We should consider those boards as DIY too, something still needs to be finished, power supply, case, etc.
E.g. Eversolo DMP-A6 and Wiim Ultra are also DIY as the user needs to hook them up, configure network settings etc. 😉
Heck, even my car is DIY as I don't have a chauffeur.
Heck, even my car is DIY as I don't have a chauffeur.
Oh, computer audio. Interesting.
Let's just go back to inexpensive ready made/encase-it-yourself Chinese DACs.
Let's just go back to inexpensive ready made/encase-it-yourself Chinese DACs.
Last edited:
My money went on a dual PCM1794 board, no opaamp option. The current outputs drive my Sowter IV transformers directly. The SQ is impressive to my old ears.
I've never liked the ESS sound - initial impressions were good but then I found them a bit bright and clinical.
I did buy a Douk dual chip dac - switchable between Ess and Akm and cheap enough (£149) from Amazon but decided to return it and buy the 1794 board.
https://www.amazon.co.uk/dp/B0D83ZLTWG
https://www.aliexpress.com/item/1005004848385830.html
I've never liked the ESS sound - initial impressions were good but then I found them a bit bright and clinical.
I did buy a Douk dual chip dac - switchable between Ess and Akm and cheap enough (£149) from Amazon but decided to return it and buy the 1794 board.
https://www.amazon.co.uk/dp/B0D83ZLTWG
https://www.aliexpress.com/item/1005004848385830.html
Almost done TDA1387x8 DAC. Very cheap 5USD board, but need some DIY and extra money investment. Excellent for it's price.
Don't look at output caps, just find comfortable place to listen to different types, it is very easy to swipe. Now it is CDE BPA series (for audio coupling).
Don't look at output caps, just find comfortable place to listen to different types, it is very easy to swipe. Now it is CDE BPA series (for audio coupling).
Attachments
You are right. Some would rather reserve the term for those that mined the raw materials themselves. Legally though, final assemble is point of manufacture.
I found my device will connect to the qualcomm BT software, but it's all locked.
Buying these cheap boards might bag us factory designs, but components that need upgrading, or adding. I think already, that the ess dac is a bit thin by my standards, but I expected that. It's the ess sound people talk of. I have decided it probably isn't for me, in any implementation. It's now in the garage with old PC speakers, as a BT receiver. I'm happy I spent so little on this experience. I might have a go with the 56 dac
Last edited:
We did that extensively before with similar ES9038Q2M boards. Learned a lot about how low-cost Chinese dac boards are designed. Fixed up power supplies, output stage, clocking, I2C programming, pretty much everything except changing the dac chip. Found out the effects of each one of those things. Its all in the very long ES9038Q2M Board thread.
Out of all the people that tried the boards in that thread, there was only one person who found the unmodified board to sound acceptable (and I still wonder about that guy's hearing).
In the end I don't think anyone else except that one guy was very satisfied with the sound, no matter how much time, effort, and money was spent investigating. For myself, maybe I spent $1,000 or more first starting with simple mods, and eventually going to fairly exotic stuff (including finding out what altering MCLK phase in 200ps steps could do).
Not only that, but every time a new similar low-cost ES9038Q2M board came out, I tried that one too. Then more expensive versions in a finished case came out. People still held out hope a little better design was all that was needed. I bought a few of those before other people did just because I could afford the cost more than some of those other people. Then I did a circuit analysis, listening tests, tried mods, and reported the results in the thread so that other people wouldn't have to buy one to find out things. Only one person, me as it turned out, was needed to serve as the guinea pig.
Every single one of those dac products was awful sounding. Unlistenable.
In the end I reluctantly decided it was foolish not to admit the Chinese designers knew exactly what they were doing: producing cheap junk dacs because it was a way to make money selling to people looking to save some money on pretty close to decent dac.
Anyway, apparently memories are short around here. Or, maybe its that history repeats itself. Now we have some people optimistic that a fresh do-over may give better results. Very unlikely, but at least it can be a good way to spend a lot of money learning about how to design a better dac than low-cost Chinese boards with fancy SOA dac chips. Older, simpler chips, maybe not so bad. They don't need the surrounding circuitry and PCB to be all that well designed.
Last edited:
That makes no sense at all. You could get a pretty decent PCB for a multichannel '9038Pro for that. Well maybe not you but I'm pretty sure I could.I did say I tried some exotic things, some of which involved differential clocking in order to do the 200ps phase stepping. Also tried as close as I could get at the time to running AK4137 synchronously for conversion of PCM to DSD. It was a lot of stuff and it took quite a bit of time. Extended the ground plane of the dac PCB to make room for a better output stage, replaced the clock and powered it on a low noise regulator. Tried various film caps, some a little exotic. Tried some Iancanada products for his version of external clocking and I2S reclocking, and tried FIFO_Pi (the one product of that lineup I thought helpful at the time). Tried various approaches to powering AVCC. Also, tried various clocking, AVCC, power supply, and output stage designs from Twisted Pear (all junk too, but for different reasons). Tried published external clock board projects. Then there were multiple variations of the same Chinese dac board, there was the blue one the green one. There were some other ones too. I did mods to all of them to see if there was a best one to recommend to mod'ers and or experimenters. The final version I did was pretty complex and sounded slightly better than Benchmark DAC-3 in some ways but not all ways.
