pm

I suspect what you are seeing here is ADC overload. Moving the RF gain from state 0 to state 8 reduces the RF gain. In my tests, I can see a warning for an overloaded ADC.

Steve can perhaps comment better on this.

Sincerely

Tech_Support

Reason: No reason

Paul
Posts: 242
Joined: Fri Feb 17, 2017 9:52 am

Re: Feedback for version 0.9a here

Postby Paul » Thu Jun 14, 2018 5:12 pm

[/quote] The waterfall display is being overhauled, along with a lot of other code. I'll be offering the user an optional colour display instead of fixed monochrome. The waterfall's dynamic range will also be increased which should result in a better display as well.

If none of the above gives you a better display, please let me know, it might be that the waterfall needs some attention.[/quote]

Thanks for your response Steve, I'm not going to go on nagging about the waterfall - you have plenty nough to deal with, without that.

Best Wishes

Reason: No reason

bobh2k
Posts: 1
Joined: Sat Apr 21, 2018 8:20 am

Re: Feedback for version 0.9a here

Postby bobh2k » Thu Jun 14, 2018 5:44 pm

Excellent piece of software with only one major problem -- Centre Frequency step size is always 1MHz regardless of span size i.e. with a span of 200kHz the centre frequency always increases and decreases in 1MHz increments using the up/down control but it changes in 2MHz steps when using the scroll wheel. The only way around this is to enter the exact frequency by keypad. Thanks for an excellent offering - it is really appreciated.
Attachments
Image 1.png
Image 1.png (10.55 KiB) Viewed 928 times
Last edited by bobh2k on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

SteveAndrew
Posts: 119
Joined: Thu Nov 16, 2017 11:06 pm
Location: Gold Coast, QLD, Australia

Re: Feedback for version 0.9a here

Postby SteveAndrew » Thu Jun 14, 2018 9:57 pm

g1hbe wrote:Transferring the same input to the SAS with default settings and set up for the same CF etc, I got a most peculiar display. See the 'picket fence' photo below. Shoving the LNA STATE setting from 0 to 8 corrected it and we end up with the final photo. I'm assuming the central spike is because of the zero IF.
What concerns me is that novices with no other equipment with which to compare may not realise there is something odd going on and assume they have unwanted peaks in their Device Under Test.

I agree with Tech_Support, it looks like an overload problem to me. It's a bit difficult to know how to address this problem for novices. Some form of auto-ranging may be called for, but then again, that is pretty well what the AGC control does anyway. The instruction manual (such as it is) will be expanded for the next release, and will include a separate tutorial section that covers the use of spectrum analysers in general. I agree, this is something that could cause a bit of confusion for the unwary. I think "before and after" screenshots will probably be the way to go -A picture is worth a thousand words.

Reason: No reason


Return to “Spectrum Analysis”

Who is online

Users browsing this forum: No registered users and 4 guests