I'm going to order a buffalo dac with the ESS9038 but I have a few questions.
I haven't been able to find a clear answer to the current output capabilities of the ESS9038 in stereo mode, I gather it's probably like 60ma right?
The output stage I designed can't handle this current swing and I'd like to avoid redesigning it.
I see mention of a digital attenuation option in the dac, is this the same as turning the volume down on the PC? As in the audio data remains in tact as long as you don't turn the volume too low?
If I can attenuate the current swing I can avoid redesigning the output stage.
Otherwise I might be looking at the es9028 instead.
I haven't been able to find a clear answer to the current output capabilities of the ESS9038 in stereo mode, I gather it's probably like 60ma right?
The output stage I designed can't handle this current swing and I'd like to avoid redesigning it.
I see mention of a digital attenuation option in the dac, is this the same as turning the volume down on the PC? As in the audio data remains in tact as long as you don't turn the volume too low?
If I can attenuate the current swing I can avoid redesigning the output stage.
Otherwise I might be looking at the es9028 instead.
You could use a es9038Q2M, same chip as on the recent Chinese DACs. It appears to be mostly a 9038, but with only 4 channels instead of 8 and maybe lower output currents. Using a single channel for left and one channel for right can easily be handled by a couple of opamps per channel such as with LME49710's, or maybe one LME49720 per channel.
The output current issue is biggest when people run channels in parallel to achieve very low noise levels that the total current gets really high. Even in that case it can be handled by multiple IV stages running in parallel, then summing the outputs together after conversion to voltage.
The output current issue is biggest when people run channels in parallel to achieve very low noise levels that the total current gets really high. Even in that case it can be handled by multiple IV stages running in parallel, then summing the outputs together after conversion to voltage.
Last edited:
I'm beginning to wonder. Is it better to split the 8 channels into 2 differential channels and then sum the differential channels into stereo in order to cancel out common mode distortion?
So instead of running 4 outputs in parallel per channel it would only be 2 with the bonus of distortion cancellation.
Wouldn't that give better performance than just running the channels in parallel as stereo?
Also is the 9038 really that much of an improvement over the 9028?
So instead of running 4 outputs in parallel per channel it would only be 2 with the bonus of distortion cancellation.
Wouldn't that give better performance than just running the channels in parallel as stereo?
Also is the 9038 really that much of an improvement over the 9028?
Last edited:
What do you expect from a DAC chip offering up to 140 dB DNR and > 120 dB THD&N? It is measured to have the worlds best perfomance so it must offer worlds best sound experience also? I cant speak for the new ESS Chips, I only owned a Buffalo 9018 and a Exasound E18. They did not sound bad but caused me over time stopping listening to music. The remote control was best friend because I did not like to listen a song from begin to end. Listening a few seconds, fatigue - skip to next song and so on.
Delta Sigma was not developed to be soundwise superior to R2R DACs. The main reason was cost reduction. I am wondering no one consider suspicious the ESS behavior to require non disclosure agreement to send out some specs and measurements. If you compare data sheets of good old days R2R DACs (not the cheap ones) with DeltaSigma datasheets it is already funny what specs (the really important) were measured in the past and what you find in the Datasheets of modern DACs.
I sold meanwhile all my DACs and purchased a 18bit Multibit DAC from a US company. There it is, analog like natural and extremly good low level detailed sound. Listening to complete albums again and not skipping after a few seconds to next song because of listening fatigue.
Just my 2 cents
That's a post dac design problem. My ESS9018 dac could not be more silky, delicate, subtle and non fatiguing. It's all about implementation. I'm pretty good at analog design but I'm only just beginning to learn about dacs. I'm hoping the buffalo will give me a lot of insight during the "build".
I'm really hoping to get a response on my common mode distortion cancellation idea.
This is a quote I found on the internet
Is the 9038/9028 what he refers to as "true" balanced" Is there common mode distortion to be removed?
Last edited:
So what's the secret sauce to getting that kind of sound out of an ESS DAC? Are you going to give us any pointers?
Sadly not. My design is truly unique and I'm going to have to apply for a patent. I have a few patentable designs that are quite unique and took me years to develop. My end goal is to start a company but I have much to do before that point. I have a lot of interesting and unique stuff cooking up.
I would kill for R&D assistance to speed up the process, it's just a lot of work and money for one person, but everyone I know is completely uninterested in electronic design.
