Sign me up for one as well!
I would prefer BNC, but XLR will work for me too.
Question: what would be the recommended “horsepower” for a windows-10 based PC, to operate the new native SW?
Excited to see more input voltage. This has been the biggest challenge for my product reviews. The car audio guys think you need a few hundred watts per channel for a tweeter! Hahaha…
I have a buyer for my QA401, anxious to see the QA402 come to life.
1 Like
It should be similar to the QA401. And the QA401 runs fine on Intel Celeron N4000 with 4G (about $120). These small machines come with Win10, and they include mounting hardware to attach to the rear of an LCD. If you get an LCD with touch (about $190) and hang a powered USB 3.0 hub from the PC it’s a pretty nice system that lets you run tests without a keyboard (assuming you are scanning serial numbers).
The link below has more info on fully isolated split rail generation. The QA402 uses SN6502B as the DCDC push-pull transformer driver and a TPS7A3901 for the LDOs.
This is REVJ HW, -1 dBV input from QA480 (super-low THD oscillator). QA402 max input set to +24 dBV, 2H and 3H are below -133 dBc and look to limited by FFT size (1M in this case)
This plot below is same setup, but at 192K. Out at 70 kHz you can see the ADC noise bump appear (the AK5397 used in the QA401 was better in this regard). The DAC won’t be functional at 192K–just 48K and 96K–due to timing issues across the isolation barrier.
Note the front-panel additions of the sample rate buttons. Those are no longer buried in the UI settings. Also to the right of the 48K sample button you can see a button labeled “Idle”. When this is active, the QA402 DAC will generate a continuous tone at the Gen1 settings whenever the analyzer isn’t running. This replaces the “generate fixed tone” menu. And this will respect your Gen1 units settings (dBV, dBu, etc)
Calibration data lives inside each QA402. So you can move the QA402 from setup to setup without needing to do another calibration.
The QA402 also moves away from an FPGA, and thus doesn’t require configuration. It’s ready to run within a second of being plugged in.
Finally, each of the plugins will allow you to specify units. Initially, just dbv and dbu, but more can come. The plugs-ins are being built into the exe and are no longer individual DLLs.
I don’t want to be an armchair quarterback here, so please just take this as a point of interest.
Have you looked at the Maxim isolators?
Wow thank you so much I was not expecting such detailed response.
What an elegantly simple solution!
This new device looks really great!
How about planned digital I2S port? Is it input/output or just an output?
Cannot wait to see QA402 released 
Will there be a possibility to write custom plug-ins for QA402?
Hi @ChUml, I2S in/out is probably the biggest risk for March. It might be left off (no front panel access). The hope right now is a very quick confirmation test in the next two weeks to verify the signals made it out as expected and then software that evolves over the coming months to allow you to configure I2S parameters such as bits per word, bits per frame, justification, etc. Initially only 48K will probably be supported. But higher rates are possible.
What the I2S will require is a breakout board. At some point, we’ll offer a breakout board that can be connected with a 10" or so ribbon cable to the front panel. And that could potentially have an array of interfaces on it too (SPDIF, etc)
In the attached PDF, you can see the test breakout board that will be used for initial verification. U41 is the connector from the QA402. This is where the ribbon cable will attach. Note on U41 there are data in/out lines, as well as power and I2C.
The power signal is 3.2V at 10 or 20 mA or so. This allows a pair of digital isolators to be powered (one for I2S, one for I2C). And on the other side of the isolator you connect whatever you want. In this case, you see a class D chipamp with a digital input in the PDF.
The expectation here is that a company that is testing a product they have created will invest in doing a board design to adapt the generic 3.2V logic signals to whatever they might need. But it could be a full-fledged ADC/DAC codec with I2C control.
In any case, I just want to set expectations that the I2S interface won’t give you the ability to set IO voltages–it’s fixed voltage and an adaptor board must be built if 3.2V won’t work. There probably won’t be a ton of flexibility on clock ranges initially. But if the I2S port proves useful there will definitely be some adaptor boards that could be done that allow things like SPDIF and TOSLINK. As Demian mentions, maybe a separate box at some point is where it all goes. But for now the focus will testing of digital in/out chip amps
QA402 I2S Breakout.pdf (49.7 KB)
1 Like
Not at first. That was an option for the QA401, and I know at least a few folks were interested in figuring it out. But I think something better will be available in the first release that might not make them needed.
Some background: The QA401 code base goes back to 2012 or so, and it’s pretty conventional C# Winforms. But around 2015, Microsoft release some async extensions to C# that were really innovative. And by 2016 developers were starting to figure out how powerful these could be. And so, a large goal with the new code base was to take advantage of the async benefits while moving towards the REST interface used in the QA401H.
