ackoDAC based on ES9018

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
both :) You know this is a very mixed IC (ASRC included),
...and this Q is difficult to answer. For 88.2 , 96k material it might be as good as it gets, but im not so sure about 44.1 , nor DSD.
We will see once there will be Sabre32's with DSP front end. (If they subscribe to the 'hyperstream' )
Until then, I recommend reading this:
Page Title
 
Last edited:
Thanks guys, your points are well taken.

I am doing this all for fun and sharing it so contributions will make it more interesting. Yes, Jkeny you have done you your share in this project and it I will offer a free bare board (+ free shipping) to you once ready. How about that for carrot:)

Yes, it is not the pcb costs that will matter. I could flog it of to china and get it for 2 a penny but at the end of the day the sabre chips and high end crystals etc dominate pricing.

Ok then, I will have to get back to rest of you all. Initial batch of boards have now all 'sold' out. I will get it touch with those who have contacted me.

Regards, acko
 
Last edited:
Thanks guys, your points are well taken.

I am doing this all for fun and sharing it so contributions will make it more interesting. Yes, Jkeny you have done you your share in this project and it I will offer a free bare board (+ free shipping) to you once ready. How about that for carrot:)
That's very kind of you acko, I was not expecting that - I was only replying to Curly Woods when I made my remark. I'll PM you :)
Yes, it is not the pcb costs that will matter. I could flog it of to china and get it for 2 a penny but at the end of the day the sabre chips and high end crystals etc dominate pricing.
That's why I wanted to use the lower price Crystek clock with a good PS ( see Jocko's low noise board on diyhif.org) - it can perform as good as some of the "audiophile" clocks mentioned. Remember with any low jitter clock you are going to need a superlative PS to realise anything like the jitter figures quoted for the clock (this is the fact that the clock specs don't reveal)
 
Hi Guys

Oppo Digital has just announced the SE version (add USD $400 over the standard model) of their Blu-Ray player. The SE improvements are purely on the audio side, going from using CS4398 Stereo DAC and CS4382 Multi-Ch DAC to ESS ES9016 (paralleling the outputs) Stereo and ES9006 on the Multi-Ch.

I would love to see PDF data sheets and the suggested application notes they usually include - how is the ES9016 paralleling up done, I versus V modes etc.

Does any have these or know where they can be downloaded?

Cheers.

Joe R.

PS: Acko, I'm in Sydney too. :)
 
Any updates?:)

Peter.

Well, it seems that NDA provisions by ESS means you cannot reveal the datasheet and also you cannot even reveal your own schematic of how you employed (implementation) their ESS DAC?

For example, the SE version of the Oppo Blu-Ray uses the ESS DAC (ES9016) in the dedicated stereo outputs, is it in voltage mode or current mode? Is it OK by the ESS NDA to let us know how the DAC is put into either mode and which, if any, there is a default mode? I am getting an SE within a few weeks and perhaps I will be able to answer which mode it is in as I gather it can be put into one or the other? Or is it possible to have both and if the Oppo uses voltage (is this an inbuilt opamp I/V converter?) can the current out still be used and a separate I/V stage be built and go to other RCA out while still maintaining Oppo's RCA outs.

I think what ESS is doing does not make sense to me. It only builds up frustrations.

ESS should learn how to deal in good faith, even if an NDA is required it should have reasonable lattitude. Why should the ESS user be prevented to reveal his work by way of showing his schematic, whose property is that? I would never sign an NDA that would prevent me from sharing something that is mine to share. No NDA I have ever seen has prevented any manufaturer to show a schematic, not even a block diagram? Also, I wonder if this was thought up by a tech guy? I think not, rather a stuff shirt with a lawyer standing by his side.

Who does ESS think they are? Microsoft?

But seriously, what does the NDA allow and what does it not?

Joe R.
 
Last edited:
Joe,
I have the datasheet & can give some of the info that has already been revealed publicly by Dustin - each DAC puts out 0.924*AVCC = 3.04V PP on the analog outs - it also puts out 0.924*4.244mA = 3.903mA PP (so 4*3.903 = 15.6mA PP for 4 channels) with a bias of AVCC/2 (1.65V)
 
Hi Everyone,

Sorry for the delays but the design team has been working very hard to provide a highly optimised layout and options to bring out the best from the Sabre chips.

Prototype boards are due out next week for tests and will post updates soon.

As for queries on Schematic/design info of Sabre chips on public domain, we need to abide by NDA restrictions (sorry). But negotiating to allow those who purchase the DAC unit itself to have schematic included in the user's manual for reference purposes. Will see how this goes.

Thank you for your patience.


Regards, Acko
 
Last edited:
Joe,
I have the datasheet & can give some of the info that has already been revealed publicly by Dustin - each DAC puts out 0.924*AVCC = 3.04V PP on the analog outs - it also puts out 0.924*4.244mA = 3.903mA PP (so 4*3.903 = 15.6mA PP for 4 channels) with a bias of AVCC/2 (1.65V)

Thanks.

The voltage outs have "+" and "-" outputs, in which case 3.04V PP is would be across those and the outputs sits on 1.65V. That makes sense. Is this an internal opamp solution? If it's an inbuilt I/V then the current output would be disabled if used in voltage mode. If both are available at the same time, then an I/V is not used for voltage outs. The volts would have to be derived separately.

Now for the current outputs, the current output is certainly healthy, no doubt because of paralelled DACs. But if what you call bias is what I call offset voltage then the I/V conversion has to float on 1.65V whereas B-B PCM/DSD/1792/1794 etc does not, no DC on the output. The I/V converter I use thrive on the B-B's 0V (and other which are the same) but are a pain when I have used it with AD1955 (2.39V). Does the ESS have a Vref 1.65V or does that have to be generated externally.

Also, does it have a bias/offset current. In the case of AD1955 it's 3.04mA - if zero that's makes it simpler than AD1955.

You know, stuff like this should be available, absolutely no reason whatsoever that it shouldn't.

But thanks anyway. Anything you have along these lines are appreciated.

Cheers.

Joe
 
ackoDAC Preview

Hi Everyone,

I am nearing completion of the new DAC version and also all the support modules required to build up a complete system. See attached files for preview. All these will go for fab next week and boards should be ready by mid-Jan for assembly/test and then release. Happy New Year!
Acko
 

Attachments

  • ackoDAC Preview.zip
    918.3 KB · Views: 714
Hi Everyone,

I am nearing completion of the new DAC version and also all the support modules required to build up a complete system. See attached files for preview. All these will go for fab next week and boards should be ready by mid-Jan...

I am quite interested in what you are doing, despite my being miffed at ESS unfriendly policies, and have some questions:

A. What do you use as I/V converter? Surely no NDA can prevent you from revealing that. This is such a key to great performance even with the best of DACs.

B. What or where is the source of clocking. What if I use the SPDIF, does the transport become the "master" and the DAC "slave." I don't necessarily pooh-pooh that out of hand as some would. But if this is the case, the transport (if used with such) becomes crucial. Or can the clocking easily be incorporated/updated. Very much feel this is a BIG part of getting the best that digital audio can do.

Link to Terra Firma Article.

C. If able to come around to my place I would like to discuss with you a simple and yet confidential trick that you may not be aware off, one that I am sure can be applied to your PCB (close to where the DAC is located... hint) without any changes. Would also love to hear it at the same time.

Cheers.

Joe R.
 
Last edited:
yes joe, you might consider reading the info ;) all your questions and more are answered. the I/V stage is what you make it. there is an opamp version available or you can use whatever you wish, beit tubes transformers buffers etc etc. same with regulation. all regulation points are exposed and very near their target.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.