Chord DAVE Review (DAC & HP Amp)

[majingotan]

It's odd to see a DAC there in the output, since I thought the claim was there were none.

Chord prides itself on its filters, so those blocks may warrant patenting if they think they are doing something unique that deserves protecting. I am no expert in this field, that's just looking at the blocks in the design.

Now you are referring to a post (#633) that clearly shows a dac chip. So, which is it: they wrote their own code for the dac process or they used an existing dac?

Watts original DAC design was pretty much a Pulse Width Modulation approach using microcontrollers, and the FPGA is there to optimize timing transients (not really optimized as you can see with Hugo2's jitter performance). When those microcontrollers are inside a chip then yes it's a DAC chip, but in this case it's a discrete implementation

 

[DonR]

It was my understanding that the FPGA fed a series of flip-flops that are part of a discrete R2R ladder. Sounds positively medieval.

 

[pablolie]

Ha, I see yes, there are "chips" used in the design. What I refer to a DAC IC is a IC that perform all the functions of the conversion from analog to digital. There are in this designs some single functions circuitry yes. Which box is the DAC chip exactly in post 633?

There is no standard definition of what a discrete DAC is, as far as I know. I *guess* that instead of buying an off the shelf DAC from the usual people, you build it out of standard "discrete" components, ie transistors and a filter circuit. I assume Chord does that because they do a lot of the "D" in DAC with their filters in the FPGA. But I am just guessing. Basically, if you think you do a better "D" in your FPGA, why pay for that part from a top end off the shelf DAC? There are also FPGAs with analog capabilities, so that block *may* be integrated in the FPGA. The blocks don't necessarily mean those are separate physical components, they may be logical blocks.

 

[PeteL]

It was my understanding that the FPGA fed a series of flip-flops that are part of a discrete R2R ladder. Sounds positively medieval.

It's Delta Sigma.

 

[Dogcoop]

Watts original DAC design was pretty much a Pulse Width Modulation approach using microcontrollers, and the FPGA is there to optimize timing transients (not really optimized as you can see with Hugo2's jitter performance). When those microcontrollers are inside a chip then yes it's a DAC chip, but in this case it's a discrete implementation

Thanks for the explanation. My question is whether RW actually designed and wrote the whole dac implementation or if he concentrated on the optimization of the timing transients and used already produced product to handle some of the dac processing?

 

[DonR]

It's Delta Sigma.

ah... so less medieval than I thought.

 

[PeteL]

There is no standard definition of what a discrete DAC is, as far as I know. I *guess* that instead of buying an off the shelf DAC from the usual people, you build it out of standard components, ie transistors and a filter circuit.

OK, I didn't use the word discrete, it is indeed a vague concept.

 

[DonR]

OK, I didn't use the word discrete, it is indeed a vague concept.

If you watch their pulse DAC video, it looks to be a solution put together from discrete components. Like an op-amp built from transistors.

The DAC resides in the circuitry between the FPGA and the output cap to the left in the thumbnail.

 

[mansr]

My understanding is that the FPGA does interpolation (upsampling), sigma-delta modulation, and some kind of dynamic element matching (what they call "pulse array"). The resulting (multi-bit) signal is output, likely via a set of synchronised latches, to the actual D/A conversion stage, the "DAC" box in the figure. This probably consists of as many (identical) current sources as there are bits in the DEM output which are summed to produce the analogue signal. Companies like AKM, Cirrus, and TI have the resources and sales volumes to put all of that on a single cheap chip. Chord does not, so they're stuck with an expensive FPGA.

 

[PeteL]

My understanding is that the FPGA does interpolation (upsampling), sigma-delta modulation, and some kind of dynamic element matching (what they call "pulse array"). The resulting (multi-bit) signal is output, likely via a set of synchronised latches, to the actual D/A conversion stage, the "DAC" box in the figure. This probably consists of as many (identical) current sources as there are bits in the DEM output which are summed to produce the analogue signal. Companies like AKM, Cirrus, and TI have the resources and sales volumes to put all of that on a single cheap chip. Chord does not, so they're stuck with an expensive FPGA.

No doubt that AKM, Cirrus, TI and ESS have development budgets and sales volume many many orders of magnitude higher than Chord.

 

[Dogcoop]

No doubt that AKM, Cirrus, TI and ESS have development budgets and sales volume many many orders of magnitude higher than Chord.

I am curious. What exactly is your purpose for being on this thread? Are you making a case for RW’s technology being superior to all other dacs? Are you looking for clarity on a specific technical aspect of RW’s design? I’m just not sure what point you are trying to make.

 

[PeteL]

I am curious. What exactly is your purpose for being on this thread? Are you making a case for RW’s technology being superior to all other dacs? Are you looking for clarity on a specific technical aspect of RW’s design? I’m just not sure what point you are trying to make.

I have already been trough that. But I can repeat. No I don’t think this technology is superior than any other DAC. But I am genuinely impressed by the amount of engineering put in this thing coming from an audio manufacturer. This is quite uncommon, especially back 7 years ago. The main thing I was getting from many of these thread was tha Mr. Rob Watts was nothing more than a Charrlatan that didn’t know what he was doing and I feel that people don’t realise the amount of engineering talent it takes to do something like that. The fact that nowadays a multi billion company like AKM or else can put similar tech in a chip that perform even better does not surprise me, and I don’t disagree that you can make an even measurably better DAC using those. That’s not What I am debating, but I am an engineer myself as I said. Putting a XMOS USB receiver and A ES9038 or else in a way that is functioning is easy, I myself can do it and actually done it. Making it optimal in term of meaasurable metrics, audible or not, is harder, That´s what companies like topping or others are great at. But something like that it’s not even close to what I could put together with a lifetime of studying audio. Yes there are just as talented people that may work at AkM that deserve even more credit, But collectively, these « type » of developpers push the tech forward, and allow for us to have affordable high performance product. Not necessarily Watt personally, but as part of the pool of select people that know enough to. be able to design this, generally speaking.And They may not do the best DACs or may not be for long, but collectively, they bring the tech forward. my point is, wheter it’s worth it nowadays, wether it « competes » to this day, is totally fine when it comes to asses the « value » of a product, But still credit is due when there is actual real hedscratching, calculation making algorythm creating… Well actual engineering. That’s all, we can disagree on the price we can disagree on the marketing methods, all that is fine by me. But it’s when it’s put forward that this guy don’t know what he is doing and that it’s just a woo pusher, than it don’t add up when you have an understanding of what it takes to make something like that.

 

[Dogcoop]

I have already been trough that. But I can repeat. No I don’t think this technology is superior than any other DAC. But I am genuinely impressed by the amount of engineering put in this thing coming from an audio manufacturer. This is quite uncommon, especially back 7 years ago. The main thing I was getting from many of these thread was tha Mr. Rob Watts was nothing more than a Charrlatan that didn’t know what he was doing and I feel that people don’t realise the amount of engineering talent it takes to do something like that. The fact that nowadays a multi billion company like AKM or else can put similar tech in a chip that perform even better does not surprise me, and I don’t disagree that you can make an even measurably better DAC using those. That’s not What I am debating, but I am an engineer myself as I said. Putting a XMOS USB receiver and A ES9038 or else in a way that is functioning is easy, I myself can do it and actually done it. Making it optimal in term of meaasurable metrics, audible or not, is harder, That´s what companies like topping or others are great at. But something like that it’s not even close to what I could put together with a lifetime of studying audio. Yes there are just as talented people that may work at AkM that deserve even more credit, But collectively, these « type » of developpers push the tech forward, and allow for us to have affordable high performance product. Not necessarily Watt personally, but as part of the pool of select people that know enough to. be able to design this, generally speaking.And They may not do the best DACs or may not be for long, but collectively, they bring the tech forward. my point is, wheter it’s worth it nowadays, wether it « competes » to this day, is totally fine when it comes to asses the « value » of a product, But still credit is due when there is actual real hedscratching, calculation making algorythm creating… Well actual engineering. That’s all, we can disagree on the price we can disagree on the marketing methods, all that is fine by me. But it’s when it’s put forward that this guy don’t know what he is doing and that it’s just a woo pusher, than it don’t add up when you have an understanding of what it takes to make something like that.

Thank you for taking the time to explain your point of view. I really do appreciate that. Having owned a DAVE and M-Scaler, I have a different opinion as to the value of RW’s contribution to DAC technology, but I respect your view and wish you happy listening!

 

[JSmith]

If he hadn't mentioned it I would never have thought there was a problem; I think it's inaudible.

Hopefully so... but still, I would imagine you would have preferred not to have seen that in testing.

I have a thick skin from being a lawyer.

Touché.

But it’s when it’s put forward that this guy don’t know what he is doing

RW knows exactly what he is doing, both design/engineering wise and commercially... it's probably the commercially aspect that upsets people. To charge this money for a DAC is nothing short of highway robbery.


JSmith

 

[Xyrium]

Are those controls for an xbox, on the right side? Nice! Perhaps my kid can dig down a few hundred blocks and bury this brick somewhere.

 

[amirm]

We have some interesting answers from Rob Watts: https://www.head-fi.org/threads/chord-electronics-dave.766517/post-17065133

So much for JA measurements showing excellence!

Good to see that he accepts other DACs do better in noise department. The challenge for him is that those DACs cost as little as $150!

As to it being inaudible, how come this noise is, and others are not?

Wow. How broken an argument is that? So his marketing is super effective. What does that have to do with performance of the DAC?

There is nothing "essential cue for the brain there" or every music producer would throw up on their own creation when they listen to non-Chord DACs!

Rob has no background in psychoacoustics yet he keeps saying, brain works this way and that way. Show a reference at least.

And transients being randomly too early or too late? Really? What DAC does this non-deterministic thing?

It is not as he hears them. it is as he hears and sees them considering how he never performs a blind test. All of a sudden when effective marketing causes people to spend money, becomes a proof point too! Even the most ineffective audio product sells. There are people who sell boxes of dirt with a wire and call it grounding. They sell them for thousands of dollars. Is this good enough for Rob Watts that they make audio sound better?

Finally:

Oh really? If SINAD is dominated by noise, then you better not play more than 95 to 100 dBSPL your will hear background noise. Given that he posts in head-fi, he better be every careful with that statement as that causes noise in sensitive IEMs. This is why DAVE DAC did poorly in this test:

You absolutely can hear this kind of noise. Even 90 dB can be audible let alone 82.

All in all, it is amazing that my comprehensive measurements are counted with words and opinions.

 

[Cote Dazur]

That's just too money for the performance.

I cannot even start to imagine what kind of “performance” a DAC would have to reach to justify that kind of money, when low cost unit can already achieve what a DAC possibly have to do, same for amplification.
The type of hype and pricing of this device is all that is wrong with this hobby.
For this type of money we should be able to have bat ears transplant, that would really be a different experience.

 

[amirm]

I think any expensive DAC in this day and age also needs to be a streamer as well with a large, colorful high-resolution display. Selling simple DACs for thousands of dollars makes no sense to me.

 

[Garrincha]

We have some interesting answers from Rob Watts: https://www.head-fi.org/threads/chord-electronics-dave.766517/post-17065133

View attachment 220692

So much for JA measurements showing excellence!

View attachment 220693

Good to see that he accepts other DACs do better in noise department. The challenge for him is that those DACs cost as little as $150!

As to it being inaudible, how come this noise is, and others are not?

View attachment 220694

Wow. How broken an argument is that? So his marketing is super effective. What does that have to do with performance of the DAC?

View attachment 220695

There is nothing "essential cue for the brain there" or every music producer would throw up on their own creation when they listen to non-Chord DACs!

Rob has no background in psychoacoustics yet he keeps saying, brain works this way and that way. Show a reference at least.

And transients being randomly too early or too late? Really? What DAC does this non-deterministic thing?

View attachment 220696

It is not as he hears them. it is as he hears and sees them considering how he never performs a blind test. All of a sudden when effective marketing causes people to spend money, becomes a proof point too! Even the most ineffective audio product sells. There are people who sell boxes of dirt with a wire and call it grounding. They sell them for thousands of dollars. Is this good enough for Rob Watts that they make audio sound better?

Finally:
View attachment 220697

Oh really? If SINAD is dominated by noise, then you better not play more than 95 to 100 dBSPL your will hear background noise. Given that he posts in head-fi, he better be every careful with that statement as that causes noise in sensitive IEMs. This is why DAVE DAC did poorly in this test:

You absolutely can hear this kind of noise. Even 90 dB can be audible let alone 82.

All in all, it is amazing that my comprehensive measurements are counted with words and opinions.

Well, these unfortunately are no real answers, as they firstly are not directed to the person who brought up the criticism and secondly just dodge the points with words and subterfuges. No real argument to the point, own measurements or anything substantial, but this, alas, is not to be expected.

 

[bidn]

I am an electrical Engineer working in audio, for a company that I won't disclose because I am not commenting here on their behalf, it's not my rôle. I have been developping and manufacturing audio products in the past.

If you were truly an electrical Engineer working in audio,

- Why did you spell the word algorithm many times, and each time, "algorythm" ??
Not only no engineer, but no person a bit technically knowledgeable would do that.

- Why do you make so many technically wrong statements?
( I know you are trying to create lengthy, derailing discussions, and will ask me to list them and explain what is wrong. I am not falling into this trap, your posts speak for themselves).

If you were truly an engineer from a quite different field (i.e. mechanics):

- Why do you make so many wrong statements about things you don't understand?
An engineer or a scientist knows things are complicated and that you just cannot make statements about things in a field you don't know. You would instead first e.g. pick up some textbooks and learn the basics, then intermediate things, and finally more advanced and specialized ones.

For me it definitely can't be true that you are an "electrical Engineer working in audio" with such a C.V.
Instead, with your too numerous wrong posts trying to defend Chord, you come over rather as a paid shill (this kind of situation happened with some posters doing the same in several threads, e.g. threads on MQA).