OK this is a DIY site mostly about stuff people build themselves. What we see here is critical questions concerning a ready made device of which no schematic is provided (normal AFAIK). As we can see this has some consequences as many questions arise. This will always occur when no schematic is around. And when it is around it will either be copied and end up as a new Ebay device and/or pages will be filled with even more critical questions, either out of curiousity or even jealousy. This a a kind of circle no one wants to be in.
I suggest to move this thread to the vendors forum or, better, to the Group Buy section and to make a Google sheet of buyers to keep track as it will be difficult to keep up with posts in this thread. Especially the tedious "I want one with 0.02 % resistors" (often followed by an "if" question so no clear yes/no situation) which apparently should be enough for the one offering something to note that person down and keep track of buyers himself which is practically hard if not impossible. It has already been mentioned by Søren that the DAC will be sold on websites so it will be a commercial product (a bit like the VSSA project we had on diyaudio.com). For those who are willing to be early adopters it is possible and an incredible opportunity to buy and be one of the first testers. See this as an advantage but IMO expecting to get replies on all questions is a bit too much. Maybe the first post can be updated by Søren so that all relevant technical info is there. Call it the datasheet of the product if you like. It would avoid many many questions. There should be no need for him to explain every detail of his design (which is in fact "fishing" for the schematic) as it is a ready made industrial produced module that can be obtained. A connection diagram would help as an attachment to the first post as people obviously already start to think the practical side of matters how to connect and encase the board. It should be clear that an R2R DAC so nicely executed starts a buzz in the audiophile community and suspense is in the air.
Just some tips, we have some experience ourselves with all the extra things offering a device to the public can bring. It should be fun for the guy offering the boards too (and not start to be a hassle). No more no less.
I suggest to move this thread to the vendors forum or, better, to the Group Buy section and to make a Google sheet of buyers to keep track as it will be difficult to keep up with posts in this thread. Especially the tedious "I want one with 0.02 % resistors" (often followed by an "if" question so no clear yes/no situation) which apparently should be enough for the one offering something to note that person down and keep track of buyers himself which is practically hard if not impossible. It has already been mentioned by Søren that the DAC will be sold on websites so it will be a commercial product (a bit like the VSSA project we had on diyaudio.com). For those who are willing to be early adopters it is possible and an incredible opportunity to buy and be one of the first testers. See this as an advantage but IMO expecting to get replies on all questions is a bit too much. Maybe the first post can be updated by Søren so that all relevant technical info is there. Call it the datasheet of the product if you like. It would avoid many many questions. There should be no need for him to explain every detail of his design (which is in fact "fishing" for the schematic) as it is a ready made industrial produced module that can be obtained. A connection diagram would help as an attachment to the first post as people obviously already start to think the practical side of matters how to connect and encase the board. It should be clear that an R2R DAC so nicely executed starts a buzz in the audiophile community and suspense is in the air.
Just some tips, we have some experience ourselves with all the extra things offering a device to the public can bring. It should be fun for the guy offering the boards too (and not start to be a hassle). No more no less.
Last edited:
I have a -4V reference that also sent though an inverter with 0.01% resistors generating the +4 reference, both are then filtered and buffered for each rail and channel, all with precision opamps.
Unbalance in that system will generate only increased even harmonic distortion. Just for testing, I created 0.5% unbalance, that resulted in:
2nd harmonic at -60 db
4th harmonic at -74 db
6th harmonic at -82 db
8th harmonic at -89 db
So if you want it sounds like a tube DAC, that's how to do it
IMHO you are absolutely right when you say that's a commercial product, so this is not the right place for it.
Since it was exposed in a diy thread, one expected a detailed discussion around the design, that's usually means sharing all the information about the whole project.
Also, Soekris ask to keep the discussion technical, exactly what some members are trying to do.
For example, I'm particularly interested in the projectual choices about the master clock of the DAC (crucial in a DAC, IMHO), so I expect detailed answers in such this thread.
Maybe, the schematic could not be essential, but if Soekris won't share all the design information, no discussion will be possible.
Therefore, better if he moves this project in the Group Buy section as you suggest.
Andrea
Last edited:
OK, Søren... I guess there has occurred a misunderstanding between what I intended to ask and you percieved being asked.
So, please allow me to try again, but keep in mind that you've probably forgotten more about digital circuitry than I've ever known (
):You made a decision (as I understood) mainly on cost-effectiveness to use less accurate R's for the LSB's in the ladder. But what I liked to know was: If you decided to use the same accuracy R's in the LSB's, would that be noticable in the SQ? Me mentioning an amount of money (in case of the 0.05% version) was not the main priority in that question, but only an indication to what extend the impact would be acceptable, only if SQ improved. An other reason for me to ask this question is with regard to the title of this tread ( "Reference DAC module" ).
Can you further explain (I'd like to learn) what the criteria were to make the split between MSB's and LSB's around the 13th bit? One could also imagine to make that split around the 16th bit (or the 10th bit for that matter)...
Thanks in advance
Last edited:
A kind reminder for alternative resistors at a higher price
Precision Power Resistors | Surface Mount WireWound Shunt & equivalents
Precision Power Resistors | Surface Mount WireWound Shunt & equivalents
When I posted I first looked around and saw some posting about commercial module products in "digital source" or "digital line levels", and as mine wasn't even a finished products yet, I posted in "digital line levels", mostly to gauge interest, to check if my design was good and hopefully get some technical feedback.
It's commercial in the view that I run a commercial business already designing, manufacturing and selling computer products.
It's diy in the view that it's a flexible module that I hope that people will use to make great sounding DAC's, customized to the individual needs and preferences.
If the masters of this site believe it belong elsewhere on the site, then I'll move it, probably to "vendors" as it's not a group buy.
