SneakPeek: The QA403

Discussion or questions from the blog post linked below

Can we order these 3 devices now?

Hello,
The ES9822PRO for the ADC, and the ES9038Q2M for the DAC sound very good.
I have some questions, if i may.
Do you have some preliminary measurements?
Do you plan to use distortions compensation registers for H2 and H3?
If so, the compensation will be automatically via QA40x software?
Thank you.
Best regards!

Yes!!!

Awesome news!

I am using the ES9822 and the distortion setting register affect 2nd and 3rd rather dramatically, so it would be good that they are factory trimmed to the DAC so the whole system is lowest THD possible.

USB isolation can be made with external devices if needed but of course it would be most practical to have it inside the 403.

/Sigurd

I think that you want the DAC and ADC to be trimmed separately. Otherwise, you’d possibly get cancelling predistortion effects that would give inaccurate DUT measurements. Stupid example: If the 2nd harmonic from the DAC at 1 KHz is -100 dBc, you could potentially use the registers in the ADC to compensate and end up with -150 dBc as the combined result. Great, right? But, the DAC output is still -100 dBc and a -140 dBc DUT might show as -100 dBc because the cancellation was disturbed. Perhaps there is a clever software solution to this; I dunno.

From what I can tell, it may be more important to have temperature controlled ADCs and DACs because the trimming is somewhat temperature sensitive.

I own an E1DA Cosmos ADC and the distortion does change a bit if the room is hotter or colder than normal. But, it’s still so good, I’m not really concerned.

:slight_smile: I saw the future … ES9822PRO maybe the best candidate for the future qa403?

Put me on the buyer(s?) list please.

Rgds,

Pierre

Hi Matt
Thank you for persisting with trying to bring a high performance device to market given the global supply problem.

Please, please, please, bring a 384khz device to market.

For someone like me who owns a QA-401 a QA-402 or 192khz QA-403 is tempting but not a huge step up. So I’d have to wonder if it’s worth spending the money.

But if we are talking a 384khz QA-403 that’s something that really makes me sit up and know that I’m happy to part with the cash.

Please keep us posted.

Kind regards

Stuart

Hi Matt

I’m sorry to hear that the electronic components supply issues are hurting you so much that yet another redesign of the QA40x is needed. I’m working with communication and gps module manufacturers and they have the same issues. I’m afraid we have to live with this situation for at least one more year.

Concerning the new QA403, I have a feature request.

  • On the QA402 there is an analog 80kHz LPF (anti aliasing filter) and the frequency response at the low end varies with the sampling rate. Is there any chance to implement a LPF with higher or switchable corner frequency on the QA403? I assume the higher sampling rate of the ADC will anyway allow an anti aliasing filter with higher cutoff frequency.
  • I think the corner frequency at the low end of the QA402 is given by a digital filter in the ADC (or the DAC?). A lower, fs independent, corner frequency would be nice.

I hope that the QA402 will still be supported with software updates for a while.

I’m curious to hear the latest news on new features and specs of the QA403. Depending on improvements, I even consider to upgrate to the new device.

Thank you for developing and bringing to the market these really good and afordable audio measurement tools.

Best regards and good luck
Andy

Hi @Avo, yes, the QA402 and QA403 share the same software, so they will evolve together.

The freq response of the QA403 in loopback at the 48K sample rate is shown below. In both the DAC and ADC the low-ripple filter options of the ESS parts are used (+/- 0.002 dB). There is a bit of peaking in the digital filters in the ADC and DAC that is to be expected at the corner. This scales with sample rate in the silicon, but I think at 192K sample rate the board itself (due to the R’s and C’s) will be ~1 dB down at 90K.

The silicon also has a brickwall option, but it has about 10X more passband ripple.

Hi @Stuartmp74, the 384K will be a stretch goal, with the challenge being the isolation barrier. Without the iso barrier it’s not a challenge.

Hi @BKDad, you are right they will probably need to be done separately. The ESS parts have a lot of options for tweaking, and there will need to be a way to expose the options that are useful.

@bmsorin

Here’s a plot of ADC THDN (QA480 into ADC, single ended, no weighting, 20K BW, no notch). I think another 3-4 dB of improvement will coming. The ADC is spec’d for -114 dB (worst case) and -117 dB (nominal) at -1 dBFS. Noise is the challenge. I hope the yields show THD+N better than -110 dB for all inputs from -4 dBV to 18 dBV.

Hi @Moto and @PFB , we will figure something out. I don’t like the idea like kickstarter where you pay everything up front and wait. But just names on a list don’t work very well either as it’s too hard to manage. Maybe a $X non-refundable reservation which gives $2X off product price when ready, where X is $10 or so. That way the list is clear and automated and it’s easy to manage and we can be sure we don’t overbook. We are still studying what is manageable on Shopify.

@matt personally I’m not concerned with deals. Happy to pay full price in advance. Just let us know.

I would gladly skip the isolation and instead get an Fs=384kHz. Isolation is easy to buy a device for but to get 384kHz is not that easy. The COSMOS ADC has 384kHz using the ES9822 chip, but it is not a complete system with DA/AD/attenuator+SW…