QA40xPlot Thread

Hello, i need to say thank you MarkZ for that nice Work you doing.
I have download the latest Version and i have a little question.

I use the QuantAsylum Devices most for Speaker design.

So i got also his MIC Preamp and want to use my Beyerdynamic MM1 Microphone with it.

On REW i can use my Compensation File on Preferences to get the Best Results.

Any ideas to implement an Compensation File input for youre Software???

May be also to Compensate an Chain with QA403 OUT → Any Amp or QA462 IN → QA462 OUT to DUT (Speaker)
Microphon MM1 (CompFile) → QA472 MIC IN → QA472 OUT → QA403 IN → QA40X Plot

Probably no need for that because of the good Linearity off the QA Devices but if you are using any aviable Amp for an big Subwoofer probably you need also to Compensate the Amp Frequencyresponse :wink:

Like on REW

This is how my File looks like:

;for beyerdynamic MM1 with serial number XXXX
;Frequency[Hz] rel. Level[dB]
60 0.5
70 0.4
80 0.4
90 0.4
100 0.4
110 0.4
120 0.4
130 0.4
140 0.3
150 0.3
160 0.3
170 0.3
180 0.3
190 0.3
200 0.3
210 0.3
220 0.3
230 0.3
240 0.2
250 0.2
260 0.2
270 0.2
280 0.2
290 0.2
300 0.2
310 0.2
320 0.1
330 0.1
340 0.1
350 0.1
360 0.1

and so on (TAB seperated)

Sure i can still use REW for that :wink:

Hi @Frunse

That’s a good point. I haven’t done much with actual mic testing yet because the chirp required to do speaker testing is kind of specialized (see REW) and is very long.

Compensation is definitely a requirement and I’ll see about moving it forward in queue.

@cfortner note this release contains settings for an external gain/attenuation block. PLEASE test this cautiously since it’s obviously dangerous.

On another note - New Installer

There is a new installer and a new installation method. It’s written up on the standard Github release page but summary->

  • Manually uninstall your current installation
  • Go to https://mzachmann.github.io and there is a web page with a link to the installer.
  • After this install QA40xPlot will check at startup for new versions and tell you accordingly (I hope).

Please let me know if this fails for any of you.

Thank you for the testing.

Mark

2 Likes

I did it and the Installation worked well!!!

Now on 1.1.15

Thank you :slight_smile:

1 Like

I’ll test it with low levels.

1 Like

How do I install this new version to computer that has no internet connection?

Wow, that’s a good question. At least for now, here’s what I think makes the most sense. I have not fully tested this but it should work.

  1. Go to Github and find my echo repository here mzachmann github io. Note this is not the website of the same-looking name.
  2. Download the QAPlot/Application Files folder by using Git or the Github-to-zip option.
  3. Inside of that folder is the full release. You can put that on a flash drive and copy it anywhere. Remove the .deploy suffixes. It will give the usual security warning when you start it (hopefully just once).

Mark

Postscript: I’ve updated the installer and avoided the .deploy suffixes. So, the process above is much simpler. The Application Files folder is the release.

1 Like

Mark, the installer updated itself to v1.1.16 properly. Thanks again.

2 Likes

Thanks for the info. Nice to know it’s not just me : )

I’ve just posted 1.1.17 and since it’s not clear what the changes are:

This change adds unit-ed voltage input. I’m not overjoyed with the current implementation, but it seems to work well. Click the “V” symbol after the voltage and a menu will pop up with the possible unit settings. Pick one and it will be used.

This setting is not saved for technical reasons. The data value is correct but saved files revert to “V” (or “W”).

If you switch DUT from power to voltage or voltage to power the value will not change. This is consistent with the prior releases and seems right to me.

@peppermint this is the dBu change you were talking about.

Mark

1 Like

That’s great. Thank you Mark!

1 Like

I’ve been checking my noise level more carefully since the noise post and it turns out I must be dyslexic and had Right/Left Noise Floor values swapped since May 1st. This fix will be in 1.1.18.

Sorry.

