Hi @ChUmI, yes, the ADC is usually the limiting factor on the QA402. Remember, the goal with the QA402 was a replacement for the QA401 with extended ranges of operation. The QA401 has +6 dBV max out, the QA402 extends that to +18 dBV. The QA401 had a max of +26 dBV in, the QA402 extends that to +32 dBV. The QA401 delivered optimum THD and THD+N at a single point, the QA402 delivers that across a much wider range of input/output, etc.
The RME ADI-2 PRO is indeed very nice…but it’s $2000, isn’t isolated, and runs from a 24W power supply. The QA402 is $499 and runs from a 4W USB connection with isolation. There’s not really any magic needed to build a high performance converter: You spend more money and consume more power and you get better performance
BTW, I think we’ll be ready to accept orders beginning on April 26 and start shipping on May 3rd to North America. I’ll share some more details as we get closer on what the initial limitations will be. For example, in the first release there won’t be a way to change front-panel I2S bit depth, sample rate, etc. I’ll make sure that as we get closer to April 26 these are all clear and when it can be expected they will be addressed.
@matt yes, RME is a different price tag, Nevertheless you cannot do any audio measurement straight out of the box with RME and must be very careful not to damage it measuring any power amplifier Your QA402 looks like the absolute bang for the buck With QA402 at its price tag, performance, all features like digital audio I2S and I2C port and the SW you’re giving the other analyzer manufacturers hard time
I, although I have two other audio analyzers and my RME, will definitely buy the QA402 as well. Maybe not at first, so that those who need it more will get it, but in the near future I will also want to have QA402 on my bench
Hi @matt I hope everything is going well with QA402 release I have some more questions - can I do a digital sweep on I2S, simply I2S out → in, without any additional settings? How would look such a sweep like - will I see simple the dynamic of a bit number like -144dB THD+N @ 0dBFs and decreasing linearly with decreasing digital signal amplitude? Do you have such a sweep graph example
What is a standard FFT setting for the THD+N measurement? Is there any FFT averaging like x4/8/16/32… to get smooth results? How many steps can I put for single sweep? Sorry for so much questions but would love to learn more about QA402 to know how good it fits to my measurements
Hi @ChUmI, yes, I think ultimately what you outline should work as you describe. To date the testing has been 16-bit output only because the target is bluetooth and lower-fidelity class D digital-input amps.
But some 32-bit DAC boards have arrived (ES9038Q2M) that will allow 32-bit testing of the front-panel I2S. My hope/guess is that the 3.072 MHz bitclock won’t be an issue (32bits per channel, 2 channel, 48K sample rate), but there will certainly be an an issue at some point (eg 8 32-bit channels at 192K = 49 MHz bclk probably won’t work).
But where, precisely, everything breaks down on the front-panel expansion port isn’t known. Is it signal integrity? Is it the ability to keep the USB fed? For now, I think the I2S will be spec’d as output only, 48K stereo 24-bit. And then as more testing completes it will be expanded from there.
Hi @Matt, all look great I think the 48kHz 24/32bit is the best option for analyzer extension thinking of best audio performance. Most of AD/DA converters show the best performance at 48kHz. But I definitely think we need input and output here. Having input and output will open brand new possibilities for this device. I guess data input and output in my opinion is much more important than a high-res output like 96/192kHz…
Hey Matt, is there anyway we could disable the software limit of 36dbv or allow 42dbv? Provided we change the capacitors to 100v.
Also, is it possible to measure peak power for something like 20ms, 100ms or 200ms burst? I’ve got some old Proton Amps that have high peak power that I would like to test.
Hi @Tonywood, there is a command line option in the first release (should be posted next tuesday) where you specify
qa40x.exe --nolimits
on the command line and that will greet you with a one-time dialog at launch that says:
And from that point on, you’ll get a small nag on the main screen and the software limits will be removed for careful testing up to +42 dBV. BUT you will need to take suitable precautions. The attenuator is locked in the on position in this mode, and you’ll need to verify you are not connecting to a high-energy circuit (QA402 is only CAT I rated).
The input electrolytic are bipolar 4.7u/50V and rated for 2000 hours at 85C and 31 mA of ripple. In safety testing at 120Vrms the input caps don’t get hot at all. The currents in the caps with 100K input impedance is 1mA. Absent the heat, the failure mechanism will probably degraded capacitance over time, which could be detected by looking at frequency response in loopback. And after some time in the “nolimits” test mode, they will need to be replaced. But I didn’t detect any change in capacitance in testing.
You could measure peak power in a small burst by picking a very small FFT. The smallest FFT is 1K, which at 48K sample rate will yield a 21 mS burst (1024/48K). It would be very nice to have a plug-in that graphed power over per-cycle so you could see power limiting with millisecond resolution. A lot of prosound cabinets have a DSP inside that is handling everything. That said, I’ve tested a lot of commercial amps that advertise very high peak transient power, but they must be considering very, very short peaks. If anyone is aware of standards for amps and musical transients please share.
Short-term output is the subject of the EIA dynamic-headroom specification, which is based on a 20-millisecond tone burst repeated twice per second. I believe it’s at 1,000hz.
I am sure I don’t understand the motivation for I2S. My understanding is that I2S was never intended as a box-to-box interconnect. Reference: ASR I2S Comparison So I wonder how you see this actually being used?
The “Max Input Level dBV” first row of “0, 6, 12, 18” don’t have “atten active”? The text says “…addition of +18dBV without atten active.” So how are the input levels of 6, 12 and 18dbV achieved without an attenuator? Like I said, probably a dumb question.
Hi @proj964, no, not a dumb question. The “atten active” is for the big passive attenuator at the input. That attenuator is before any active electronics. There are other gain/atten stages, but they are active. The naked ADC wants to see 0 dBV full scale. With 6 dB of active attenuation, then your max input increases to 6 dBV. And with 12 dB of active attenuation your max input increases to 12 dBV, etc.
Orders will start shipping today to the US. For orders from outside the US, it will be June. If you have ordered and your financial situation changes before the order ships, send a mail to sales and we can cancel no problem.
There will be a lot of software updates in the coming weeks/months to get the QA402 software to parity with the QA401 and to fix the inevitable bugs. Note the QA402 is a re-write from the ground up–work that had to be done to deal with the wide range of hardware input and output levels. It seems like a trivial problem, but the QA401 software held the data internally in dBFS, and would then map to dBV as needed. The new software holds all the data internally in dBV. The old plugins were DLLs, the new plugins are part of the EXE. The processor used in the QA402 is an NXP LPC55S16. I cannot say enough good things about the toolchain and the silicon if you are looking at a starting point for a new audio project.
In short, just as we were able to get 5 years of active platform evolution from the QA401, the new platform on which the QA402 is built has the legs to do the same.
Some major omissions in the first releases are on the QA402 product page, but I’ll repeat them here:
Cross platform will take some time to get working reliably. It runs on Linux now, but it crashes easily.
Tractor support is a few months out. But the new API should make standalone test apps very easy
PC mirroring isn’t part of the build yet
Open loop testing isn’t yet enabled
Many plugins are missing
The product sheet on the QA402 with specs and measurements is located here.
Your QA402 looks like the absolute bang for the buck 
With QA402 at its price tag, performance, all features like digital audio I2S and I2C port and the SW you’re giving the other analyzer manufacturers hard time 
I have some more questions - can I do a digital sweep on I2S, simply I2S out → in, without any additional settings? How would look such a sweep like - will I see simple the dynamic of a bit number like -144dB THD+N @ 0dBFs and decreasing linearly with decreasing digital signal amplitude? Do you have such a sweep graph example 
