Hi,
I keep getting frequent dropouts (for lack of a better term) and from what I can tell it’s not a voltage issue. Any ideas? It’s pretty annoying so I’d like to find a solution.
The rate is not as high as it’s shown below but it does occur every minute or so.
Thanks
w.
![vlc record 2025 02 18 12h30m52s QuantAsylum QA40x Audio Analyzer 1.202 [My Default.settings]](https://canada1.discourse-cdn.com/flex032/uploads/quantasylum/original/2X/4/4ff15891a309eecfe19aa5022f72f0050d436087.gif)
Hi @peppermint, do these happen at 48k? Also, on a PC, most all USB ports are related to an upstream hub inside the computer, and often use different USB drivers. If you plug into different USB port do you get different results at 48k and 192?
Hi Matt,
Changing the usb port seems to fix the problem but I need to spend some more time to be sure. Thank you for the suggestion! For some reason it didn’t cross my mind that something as simple as changing the port could be the issue.
On a different note, I noticed that changing from 192k to 48k, slows down the update rate considerably. Most likely I don’t understand the theory behind it, but is it normal to have slower update rate with smaller SR? I would imagine that it should be the other way round…
Hi @peppermint, Windows is a bit strange that way, but it does make sense. Some ports are high-speed, some are USB3, some can still be low speed, etc. And behind each of those is PC motherboard hardware. And so, the user it looks like USB ports, but to Windows each port is unique.
A good experiment to do is set to 48k sample rate, set 1k fft size, start running and then go have lunch. In the lower left you’ll see a count of acquisitions. This should run for a very, very long time without getting stuck. You can try this on different ports and see if there is different behavior.
Note the smaller the FFT the faster updates will happen. The update period is roughly
So, a 1024 FFT size at 48k will about 21 mS to complete. Your observation is correction: When you reduce the sample rate and keep the FFT size the same, the acquisition will take longer.
I also have the dropout issue. I have tried USB ports on the motherboard and on the front of the case. I get dropouts quite often and get messy looking graphs because of it. This is a recent happening. I was using version 1.202 but getting the same issue with 1.213.
@DaveD It may not be the USB ports, but the cables themselves. Many cables simply afford too much voltage drop when the current is around 800-900mA.
Bear in mind the current draw of the QA403 is not constant. You can monitor the voltage and current and when relays are operated or the gen is supplying a decent output, you’ll notice the voltage drop from cable resistance.
I went through my pile of USB cables and can suggest you use a USB C high current data cable with an adaptor from C to B (or A).
Hi @DaveD, this is at 48k?
I have used 3 different cables in 3 different USB ports on my computer and all have the same issue with dropouts. My main cable is a high current USB-A to USB-B with 20 AWG and 28 AWG wires. I get over 4.8 volts with this cable plugging into the back of my computer. My other cables are just standard USB-A to USB-B cables with some cables giving only 4.3 volts at the front panel USB port. I am running Win 11 on my desktop.
On my laptop running Win 7, I don’t have these dropout issues.
Back on my desktop, I ran a bunch of simulations. I was using the THD vs Power Output test, with 27 individual tests per run. I set the “distortion measurement stop” to half of sampling rate or 80 kHz for 192k sampling rate.
With an FFT of 64k, 128k or 256K, I had multiple errors (per run of 27 tests) with a 192k sampling rate.
With an FFT of 64k, 128k or 256K, I had 0-1 errors (per run of 27 tests) with a 96k sampling rate.
With an FFT of 64k, 128k or 256K, I had 0 errors (per run of 27 tests) with a 48k sampling rate.
The dropout happens with other tests as well. Here is a quick THD+N vs Freq test using 96k sampling rate and 128k FFT.
Here is a subsequent test using 48k sampling rate and 128k FFT. Only 1 dropout.
The next test I did was 192k sampling rate and 128k FFT. Lots of dropouts.
Last test I did was 192k sampling rate and 256k FFT. Lots of dropouts also.
Hi @DaveD, thanks for the update! I think the focus should be getting 48k pristine, and then moving on to other sample rates. The USB code in the releases hasn’t been touched in a long time, so I don’t think you’ll see a difference bouncing between the various recent releases.
First, can you verify that the USB voltage reported by the QA40x app isn’t showing orange or red? This can be done with a 20 awg cable as you’ve done and if that isn’t solving the issue completely, then try using a powered USB hub.
Second, if you are still having issues after that, I’d see what it takes to update the BIOS and motherboard USB drivers. I know that’s a pain, but something in your current desktop is probably responding slowly to USB packets arriving and that is causing the apparent loss of data.
Next, re-start the app, do a File->New Settings, set the FFT size to minimum, enabled GEN1 and just let it run in loopback while you do other things on your machine. It should run for 10’s of thousands of acquisitions (which will take maybe 30 minutes) without any problem with a consistent spectrum. If packets are lost, the spectrum will turn to garbage and often won’t recover right away so it’s easy to see.
If the issues at 48k can’t be solved, then the issues at higher sample rates won’t be solved either. And so, the 48k is a good place to start. I know you did 27 tests without errors, but the aim should be to do 10’s of thousands of acquisitions without an issue. Remember as shown above the counter in the lower left to help you keep track.
Please report back what you find when time permits
I updated my BIOS. It was from 2020. If it aint broke, don’t fix it was my motto. Well, I guess it was broken because now I don’t have anymore dropout issues. I ran a bunch of distortion vs power out, distortion vs frequency and SMPTE tests. Not one dropped value. I just did it at 256k FFT and 192 kHz sampling. I used my regular USB cable, 28/20 AWG, and was getting 4.80 volts during the tests.
Thanks for the help. Hope this helps other users as well.
Hi @DaveD, thanks for the report back. Yeah, I’m with you on BIOS updates. I don’t like doing them either unless required. With TPM, I’m always worried the hard drive key might get forgotten rendering the drive dead. I know there’s a separate TPM module that holds the key, but still…
Glad you are back in business!