OK, Søren... I guess there has occurred a misunderstanding between what I intended to ask and you percieved being asked.
So, please allow me to try again, but keep in mind that you've probably forgotten more about digital circuitry than I've ever known (
You made a decision (as I understood) mainly on cost-effectiveness to use less accurate R's for the LSB's in the ladder. But what I liked to know was: If you decided to use the same accuracy R's in the LSB's, would that be noticable in the SQ? Me mentioning an amount of money (in case of the 0.05% version) was not the main priority in that question, but only an indication to what extend the impact would be acceptable, only if SQ improved. An other reason for me to ask this question is with regard to the title of this tread ( "Reference DAC module" ).
Can you further explain (I'd like to learn) what the criteria were to make the split between MSB's and LSB's around the 13th bit? One could also imagine to make that split around the 16th bit (or the 10th bit for that matter)...
Thanks in advance
I don't believe that there is any effect to sound quality when using less precise resistors in the LSB's, and selecting the 14 MSB is more a question about gut feeling, something like still having 4 MSB's at the full precision when the volume control is at -60 db.
I also believe that using 0.01% resistors (beside in the reference) is not going to affect sound quality, although it will probably measure a little better.... That's why I was originally only planning for 0.05% and 0.02% versions.
Member
Joined 2006
Just keep doing what you are doing, there are only a few people who are at your level who probably know more and who will give you valuable advice, most don't even read the full thread, obviously...
If you ever mention a bigger/smaller number regarding whatever people will automatically want "the best", in your case 0.01%SMD resistors, no matter that this is just a number and there are probably more important parameters about it to consider. It doesn't matter if it might as well work with 1%, just mention the damn thing and there is no going back, so better make it 0.02% or 0.01%, just forget about the 0.05% and hopefully that will give you a better discount and the price can go down a bit.
The price is good for what you offer, there is enough DIY in it, at least for me (actually too much DIY) so just push the guys and give us a link to the shop.
If you ever mention a bigger/smaller number regarding whatever people will automatically want "the best", in your case 0.01%SMD resistors, no matter that this is just a number and there are probably more important parameters about it to consider. It doesn't matter if it might as well work with 1%, just mention the damn thing and there is no going back, so better make it 0.02% or 0.01%, just forget about the 0.05% and hopefully that will give you a better discount and the price can go down a bit.
The price is good for what you offer, there is enough DIY in it, at least for me (actually too much DIY) so just push the guys and give us a link to the shop.
Member
Joined 2006
Thank you for your answer, Søren. It's clear to me now. Also I needed some assurance on the possible effects of the several tolerances on the final SQ.
I'm looking forward to the product itself!
EDIT: and the first auditive evaluations, of-course...
Last edited:
Hi Søren, you know best if your product is good and if it measures and, probably more important, also sounds good. Even if members want things changed I don't see that happen as the hardware design is quite ready in your eyes (also distilled by reading your answers) except for the choice for resistor tolerance. BTW I fully endorse the opinion of Nikola Krivorov that offering choices in resistor tolerance will only lead to more wishes and more questions not to mention more work for you by keeping track of several versions to various buyers and also supporting all versions. The best DAC in the world should have the best resistors in the world
. That it seems a good product that will gauge interest you can already see by the enthousiasm of potential buyers. So it is indeed a commercial endeavour/tryout and no DIY. Nothing wrong with that, please don't get a negative impression because of these comments.The point being that technical feedback without schematic is hard and odd for a ready made module from a manufacturer. It should be either fish or flesh IMHO. You can read that people want things changed (the never ending story of output stages/opamps) as it is in a DIY thread while probably few things will be changed, if at all. I just wanted to give some tips to let this thread not derail into a "question-walhalla" and a feeling of participation in the final design which is kind of chaotic as no schematic will be given if I am not mistaking and participation is out of the question. Something that will give frustration on both sides in the end. That is why I advised on making a technical datasheet of the product concerning all technical merits and possibilities, a connection and drilling diagram and add it to the first post.
Last edited:
+1 Jean-Paul, you have distilled what should have been said right after post #1. The design is really not up to discussion. Søren is obviously an excellent engineer, knows exactly what he is doing, and needs no further input from anyone.
I for one cannot wait for the kits to be produced, and buy one (or two!) as soon as possible.
Roger
I for one cannot wait for the kits to be produced, and buy one (or two!) as soon as possible.
Roger
True. It strikes me he wants feedback on a, for the buyers, not yet existing product of which no schematic can be seen. A bit like discussing the road behaviour of a new Ferrari by looking at a picture of it. If technical feedback is wanted it would be better to discuss the design with engineers on equal level or disclose the schematic (a big nono if you ask me). He should not be insecure with such a nice design and skills like that. Besides that, we can not rely on anyone else than Søren and we only have 1 picture to judge the product so technical feedback is hard.
Now only if my technical skills were like my posting skills
.....seriously, it is by no means meant as negative comment. We made many mistakes with the GB of our humble DIY DAC. Well you learn from them isn't it ? It is now becoming silent but there were many months in a row that I was replying emails and support questions 7 evenings a week. The question is if I would have listened to warnings/tips beforehand ...
Last edited:
I think we should stop any non-technical discussion from now onward. It really does not provide any useful ingredients.
Soekris has been giving many valuable technical input and we really appreciate that, let it continue that way
BTW, is there any consideration to use LME49990 on the buffer stage? Some say it is a better choice.
I saw another similar implementation from the east. No GB and not a commercial project though.
Soekris has been giving many valuable technical input and we really appreciate that, let it continue that way
BTW, is there any consideration to use LME49990 on the buffer stage? Some say it is a better choice.
I saw another similar implementation from the east. No GB and not a commercial project though.
Attachments
Last edited:
