I appreciate the feedback. The software is starting to be pretty usable for me and I’m going to cut back to a release a week and I’m going to be checking this less than daily – just so you all know.
I have found a bug or few and they’re being fixed. The worst is that it’s using the wrong reference resistance for impedance tests - the global amplifier load rather than the value shown in the tab field. The next version fixes that since they aren’t the same thing at all.
Yes, I know the averaging isn’t right. It’s in my moderately short TBD list as I noted in an earlier post - it doesn’t even work at all on REST. It just wasn’t in my personal must-haves. It should be soon (maybe next week) now that all the refactoring is done. I’ll bump the max too. Almost certainly the REST version will be slow.
re 384K support. I’ve tried it and it works fine on the ADC but it’s not the same as other code which has a single FFT value so I’ve got to cobble in something to use the dac at 192 if that’s even possible. It’s in queue but probably not next release/week.
I also do have external generator as a thought but not yet in my TBD list.
Power measurements… that’s a valid question. I have in my queue adding offsets to the graph plots but that’s way different than integrating it into the sweep UI. How do you feel about a global setting? At this point I’m really reluctant to add more tab-local settings until I add collapsers.
I honestly haven’t decided about antialising. I may do it but
a) I think it’s not a good feature
b) I hate to ‘offer’ features that I don’t think people should use
c) it will require additional effort from me since ui changes need to be double-tested
Mark,
Serious THANK YOU for all you have done on this. Beyond having a stable and usable product, the ‘wish list’ the community could keep throwing at you could probably consume you forever. I think finding a balance of personal life and how much you give to a “free” product is more than warranted. Again, thanks for what you have done to get it where it is.
Looks like there is something wrong with signal generator1, there’s no output if I select voltage lower than 1V. This happens in spectrum and scope views.
In the Intermod view same problem with gen1, but generator2 kinda works. But looks like its level is not correctly lowered. There is just one step where (measured) output drops -40dB when I select any of the voltages below 1V.
Hey Mark, I´m using a QA402 and it seams, that Left an Right Inputs are mixed up compared to the original software.
Also I´m wondering if it would be possible to switch units for the geneator? I´m always working in thinking in dBu.
Thanks for your work!
@TimoJ
I default to 0.1 volt for everything and it’s working ok for me. I haven’t looked at internationalizing yet - are you using commas for decimals? Is your DUT setting “Input Voltage”?
@cfortner
Yeah, I agree completely. It’s why I don’t like global settings much since they’re invisible. I’m going to have to free up some screen real estate so I’d like to do this in two passes. First, invisible quickly (yuck) then visible when I figure out how best to add some space. I can easily add it to the progress snipe on top during runs, though that’s kind of cheesy.
@passmedian I don’t have a QA402 so you can help answer a couple of questions. Are you using USB or the REST interface? Does it happen on both or just one of those? Can you please test a couple of the tabs to ensure it’s consistently wrong? Thanks.
@TimoJ Re:antialising. I don’t want to come off as a pain here but if you don’t like the ClearType rendition I strongly recommend you go to Settings/ClearType and try to optimize it for your display or just disable it there.
One last thing… someone asked about colors - yes, I did change the right color so it’s no longer translucent on overlay. I’ll be smarter soon but as part of the refactoring it uses the same method (fully opaque) as the left channel in all cases for now.
-Mark
@cfortner how about if I display it in the mouseover tooltip for the Set DUT option?
Yes, I’m using commas for decimals. I’ll try with dots. Official QA40x software also has comma issues in some places.
About antialising: of course I have disabled Cleartype but QA40xplot uses its internal settings and overrides that. But lets forget this, I’ll manage with antialising or modify source code and disable it.
Yeah, sorry about that. I haven’t gotten there yet - so period (.) = decimal place for now. I don’t override antialising settings anywhere afaik…
I just posted the 1.1.12 release with frequency averaging which you’ll like a lot better than time averaging (though Scope still does that). I’ve tested it but it may have bugs as usual.
Also done: added 8k,16k fft sizes, increased number of averages to 20, added an option to set graph background color.
Finally, 1.1.12 has slightly different plot colors/translucency since somehow that last edit didn’t go into 1.1.11.
and… @passmedien - units for the voltage entry values is in queue but not for a bit since it’s so risky.
Regarding the channels being reversed on a qa402, I am seeing the same thing. I did the test to compare to see if it is happening with both rest and USB. It only happens when running with usb. Also checked scope and response tabs, same thing. I’m also happy with the averaging change. Thanks again for all your work
@MarkZ Would you please consider making the noise measurements in uV accurate? That would obviously need to take the measurement bandwidth into account and any weighting (if any). It seems to me the number displayed is a narrow ‘spot’ measurement, not an integrated RMS across the specified* spectrum.
we have no ability to specify a bandwidth so what is the default measurement bandwidth you are using?
We know the shorted input 20Hz-20kHz UNWTD noise floor of the QA is around 1.2uV.
I think what’s happening is that you have averaging set to 5, which is going to lower the value at 1KHz substantially and the upper right value is the fundamental signal strength (for a THD measurement), not the noise value.
Note that the noise is shown as -115dBV which is about 1.7uV. So, the math may be incorrect but it looks in the ballpark. Let me know if not.
Also, note that if you turn off the “% - Data Summary” it will show noise in uV in the popup.
Also… @BarryLambert I’ve just posted 1.1.13 which I think fixes the QA402 channel swap. Let me know if not.
-Mark
oh, and I do 20-20Khz unweighted. I want to add weighting but needed the 1.1.13 GUI upgrades to have room on-screen.
For everyone watching I found a very bad issue in that the right channel data in 1.1.13 is incorrect due to the averaging upgrade (typed Left when I meant Right once) and so I’ve updated the Github release to 1.1.14 in the 1.1.13 release area. Just go back and get the latest release from there. The next release will be 1.1.15.
My bad (though it was in 1.1.12 also I think…).
Also @TimoJ as far as I can tell the arrows point the right way. They’re triangles and came with the sample code I had to rewrite. I’ll look for real arrows or something more obvious.
No need to change arrow graphics. This is what I see and here left arrow points up and avg goes up. But shouldn’t left arrow point down and values should go down? i.e. arrow positions are reversed.
QA 402 on version.14, USB channels now are good. Under REST now getting the error message, in the Spectrum tab, response status code does not indicate success: 400 ( bad request: exceprion: ( object reference not set to an instance of an object )). Scope and response tabs work fine under REST .
Hi @BarryLambert I can’t replicate your issue with the REST interface with my setup. Can you please run the Scope Test first and then the Spectrum test and see if it still fails that way? If so please include a small photo of the Spectrum Test screen so I can see your settings.
Mark, I did some experimentation this morning and found that I had a CFG file from v.12. Or maybe even earlier. Deleted it and now everything works fine under both REST and USB.
