Anybody using the new ESS Vout DAC (ES9022)?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I bought some of these to basically have a play & hear what it's jitter reduction can do for an ordinary Vout DAC.

I have a Musiland USB DAC which uses PCM1793 DAC chip in I2S mode. I was hoping it might just be a matter of tapping into these I2S lines & feeding them to the ESS DAC. but no sound just a low whine. I then read the datasheet more carefully :eek: & saw that the ESS DAC states that MCLK must be > 192fs in asynchronous mode. How can I check if the existing board operates asynch mode to the DAC & if MCLK >192fs?

I have a 100MHz scope but barely know how to use it so some guidance would be appreciated here also, if it's feasible to check this.
 
Hi Jkeny....
Please do not take this the wrong way. But if you don't know how to use an scope, what is the chance of you building a high performance DAC-system?
Making a DAC chip work in a simple implementation is quite easy. But achieving optimum performance is lot harder than just copying a app.note (Even thouhg many commercial DAC's really just is a app.note-style implementation). Building such a DAC, is much more complicated than using a scope.

Anyway... Where did you buy these ESS chips?? And what is the price?? Is it possible to get a datasheet??
 
Hey Hurtig,
I agree with what you are saying about optimum implementation BUT.... if you re-read my post I was firstly trying them out to see if they held any promise:).

Then I might get into putting together an optimum configuration but I wouldn't do this with this particular Sabre DAC as it's got an internal op-amp (oh perish the thought) from which you are known to break out in a rash at the mere mention of the word :D

BTW, I agree with probably 90% of what you are saying on the Buffalo thread & I posted my thoughts but I think it's time you stopped talking & started doing! get your DAC out to those interested & let some independent review happen. My contention with the Buffalo is that it in it's first incarnation it was value for money for those who wanted to hear the Sabre DAC as per the app note schematic (not an optimum implementation according to Dustin's account of how sensitive the PS is!).

The SABRE32 implementation is way up in price and no longer good value particularly as the PS issues haven't been dealt with properly. My own implementation will be less than optimal in a layout sense but will deal with the PS issues & allow tweaking - I wonder how it will fare against the Buffalo?

So you see, I understand what you are saying here & in the Buffalo thread but at the moment I was not going for the full body press solution, certainly with this chip - just cheap & cheerful!

You can get the chips form Ismosys.com, here in Europe - I don't know if ESS want yo to sign an NDA for this $2 chip or for their flagship 32bit chip (ridiculous policy anyway!)

Anyway, I got it working (without a scope) :spin:

Now I need to optimise it somewhat to give it a fair chance
 
The good news is I got it working, (a disconnected wire I hadn't noticed) - the bad news is it's not as good sounding as the on-board PCM1793 DAC but I have to look at the implementation - it's not optimum :D, and see where I might improve it.

It sounds a lot more like an ordinary digital playback system - the midrange or sound stage isn't as believable as the PCM. There's a bit of high pitched whine when volume is low & I'm not sure it isn't there all the time muddying things. When I feed it through a transformer (as per the PCM DACs) it cleans up somewhat. Maybe there's more to be had from it
 
Nobody else using these?

I have a question, anyway - I tapped into the I2S traces going to a PCM1793 DAC to feed them to the Sabre DAC. Now, I haven't cut the I2S traces to the PCM DAC or powered it off during my listening to the Sabre DAC. Why, because it would be somewhat irreversible & I didn't want to do that yet, anyway.

This is one of the factors that I thought might be causing the whine. Anybody got any info on re-routing I2S?

I use a 4 " flat cable of 9 wires - 4 signal wires & 5 interleaved ground wires. I know the cable should be as short as possible and all wires of equal impedance (length) but I don't think this is causing the whine!
 
Yes Kurt, them's the prices :). I have a 9102 which I will be building a board around at a later date so I wanted to experiment a bit with the cheaper 9022 first & get an idea of the family sound of these chips. Hence my casual approach to the implementation - I wasn't going for the best DAC build possible just a sampler of the ESS sound & maybe a better implementation if it showed promise!

