[Jimi Zine]

Will the Qutest handle DSD native on a mac too? (as oppose to Mojo's DoP)

 

[x RELIC x]

Will the Qutest handle DSD native on a mac too? (as oppose to Mojo's DoP)

It won’t matter audibly. DSD over PCM (not converted to PCM) was created as a way to fool the USB interface to allow DSD transmition without drivers because the USB protocol doesn’t recognize DSD (it does recognize PCM). When a DAC sees a DoP signal it discards the PCM marker and plays the original DSD data contained unchanged from the original.

Basically DoP was created as a way to get around USB drivers, but coaxial and optical can use it as well.

Some reading:

https://www.northstar.it/dsd-native-vs-dop/

https://www.dcsltd.co.uk/support/what-is-dop-dsd-over-pcm/

 

[Pokemonn]

deleted

 

[jwbrent]

Very interesting. I was hoping for an updated replacement for the 2Qute to use in my speaker system. Can’t justify the cost of a DAVE, but I certainly can afford a Qutest. February is just around the corner ...

 

[jwbrent]

I would need a short micro USB to USB-B cable in order to connect my AK240SS as a transport, but they’re not easy to find without using an adapter. I wonder if Chord is packing in cables.

Also, if anyone knows a source for a high quality USB of the type I’m describing, I’m all ears.

 

[Rob Watts]

i see that Qutest measures slightly differently than Hugo2

Hugo2 vs Qutest (from Chord website)

• Channel separation: 135dB at 1kHz 300Ω vs 138dB at 1kHz 300ohms
• Signal to noise ratio: 126dB ‘A’ Weighted vs 124dB A-Weighted
• THD and noise at 3v RMS: 120dB at 1kHz 300ohms ‘A’ weighted (reference 5.3v) vs THD (2.5v RMS ref 3v): 117dB 300ohms A-Weighted
• THD: <0.0001% 1kHz 3v RMS 300Ω vs 0.0001% 1kHz 2.5v RMS 300ohms

Id be interested to have @Rob Watts comments.

The separation spec for Hugo 2 is worst case spec; my presentation gave it 144 dB, slightly better than Qutest. Having said that, separation spec is irrelevant - what is important is distortion from the other chanell - and both have none. Distortion on the separation test would indicate crosstalk via the PSU.

Signal to noise is exactly as expected. When I design a DAC, there is a detailed calculation of all sources of noise on the analogue section, assuming the DAC is disabled - the pulse array element resistors are set to DC - half to ground, half to the reference voltage. The vast majority of the noise is simply resistor thermal noise (Nyquist Johnson noise). Any design I do must have a dynamic range that is within 1 dB of this calculated value as this will indicate that the DAC part is not jitter sensitive, nor are there any sources of noise within the DAC (like switching sources or PSU noise) leaching in. Both Hugo 2 and Qutest are within 0.5 dB of theoretical. Qutest is lower DR simply because of the gain being different (3v rather than 5.3v OP). Actually getting this performance can be frustrating - Hugo 2 took 5 new PCB's - Qutest took 2 boards. Interestingly, I discovered something new when fixing Qutest (a layout issue) as one ch was 2dB worse than the other one, on the first prototype. Production units are now within 0.5 dB of each other.

When a DAC design goes through the prototype development treadmill a huge amount of measurements get done. I do not finish a design until some key metrics are passed - noise must be within 1dB of calculated; when the DAC is disabled it must be no better than 1dB (this proves no significant noise from the DAC itself); no measurable jitter artifacts using the source jitter test; no measurable noise floor modulation; no distortion via the other channel; and no in-band noise artifacts - the noise floor must be clean; no anharmonic distortion products (anharmonic distortion is distortion that is not a integer multiple of the fundamental - it is much more important than harmonic distortion). With all designs, there are often a lot of new prototypes that are needed to ensure all the above points pass.

Since the noise is 90% resistor value, why don't I simply half the values and get 3dB lower noise? Well problem is the actual impedance - there exists a tipping point when distortion will quadruple with halfing the resistor value. This distortion mechanism is understood but very complex, but basically with pulse array I choose the element resistor values so that I am just at the tipping point, so I get the lowest possible distortion - which can affect the sound quality (SQ). Fixed resistor noise is just hiss; it does not actually degrade the SQ.

THD and noise is consistent with noise being higher, and with distortion being slightly worse on Qutest due to DAC modulation index (2.5v against 3v max qutest, 3v against 5.3v Hugo 2 - Hugo 2 at 2.5v is 0.00007% THD)

 

[Rob Watts]

Here is my Hugo 2 presentation updated to include Qutest.

So that covers all the slides that are common to Hugo 2 and Qutest, So the slides covering Qutest only next:

I hope that is useful information.

Rob

 

[Clive101]

I would need a short micro USB to USB-B cable in order to connect my AK240SS as a transport, but they’re not easy to find without using an adapter. I wonder if Chord is packing in cables.

Also, if anyone knows a source for a high quality USB of the type I’m describing, I’m all ears.

Curious Cables.

http://www.curiouscables.com/buy.html

If you need any further info IMO best for price.

 

[dawktah2]

Here is my Hugo 2 presentation updated to include Qutest.

I hope that is useful information.

Rob

Yes, thank you! I am so appreciative of your input.

 

[Rob Watts]

I had a PM from @Music Alchemist asking about the PSU structure worked, so am posting my reply here.

One of the issues I wanted to improve upon, when designing Hugo 2 was isolating the amplifier (this is the single module that combines DAC amplification, filtering and headphone drive into one single feedback loop) from the battery. To do this, I created a booster voltage that was greater than the battery voltage, and then used low drop-out linear regulators to feed the amplifier gain section - the discrete output stage is connected to the battery, and it is signal dependent current drain from the OP stage that corrupts the battery voltage, which then feeds into the amplifier gain section - creating distortion. By separating the amplifier gain section from the noisy and distorted OP stage, you can eliminate this source of error. That is why Hugo 2 and Qutest both have very high channel separation, and more importantly no distortion from the other channel when it is playing into very low impedance loads. You can see the measured benefit of this from this slide:

Now this feature - of using a booster voltage, then seperate linear regulators to isolate the OP stage noise and distortion from the sensitive amplifier gain section - I have copied directly from Hugo 2 to Qutest. You may argue that this was unnecessary, as the distorted currents drawn on the power supply with Qutest is much lower - but this strategy worked so well I did not want to take the risk. And I know that distortion from the other chanel leaking through is a very important DAC issue - and having eliminated it with Hugo 2 I did not want the possibility of it degrading with a conventional PSU arrangement.

The other issue is with Hugo 2, adding the charger makes zero measured difference, and I could hear no change using HP in sound quality at all. So I replicated the design (it's actually complex with multiple RF filters and lots of separate regulators) from Hugo 2 into Qutest - but replacing Hugo 2's battery with a regulator.

Another point - like Hugo 2, Qutest has no coupling capacitors in the audio path at all. DC is kept to 100 uV by using my digital DC servo - the benefit of a digital servo is that I can remove all the distortion and noise from a conventional analogue servo digitally. The DC trim is then done digitally to the noise shaper inputs by applying a DC trim that has zero distortion and noise, as this has been filtered out digitally. The subjective benefit is of a very tight and deep bass - and measurement wise the distortion at 20 Hz is identical to 1 kHz - you normally see THD increasing with low frequencies hwn using coupling capacitors.

Rob

 

[vrln]

Hoping to see a matching headphone amplifier later on to go along with the Qutest. That would form a nice small desktop friendly setup and also close a gap in the lineup that otherwise exists between Hugo 2 (no galvanic USB isolation) and a potentially much more expensive Hugo 2 TT... Any chance of a comment on this?

 

[dmance]

@Rob Watts
So appreciate you being back on-line in a big way with incredible contributions to chord product and technology knowledge.
Regarding the Qutest chassis - which is described in the literature as being more heavy and robust to reduce vibrations and hence isolation of the circuit board.
Is this market-speak? ...or does this isolation benefit, in any way, the Qutest sound quality.
Thanks,
Dan

 

[Rob Watts]

If you want to drive HP, get a Hugo 2. Any external amp will degrade transparency - you just can't escape that. And my listening tests with HP was that galvanic isolation was not required as the USB input sounded the same as optical. You only need galvanic isolation when driving amplifiers that are mains powered....

 

[vrln]

If you want to drive HP, get a Hugo 2. Any external amp will degrade transparency - you just can't escape that. And my listening tests with HP was that galvanic isolation was not required as the USB input sounded the same as optical. You only need galvanic isolation when driving amplifiers that are mains powered....

Thanks for the reply! Sounds like a significant improvement over Mojo then. Just a little hesitant as I used a Mojo for desktop once and was a bit disappointed. Sound quality was noticeably worse via USB from my desktop computer compared to smartphones. Also had some hiss/clicks and pops on the Mojo (USB/computer) so I eventually sold it. Loved the sound from mobile sources though... Zero glitches or anything and a much cleaner sound.

 

[Rob Watts]

@Rob Watts
So appreciate you being back on-line in a big way with incredible contributions to chord product and technology knowledge.
Regarding the Qutest chassis - which is described in the literature as being more heavy and robust to reduce vibrations and hence isolation of the circuit board.
Is this market-speak? ...or does this isolation benefit, in any way, the Qutest sound quality.
Thanks,
Dan

Yes John did a great job on it - it really feels substantial. I would be surprised if it did not help SQ wise, but I haven't listened to the chassis yet - I have only just had the metalwork myself!