I've been living in poverty for the last decade so I can't afford to be throwing my designs around unfortunately, they are my way out.
Rather ironic that you'd be posting on a DIY audio forum then is it not? Thanks for clarifying that you're only after receiving help from others and apparently not giving any out yourself.
Sorry if not wanting to throw away the last few years of my life and my future to satisfy your question (which is entirely unrelated to the questions I posed in the op) makes you unhappy...Why don't you make less than minimum wage for over a decade and see how much you enjoy it.
Last edited:
I have heard using transformers as output stage and volume control for ESS 9018 should be an improvment over using OPAs. If you add a discrete output stage with your secret sauce perhaps it better things also.
Good luck with your plans!
Thanks 🙂 But boy is it a lot of work. It's consumed my life. I would kill for an assistant.
Is there anyone that knows the answer to my question about common mode distortion cancellation in the ESS9038 dac? Is the internal processing single ended or differential when you use a differential output?
Last edited:
You need to go sign an NDA with ESS, and it's not fair to ask people here to help you with your design work if you can't share the details.
The forum is really all about diy'ers sharing their hard won knowledge, ideas and experience with the other members.
If you have unusually good analog/mixed signal circuit design chops there is no reason for you to be working in a low wage/low skill job, although you do have to be careful to avoid situations of conflict of interest. There are plenty of contract engineering jobs where this would be a non issue.
I do medical electronic product design for a living, and a little bit of audio on the side, now entirely for fun. Unless you have very deep pockets you are not going to be able to defend your patents effectively against infringement, and that's assuming you have done thorough patent searches and also are not doing something that is considered established practice so that a patent is granted in the first place. Possibly treating it as a trade secret would be more effective protection.
It's hard to imagine there is someone out there in consumer/high end audio land who is going to make you a rich man. (It does happen, but usually to guys who are already well established/well known innovators in a given industry) Yes, I am extraordinarily cynical having entertained similar dreams for several decades of my life. I gave up and went to work for the man, doing so has gotten me a bit closer to my goals than otherwise...
The forum is really all about diy'ers sharing their hard won knowledge, ideas and experience with the other members.
If you have unusually good analog/mixed signal circuit design chops there is no reason for you to be working in a low wage/low skill job, although you do have to be careful to avoid situations of conflict of interest. There are plenty of contract engineering jobs where this would be a non issue.
I do medical electronic product design for a living, and a little bit of audio on the side, now entirely for fun. Unless you have very deep pockets you are not going to be able to defend your patents effectively against infringement, and that's assuming you have done thorough patent searches and also are not doing something that is considered established practice so that a patent is granted in the first place. Possibly treating it as a trade secret would be more effective protection.
It's hard to imagine there is someone out there in consumer/high end audio land who is going to make you a rich man. (It does happen, but usually to guys who are already well established/well known innovators in a given industry) Yes, I am extraordinarily cynical having entertained similar dreams for several decades of my life. I gave up and went to work for the man, doing so has gotten me a bit closer to my goals than otherwise...
It may depend on exactly what you mean by that. Only people at ESS know certain things about what is inside. What we do know is that is it uses resistor switching to produce the outputs, there are two outputs for each channel that operate in opposite phases, that the resistors effectively move the outputs in opposite directions bounded between AVCC voltage (+3.3v) and ground, that the movement in opposite directions can be in terms of voltages or currents with currents giving lower harmonic distortion, and that ESS recommends summing the outputs differentially after IV conversion if a single-ended output is desired (although they do have a recommended single ended output circuit in one of their application documents that ties one of the differential outputs to AVCC while the other output goes through an IV stage offset by AVCC).
So, how would anybody who doesn't know ESS's internal secrets know how to answer your question? How would anybody even know which of those things you might be asking about? Since the whole thing kind of seems rather confounding, my guess would be maybe that is why nobody has tried to answer. Sorry if that didn't help much, don't know what else to say.
Only the ESS designers can tell you how did they implement the DAC but does it really matter if the DAC can offer 130+dB THD and dynamic range? Or ask yourself, if the internal processing is single-ended instead of differential, how do you plan to change your design?
By the way, I am confused about your questions of the common mode distortion. The common mode distortion I have learned is about the common mode voltage of the output/input changes lead to differential mode changes. Since you are designing an I/V stage for the DAC, the common mode voltage of the DAC output should not change at all, then why do you care about the common mode distortion?
- Status
- Not open for further replies.