So far, I'm not impressed but then I don't think the chip is giving of it's best & I wanted to resolve this!

Any ideas?
 
I do not either know the sound of the ESS family, and I do not know if there is any family sound at all.
What I do know is, that there is some kind of family sound amongst the BB DACs. And if you should want to compare the BB DACs to the ESS, you should use 2 pcs. of i.e. 1794A or 4 pcs of 1704K pr. channel, to get it right @ 24 bit.
If you want to compare 32 bit performance, you should use 2 pcs. of AK4399 pr. channel paralleled, and a CS8421 and then one pc. of CS8416.
So the ESS flagship DAC does really include a lot of sophisticated digital technology in one chip.
This might be a good idea, and it might not, I don´t know.
The only experience I´ve ever had with a lot of DACs parallelled was with an Accuphase CD player, where I think there was 8 BB DACs pr. channel, operating differential and thus 4 of them parallelled. Maybe it was only half what I´ve said, I do not remember, but the sound I really do recall.
16 or 8 or what ever DAC chips and then run through 4 JRC5534 op-amps. It sounded redicoulus, even the owner heard it, he hated it from ever after.
But it looked great, my god it did.
So! I really look forward to listen to the Sabre DAC with a nice powerfull no nonsence analog stage, with a well thought throug reconstruction filter.
 
Yes Kurt, them's the prices :). I have a 9102 which I will be building a board around at a later date so I wanted to experiment a bit with the cheaper 9022 first & get an idea of the family sound of these chips. Hence my casual approach to the implementation - I wasn't going for the best DAC build possible just a sampler of the ESS sound & maybe a better implementation if it showed promise!

So far, I'm not impressed but then I don't think the chip is giving of it's best & I wanted to resolve this!

Any ideas?
From my experience I´d say, that you should not really have any doubts.
You actually should get a feal of it in some sence, if you are a bit carefull.
Do not try to listen to any component with less than AN supplies. But from there on, you mostly can rely upon your impressions, and often it will amplify when implemented correctly.
In digital design there is some mysteries though. I.e. Clocks are very special, I do not yet know their sonic nature, but until now it seems like everything is upside down compared to nerd knowledge.
Nerds normally likes discrete clocks with high accuracy, but I hate them from the buttom of my heart:gnasher: They are no go:no:
 
Kurt, this is kinda getting away from the thread topic but based on my experience with the PCM1793 DAC, I would use a transformer output stage for superior experience. Others have also stated how good this sounds when used on the Buffalo board in place of the on-board output stage!

Nobody got ideas on why I'm hearing a whine? Somebody told me the fs from the Monitor 01 USD is 100KHz - I'm not sure exactly what it means that the word clock is 100KHz
 
Last edited:
Maybe the wine is there because the reciever chip powers down, or is not sending any data when no signal is present. That might cause some preceeding components to make noises when ideling. Or the ESS DAC chip is not powering down when no signal is present.
You also might find out if the logic is set to run @ 3,3 or 5V, this also could be a reason to malfunction.

About the transformers, I have a few words about them. A transformer is a reactive component, which means that it will try to resist current changes. On top of hat it also shows hysteresis, which is really bad to low level signals. I´ve heard transformers used in various ways, both as passive preamp and coupling devices in DACs, the result was always "out of any control". It is actually a peculiar thing, because you listen to it, but it seems very hard to find the right volume to play back in a natural way.
Also in my vinyl days this problem was present in MC transformers, which were way surpassed by real headamps.
To me the worst component i can think of, when put into the signal path, is a transformer, number 2 and three must be both electrolytic caps and op-amps:D. I´d go for a buffer of some kind, I think schematics and even Gerber files can be found on the net somewhere.
If you do go for transformers, let us know if it´s any good.
 
Last edited:
Kurt,
You missed my thread on transformers as output devices - I found this particular transformer to sound excellent, best sound I ever heard from a DAC! I think you should review your opinions & try some good quality ones!