Part of the reason for trying so much off-the-shelf diy'er focused products was to see if that stuff was any good for making my dac better. I didn't want to reinvent the wheel until I saw if other people had already solved some or all the problems. Tried modifying a lot of those products to find why they didn't sound so good, and if they could reasonably be made to sound better.
Part of what kept me going was a little bit of friendly competitiveness with Allo and their ES9038Q2M Katana dac. The ES903xPRO dacs came later. From diyink, Aliexpress, etc. Sounded awful all over again.
A lot of the work was also to try to compete with Benchmark DAC-3 on SQ was towards the end when I got close enough for that to seem perhaps possible. IOW, it was to learn how to do it, not just to have another dac.
So, what I recommend today for people who want the best dac for the lowest cost is to go with the best Topping SMSL they can afford. That, or do a project with an older dac chip that is a lot easier to get good sound out of. The newer Sigma Delta dacs with SOA datasheet specs cost too much to make a one-off version that can compete with Topping or SMSL. If someone wants to diy a better sounding dac than those, then first thing is to learn how to SMD solder.
Part of the reason for trying so much off-the-shelf diy'er focused products was to see if that stuff was any good for making my dac better. I didn't want to reinvent the wheel until I saw if other people had already solved some or all the problems. Tried modifying a lot of those products to find why they didn't sound so good, and if they could reasonably be made to sound better.
Part of what kept me going was a little bit of friendly competitiveness with Allo and their ES9038Q2M Katana dac. The ES903xPRO dacs came later. From diyink, Aliexpress, etc. Sounded awful all over again.
A lot of the work was also to try to compete with Benchmark DAC-3 on SQ was towards the end when I got close enough for that to seem perhaps possible. IOW, it was to learn how to do it, not just to have another dac.
So, what I recommend today for people who want the best dac for the lowest cost is to go with the best Topping SMSL they can afford. That, or do a project with an older dac chip that is a lot easier to get good sound out of. The newer Sigma Delta dacs with SOA datasheet specs cost too much to make a one-off version that can compete with Topping or SMSL. If someone wants to diy a better sounding dac than those, then first thing is to learn how to SMD solder.
Last edited:
Got my pcm58 boards with adjustments for MSB and first bit , second and third could be modified added trimpots.
I need to install Optical imputs, some SM5813 converter to the more recent, not sure which one yet.... maybe the newer SM..
Then plug transformer/RCA out, and probably have a first listen tomorrow through USB upsampling.
Many say it is behind tda1541 and PCM63, I will see that soon, I bet I can make it sounding quite good...
I need to install Optical imputs, some SM5813 converter to the more recent, not sure which one yet.... maybe the newer SM..
Then plug transformer/RCA out, and probably have a first listen tomorrow through USB upsampling.
Many say it is behind tda1541 and PCM63, I will see that soon, I bet I can make it sounding quite good...
As said before I have not yet heard a single ES9038Q2M DAC that sounded real good regardless of price, implementation, brand etc. It is the IC itself that is not to write home about really. IMHO many a new ESS product was hyped up big time and usually later on turned out to be serving mainly instruments.
So I just ignored it altogether, that seemed more intelligent & efficient than pulling a dead horse. They may measure like they have been produced in heaven, they do not sound OK enough in my ears. Even in the field of SMSL one can quickly learn which one has AKM and which one has ESS chips. Apparently the ES9039Q2M breaks that spell. We'll see, if the chip is good it will also perform acceptable in not so good designs. Probably better than using so so ICs in perfect designs 🙂
Last edited:
To continue with car analogies if you put diesel in a Yugo or a Ferrari both will give you a very poor ride.
If it's like my PCM56 DAC, yes it's good. You can look forward to very good bass.
It is IMHO worth a try to have a good chip in a mediocre design as opposed to putting much money and effort in a good design with mediocre IC. No more no less. If only to have substance instead of words. We spend more words on the CheapESS than it has value 🙂
Glad I skipped putting much money, time and effort on ES9038Q2M (or its predecessors). Even quite some branded products with mid range or upper range ESS ICs sounded like the average Chinese DAC. Most were a poor ride indeed and a waste not worth listening to.
So that Yugo may be a good ride when given the right fuel.
Last edited:
I think that the role of the exactly DAC chip is greatly exaggerated by community. Most of the work plays source of I2S or SPDIF and if it is bad then no best DAC chip in the world will sound good.
Not so cheap (although they say it's "inexpensive " ) and not DAC precisely but a device of interest to me and sort of close enough to the subject of the thread. Anybody has experience with that USB bridge ?
Quloos QU02 USB Bridge
https://shenzhenaudio.com/products/...-i2s-supporting-dsd512?variant=43509039268083
Quloos QU02 USB Bridge
https://shenzhenaudio.com/products/...-i2s-supporting-dsd512?variant=43509039268083