May be that’s the reason I get different results between your software and the official software. I mostly use only the left channel and nothing at the right..

Yes, that would do it… sorry about that. The good thing is that I did the primary refactor a month ago and now I’m just on final cleanup - since most of the math changed form. The best thing is I’m starting to feel confident in the Summary box information, which has been squirrely ever since I did the USB support.

Anyway, at this point - folks - I don’t have any pending requests other than

  1. I have one large TBD which is speaker response testing. The Response test does ok with it but not great.
  2. Still incomplete are the Summary boxes which I think are very good for Sine and not so great for others (and missing for Scope). That’s near-term to be fixed.

Release 1.1.18 goes a long way to getting noise floor and thd+n values that are correct (the errors were small - like 15% - but there). I still expect some bugs in the processing but decreasing rapidly.

Clearing averaged data - note that I don’t use a moving average for the math but instead you’re seeing the actual average of the last N sweeps so just wait N sweeps and the value will have been totally recalculated. There’s no leftover grunge.

Mark

v 1.1.18

  • fix left/right swap of noise floor
  • repair up-down control positioning on-screen
  • save current window size and position in configuration
  • enhance color picker transparency display
  • enhance plot color choices
  • add multitone (30 logarithmic spaced sines) waveform
    warning->much higher total power than a sine wave
  • improve auto attenuation math
  • fix enbw missing in noise floor and thd+n math
3 Likes

I’ve just posted 1.1.20 and … two comments:

  1. If you want to know changes, I keep the changelog up to date. To see the changelog, click the Help button and the help page at the top has a link to the changelog (which is local on your computer).

  2. Hi @Frunse - I’ve implemented a first pass at microphone compensation. It’s only available in the Response test and uses a file defined in Settings. Give it a shot when you get a chance and let me know if it works for you. I only had a Umik microphone compensation file to test with.

Mark

changelog----
v 1.1.20

  • add microphone compensation to response test and settings
    this uses linear frequency interpolation of the loaded compensation file
  • add data summary box to the scope test
    very preliminary, but working
  • line up gui fields more uniformly
  • fix (more) .using amplifier load. for Impedance test
  • let user set windowing in frequency response test
  • improve .get. display putting name on its own line
  • improve accuracy of output gain determination for frequency response test
1 Like

Tested it, but the Generator selection didn’t work, my ears brocke away even if i select 0.05 V ouput Voltage, it goes @ full Power!!! i need to stop that.

Also if i hit the 0.05 selection it shows me 5 V

All Values with decimal Points don’t work for me.

So cant do the Test my Amp didn’t have analogue Volumecontrol!!!

Didnt want to destroy my speaker.

Same here. If I chose 50mV = 0,05V, the output is a lot higher until I I chose another level and then 0,05V again.

1 Like

Ok, wow apparently the recent voltage conversion code (when I added the numeric input stuff) automatically reads the decimal place internationally but most of the formatting code doesn’t. This is honestly really strange and I’m going to have to repair this for those of you who don’t use decimal points native. This will take maybe a day to repair as I try to figure out wth is going on.

Mark

1 Like

Ok @cfortner and @Frunse … I think I’ve fixed it but it’s hard to test international stuff effectively. Especially when Win11 and Win10 are pretty different.

There is an auto-update. Please try this (carefully) but I think it should work well and actually will support commas and periods as decimal points (I think).

My apologies. I haven’t been testing internalization.

Mark

Again – thank you both so much for testing this so quickly. :slight_smile:

2 Likes

Thanks a lot! I’ll test it tomorrow with my self built valve amp, 6 KT120, 500W RMS.

Just kidding …

Really, thank you!

1 Like

I testet it right now!

Latest Version i connected the Left + Out with the left + IN so no worrys about speaker could blow up :wink:

Without compensation:

With Compensation:

But if i switch the resolution to more than 1 step per Oktave i get this error @ the End off that Measurment

I tried 2 Steps and it happen also

If i hit OK than nothing changes on my screen still the old Measurment would show

If i use chirp the same Error occure: