ESS Sabre Reference DAC (8-channel)

Hi,

I assume that Your answer qusp addressed to me.
Well as I told before, I find both the information politics as well as the price politics of ESS ridicolous and tend to not support such politics. I regard 600 bucks for such a rather simple eval-board and 30 bucks for the chip itself way too high.
The measurements owen took are all made in differential mode. Asked, he told that the ´single ended´ mode resulted in much higher THD-values because the even harmonics wouldn´t cancel out. The BB/TI PCM1794 measured with different resistor values as I/V-converters in single ended mode already got results comparable to what ESS claims in differential mode.
But apart from that, has anyone besides Owen done any reliable measurement, or is all this just talking about marketing hypes and theoretical datasheets values????

jauu
Calvin

Twisted Pear and Wavelength have both achieved circuits that meet the specification of the DAC, so clearly state of the art figures are possible with the right output stage. Given that, calling the output stage "bugged" seems a little strange ...

As for the price, it is a little expensive, but you do get multiple S/PDIF inputs, oversampling, upsampling with jitter reduction and 8 DACs in one chip. Not such bad value IMO.

If you want to find out more about the chip, there is plenty of info downloadable from ESS' website. There are also patents available to view if you want the heavy stuff.
 
but does it improve on sonics in any way??
Whatever an analog stage of the ES9018 DAC is, I think those 2L DXD 352.8 kHz /24 bit WAV files can be recognized better on sonics than those of 192 kHz/24 bit counter parts. (I have listened to ES9018 DAC output without any so-called I/V stage.)

So is there any sense in that besides marketing??
I think there is enough sense besides marketing for me.

I will strongly recommend you to listen to sounds of ES9018 actually if you have not. Have you ever listened to them? I hope you to make the point clear before us first.
 
Hi,

@Spartacus:
The best technical results like THD and noise I´ve seen have been reached under usage of OP-Amps. Still though I would like to see the OP-Amp stage that offers true 24Bit linearity, let alone 32Bit linearity :D (Even the AD797 from the ESS AppNote is specified at max. for 16Bit linearity). ESS themselves state that the analog stages are the limiting factor.
The problem I have with OP-Amp stages is a sonical. On true high resolution speaker systems like good ESLs they imho suck. I have to come across a OP-Amp stage that really sounds like music and not like HiFi or reproduction. Thats what I meant with inferior (You called it ´bugged´).
There is indeed a lot of marketing blabla to find on the ESS-Website, but hardly any flesh. If You want to see how a decent datasheet and better support looks like take a look at BB/TI and ADI and AKM and..and... and.
And yes there are patents.....which´s purpose is to claim intellectual property. But a patent guarantees not, that the claims describe something new, something improved, or something superior.
Thats why I asked for some independant hard facts and measurements.

@Bunpei:
Since my first post #1511 I asked solely for the existance of ESS-independant reliable technical data, because ESS refuses to show any valuable information and the few numbers they offer would be good for a 24Bit DAC, but from a 32Bit system I´d expected better figures.
I have not asked for a discussion of sonics and won´t argue about that. Met too many people with different taste.

@qusp
"9018/12 spits in the face of your modern chips internal 384khz clocking"
I still try to find some intelligency in those words :magnify:
If You can prove to us any practical improvement of such a technical detail, I´ll really try to understand :D
The standard for the next decade will very probabely be 24/96 or 24/192, but even if studios would switch to the 24/352 mentioned by Bunpei this format will be happily accommodated by any modern DAC-chip.

jauu
Calvin
 
really, you struggled with that??I thought I included enough visual content for even the most challenged english speaker. pretty straight forward sentence if you know what numbers i'm talking about. you mentioned a number claimed in a datasheet, I merely mentioned that the same spec in the ESS datasheet is not even close. then you went on to be redundant, I already said the input stages (which includes source) are the things that need to catch up for the specs to have any meaning; until then its just headroom.

but TBH, this is getting boring, its clear you have no interest in trying it out and it is clear, since you keep avoiding the question, that you have no experience of what you are talking about. all you seem interested in doing is speculation and argument; its getting dull and i'll leave you to it. theres a word for what you seem to be doing at the moment WRT the sabre on the forum, but I prefer not to use it too casually; i'll leave you to wonder what it is
 
Last edited:
Hi Calvin,

It's not relevant what technology was used for the output stage of the DAC. It simply demonstrates what the DAC is capable of, i.e. that the DAC is "blameless" to use designer Doug Self's term. If a user doesn't like opamps, fine, they can use what they want, but they can't blame the DAC if their schema introduces further distortions. Therefore it is not correct to call the DAC "inferior".

As far as info about the DAC, I agree that the datasheet could be more informative (can't they all?). However there ARE independent 3rd party measurements available on ESS' site (what other manufacturers provide this?), as well as plenty of info available about the inner workings of the chip. Seek and you shall find.
 
Hi,

@Spartacus:
A third party measurement without naming this party and not naming the competing devices ...... how reliable could that be on a website that doesn´t present data and information as any other serious competitor in that market gives? Well I prefer to rely on statistics that I fake myself ;-)
I didnt call the DAC inferior and yes, there is some information about the inner workings of the chip. But how does that translate into a true practical advantage? That´s why I asked for figures, which so far no one except matejS really referred to.