I´ve tried out very expensive ones for my MC cartridge.
I´ve also lately tried out a passive preamplifier @ a terrible price but, I must say that I have no urge at all to try out transformers again.

Btw I did see your transformer thread.

Transmission transformers are in some way a 2 edged sword.
They do provide insulation, but they are prone to hysteresis. Thus you´ll see i.e. high end electred microphones made in 2 different versions, namely normal or TL (Transformer Less). The normal type can be used anywhere without any hum probs. The TL version should not be attached to a studio environment, where a lot of gear is connected together without thought about loop probs.
If you ever should get the opportunity to listen to a "normal", and a TL version of the same mike, you really should try that out.
Then you´d know about transformers in your signal path, even if they are state of the art.
 
Member
Joined 2009
Paid Member
Hi KvK and Hurtig,
How come this continuos interest in the ESS dac chips? Are you considering to change your DAC circuit (Elab digilog, I think it is named) to accommodate one of these chips in place of the "oldish" Crystal?
Nic
[Edited] P.S. Sorry jkeny for going a bit OT. I won't make anymore noise.
 
Last edited:
Hi KvK and Hurtig,
How come this continuos interest in the ESS dac chips? Are you considering to change your DAC circuit (Elab digilog, I think it is named) to accommodate one of these chips in place of the "oldish" Crystal?
Nic
[Edited] P.S. Sorry jkeny for going a bit OT. I won't make anymore noise.

No!
The digilog is designed arround CS4398, and will eventually be replaced with the next Crystal chip, which probably will be just that same, but with 32 bit wordlength, and also the AD1896/SRC4192 will then be replaced with 32 bit types, probably of the same breed.

But we are planning to do a low cost, easy assembly 32 bit DAC with upsampling, where the Sabre could be considered. But most likely it will be CS8416, CS8421 and AK4399, and then an analog stage with 2 choises of amplification, a standard op-amp operated one, and a non feed back design featuring AD844 on the same PCB.
When done gerber files, partslist and schematics can be found on the internet.

Then! ESS tech´s chips are of course of interest to us, the Sabre 32 chip shows faboulus data if THD and DNR are considered, we might not be able to squeeze that through any analog amplifier though. We might already have problems to equal the AK4399 IRL, which shows a 123 dB DNR and a 105 dB THD+N.
The most interesting part though, is how it was done. 6dB can be concidered as a result of parallelling 4 DACs p.ch. That leaves us 130 dB DNR if compared to other stereo DAC chips. How this was achieved I do not know, but rumours say that it´s done by high speed oversampling @1MHz or so.
 
Hi KvK and Hurtig,
How come this continuos interest in the ESS dac chips? Are you considering to change your DAC circuit (Elab digilog, I think it is named) to accommodate one of these chips in place of the "oldish" Crystal?
Nic
[Edited] P.S. Sorry jkeny for going a bit OT. I won't make anymore noise.

OK, Nic, I was going to say to the two of you to take this outside :D but there is nobody posting on topic on this thread - I take it nobody has tried these ESS DACs then!
 
Last edited:
Hi Jkeny....
Please do not take this the wrong way. But if you don't know how to use an scope, what is the chance of you building a high performance DAC-system?
Making a DAC chip work in a simple implementation is quite easy. But achieving optimum performance is lot harder than just copying a app.note (Even thouhg many commercial DAC's really just is a app.note-style implementation). Building such a DAC, is much more complicated than using a scope.

Anyway... Where did you buy these ESS chips?? And what is the price?? Is it possible to get a datasheet??

Hi, Product and support available from the European stockist Ismosys
 
Good to see you here, ismosys.

Perhaps you can help your many (potential) customers here. Can you tell us how to get hold of a datasheet for ESS9022, ESS9016 and ESS9018 ? Can we get them from you ? Or do we have to purchase a chip first before we are allowed ?

I have mailed ESS umteen times and got no reply. How can I decide to spend 65 USD on a chip if I am now even allowed to know how to connect them ?


Many thanks,
Patrick
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.