A typical plugin for the QA401 looked very much like most event-driven code, which for things that have lots of delays (such as waiting for a relay to settle or waiting for a web page to load) gets a bit messy with callsback, timers, etc. But with async programming, much of this goes away and the state machines previously needed get built for you automatically. Take a look at the plugin code for the AmpThdVersusInOutLevel_QA480. This is the plugin that allows the ultra-low distortion QA480 to be stepped so that ADC THD can be measured.
Note the lines where ‘await’ is being used. These all run in the UI thread context, but the await allows the the UI thread to go off and run other UI things while the current operation is completing. This is much, much simpler than what was required before. Some might think “We’ll, why not just do all the other stuff in another thread?” and you can. But at some point you need to synchronize access between the two threads and that’s where it gets tricky. If your access is just a status update in the UI, then that’s not bad. But then add in the hardware, the USB, the plugins. It gets very messy without async programming.
So, back to your original question: Yes, the ability to load user-written DLLs (as in the 1.x software) will come and when they do they will be much simpler than before thanks to async programming. But I think you’ll be pleasantly surprised if you write a stand-alone REST app in the language of your choice to automate measurements.
1 Like
Mi Matt, it sounds great! The internal connector like U41 would work PERFECT for me! Every time I need I2S signals I can make my own adapter board compatible with your standard interface for my application needs. It’s even better having I2C available on this connector. Maybe I could design a simple I2S <-> SPDIF/Toslink adapter with SRC4382/92 and connecting it with I2C you could configure the chip from QuantAsylum SW directly. Will this internal connector be already present in the coming HW release?
Maybe having such an internal connector will be good having some optional connector on the back panel. Something very simple like blinded cutout for DB9 or DB15 connector, or maybe even better two DB9 connectors would be great for future expand-ability. The adapter board could have directly the DB9 connector on the board and could be simple mounter on the back panel and connected with flat cable with the analyzer main board.
Yes, I think the U41 mating connector will be present in the front panel for March. And yes, eventually the I2C messages that travel across U41 will be controlled as part of a test script in Tractor as part of the test. So, during test you could tell the chipamp to set volume attenuator to 20 dB via I2C (for example).
1 Like
Matt, is sounds great 
such a digital I/O in the front panel will be the best thing I could wish me for an audio analyzer. Do you mean I could set 0dBFs output signal and step build-in digital volume control in the amplifier (if it has I2C controlled volume) runing a THD+N sweep over the output amplitude? It would be the greatest feature I ever seen in audio analyzer!
Any guesses on when the QA402 might be for sale?
Also EXTREMELY interested in knowing availability and price of the 402. And, yes - me too, i would prefer BNC. I dont know how to book one, but Matt at least you said it would be in March…
Yes, March still looks solid. There will be a note here when ready and orders will probably be accepted a few days before shipping. There will be some holes in the first SW releases and missing features. But they will be steadily added in the following weeks and months.
1 Like
Another board pass is going to be sent out this week to correct an issue on the front-panel I2S. The I2S was originally a nice-to-have feature that could have potentially been cut to get a product out faster, but it’s working too well at this point and is too useful to cut. The front-panel of the software will have an I2S button, and when active the GEN1 settings will be mirrored to the front-panel I2S. There is a lot of possibility for format options that will grow over time: Programmable MCLK, BCLK, framing, etc. For testing board assemblies that aren’t fully assembled with electroacoustics this should be very useful.
As part of ensuring safety, the QA402 BNC inputs were plugged into the wall to verify design goals were met. The max input of the QA402 is +32 dBV, but the hardware can readily withstand 42 dBV = 126Vrms = 178Vpk safely on a low energy circuit. The input caps are rated for +/-50V, however, these ratings are at high temp and high ripple current and expect a 2000 hours lifetime. Because the input impedance of the analyzer is 100K, the ripple currents are a tiny fraction of the absolute max rating, in 10 minutes of operation at 126Vrms they didn’t get warm. Measured performance was the same afterwards. This was with atten on. With atten off, the current limiting input resistor fused open as expected (60W peak) and opened in a single cycle. It’s not recommended you subject the QA402 to these levels, but if you accidentally do, it will survive.
Isolation between audio side and USB side has also been verified at 1 KV, yielding more than 10Gohm.
2 Likes
So, potentially (har!), with a proper 12 or so dB external attenuator to drop the voltage, this could be used to measure the distortion and other junk riding on the AC mains up to the frequency limitations of the QA402? Lots of possibilities!
This is looking fantastic!
With the XLR inputs, how is the balanced input being handled?
Looking forward to purchasing but wanting to check that it isn’t simply shorting the negative connections like some passive adapters etc.