@qusp:
No I don´t know which numbers you are talking about. You cited owen´s work, which was always a differential value. I understood Owen´s figures as of the I/V-stage alone, fed by the AP-measurement system, not the ESS. So far no other has presented any figure, especially no figure of a single ended measurement.
Rather than questioning my motives -which is btw utterly nonsense- please answer my Qs or don´t answer at all to my post.
Either You can provide reliable data or You can´t. I´m not going to discuss with You about marketing data.

jauu
Calvin
 
Hi,

@Spartacus:
A third party measurement without naming this party and not naming the competing devices ...... how reliable could that be on a website that doesn´t present data and information as any other serious competitor in that market gives? Well I prefer to rely on statistics that I fake myself ;-)

You can choose to distrust any measurement you want .... but you wanted independent measurements, and there they are. Others have told of excellent performance too. Again you can distrust them. But the available evidence suggests the DAC's output is not inferior, as you claim.
 
Hi,

where the **** did I say anything about the ESS having an inferior DAC output??? :mad::mad::mad::mad:
I just said that I dislike the company´s information and price politics and that I´d expect rather better than the claimed values for a 32Bit DAC. Then I asked if someone could provide for some decent and neutral measurement data, especially in single ended mode. Thats all. Obviously no one is interested in such facts or engages in that theme or is able to comply.
But cheer up, I won´t disturb Your hiping session any longer :rolleyes:

jauu
Calvin
 
Hi,

where the **** did I say anything about the ESS having an inferior DAC output??? :mad::mad::mad::mad:


Obviously the quality of the ´current sources` of the DAC is low, presenting just 200Ohms output impedance (similar to Microsoft I get the Impression that ESS seems to sell bugs as features, telling You for example to have the ´choice´ between current steering I/V-conversion or voltage output, when in fact the impedance is much too low for a decent current source and too high for a decent voltage source)
 
Last edited:
Hi Calvin,

Im not 100% sure what type of information you are after, but I can assure you the numbers posted on the ESS sites are true. The third party measurement was done by Steve Peterson of ATECS. I think you will find he would be a reliable source.

ATECS - Audio Test Engineering Consulting Services


Since the numbers were posted in the datasheets, we have actually made better measurements, but have chosen not to bother updateing the datasheets for confusion. The AD797 has consistantly provided the best numbers getting -135dB of DNR (A-weighting) and can do -120dB of THD+N unweighted across from 20 to 20k. We have shown other OPAMPs can do >118 THD+N at 1kHz, which is impressive, but, try measure at 5k and 7k. Thats were almost every other device tested showed "issues" dropping into the -110dB THD+N or lower. The DAC is sensitive to the power supply, but this is acutally on purpose, It would have been very tough to generate a analog reference that can achieve -135dB A-weighted 20-20k on the chip that has 1024 DAC's (64 per pin, 16 pins).

As for the sound of the DAC, it is greatly influenced by the parts connected to it. I believe this is where audio companies can come in and use their knowledge and expertise to have some product differentiation.

For the output stage:

I dont think we ever claimed a current source, but rather a "current mode". The current mode is simply when the current going in and out the pin of the chip is sensed. This mode has the benefit of cancelling 2nd and 3rd harmonics of some of the internal ananlog circuitry. The "voltage mode" is when the pin of the chip has a voltage on it that is being sensed. While this has the 2nd and 3rd hormics (at the -100dB level or so), some people have even claimed this mode is more "tube-like" Its all personal prefference.

I hope this helps out a little.

Thanks

Dustin
 
Hi Calvin,

The question cannot be answered with just a simple statement the way its asked. What type of circuit do you plan to put on the DAC output, what do you mean by "single ended"? If you mean taking the DAC output into a DIFF to SE configuration, the performance of the THD will very from about -90dB or so (votlage mode) to -115dB or so if you use the 2 I/V converter then a DIFF to SE following it.

If you mean just completely ignoreing the DACb outputs, the THD will again be mostly influenced by how much you let the output pins of the DAC move. You could ground the DACb and take the DAC signal into a buffer stage, this would be around the 90dB level at 0dBFS. If you took the SE output into an I/V conveter, you would be back to the -115dB level. If you provide a circuit you have in mind, I can give you some better idea.

Dustin
 
Hi Dustin,

I'm using transformer output with my Buffalo II. Wired across + and -. I like it alot.

Was wondering if the ESS Sabre chip likes to see the output's + and - current returned to the DAC's AC signal ground @ AVCC/2 (by using the transformer's primary center tap), or is simply wiring the transformer from + to - (and ignoring the center tap as I have done) complete the shortest current path ?

Great job on the Sabre chip, BTW.

-Steve
 
Hi Dustin, Calvin

I too am extremely interested in using the ESS9018 in a single ended way, with a discrete component I/V stage. I intended to use "half" of the differential output- i.e. to feed the I/V converter from either the + or the - output, and ground. The I/V circuit can be set up to provide 1/2 AVCC on its input.

But you mention grounding "DACb". What did you mean by this please?

Many thanks

Paul
 
why wouldnt you use the diff out and then sum the output of that into SE? seems the best way to gain all the performance, otherwise you really take a hit to DNR and THD no?? not only by not having all the information there, but also because you throw away ALL of the CMMR and none of the dacs own self-noise is cancelled either