Needing help about chineese clone DAC

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,

I need please little help about Chineese DACs' IC and the best DAC I/V way ?

I read many threads here, in particular about AD1865/62, TDA1541/43, PCM56/63 DACS because chips without internal aop seem to sing best.

I have a Raindrop'hui AD1865 from China (because of the low price and low skill hability :eek:) with a discrete aop design with BC560 and BC550. DAC dection is I think a copy from Andrea Ciufolli DAC-END and the work of some DIYers, idem for the DAC I/V stage.

Could I have to source a genuine AD1865 N-J to improve the sound because there is a possibilitie that the AD 1865 N-K from the board is not ?

WHat about the caps, are they fake in some chineese IC ? For example the plastic body of nipon KZ-muze have not the same look on tje IC that ones I buy from a serious local dealer ?

AD1865, TDA1541 in China are they NOS or rebuild with low quality or without the originals specs or totaly sourced in the same locals factory ?

For DAC I/V conversion, I can't understand in the forum if discrete is superior to the best OPA, or if tubes is a best way ? (not for the scopes but for the ears;))

Sorry for these simple questions, i'm a newbie with just poor skill (just little knowledge with powersupply and testing aops in a poor "lego" way). Thnaks in advance if anybody have a sure opinion about Raindrop'hui products and chips).

Eldam
 
I have some experience with raindrop_hui's DAC boards - not the AD1865 one though. I doubt very much you have a fake chip - problems with the sound quality I've found with these boards are due to layout errors. In particular grounding. The most recent one I bought was 4 * TDA1543 and it mixed up digital and analog grounds with shared impedances which muddled the sound. After quite a few cuts and straps on the grounds it sounded rather good.

My advice - get the layout right first before playing about with different I/V stages.
 
Hi Abraxalito,

Thank you for you reply.

I didn't think about layout mistakes... It's worse that I thought but I'm not surpised : it's difficult to follow the shematic given by Hui and the pcb is black ! I bought it after reading lampizator but the sound is not bad... but not very good too. I feed it with a squeezbox duet, I bypas the SPDIF buffer an the little transformer to connect on the legs 9 & 10 according Lampizator advices : it's better but not really better than a simple standalone Philips CD player with just few changes. Maybe I have to change for genuine but very expensive Quangh hos' DAC End PCB !

But the problems for newbie like me is cms soldering without spend a lot of money in tools and parts for an aproximative result. Easier to work with low temperature and almost already pcb's.
I'll try first to work on the pcb I'have, thank to you again for you reply.
 
You might just possibly be interested in a thread I'm running elsewhere - I'm building a NOS DAC based on the DAC-AH as a starting point. You can buy that for about $140 on eBay then modify it by following my instructions. Or you can buy just the board for around $100 and add your own case and trafo. Its going to use passive I/V with a very steep passive filter before the output stage - this I think will give it a valve-like sound, although its not built yet so this is just conjecture. Anyway the thread is here: Digital that sounds like analog
 
Thank you, I've to read it, it seems that is I'm looking for: just few tweaks or minimal work to doesn't break chips with high heat (400°).

I just finish the thread about the ultimate NOS DAC with TDA 1541 (and TDA1543): very interesting...for the part I understand, but for high level DIYers (knowledge & skill).

I assume "Good Sound of tubes" is clear sound without harchness and a big soundstage & some liquid without too much smooth but material... maybe I try one day tubes like 6n16 (miniatures) but afraid of high voltage !

I like too sometimes transistors schems because some of them have are smooth, kindly fat sound, not too transparent and good bass like one inegrated amp of thr 80's I have : Pioneer SA 5300 (4x toshiba PNP c1061, Elna caps bypassing speakers out)
 
Yes I've come to that conclusion about the sound of tubes too, despite not having any of my own to listen to :) Like you afraid of high voltage! Transistors are getting me such sound too, after a lot of effort in controlling noise. Without adequate HF noise control, transistors grey out the sound and diminish the soundstage depth. I surmise that valves are much more linear at higher freqs and this results in less noise modulation in the audio band.
 
"I surmise that valves are much more linear at higher freqs and this results in less noise modulation in the audio band"

Hi,

These things are not clear for me because I am not an engineer. I remember some old readings in the french review: L'Audiophile (Jean Hiraga), about the thermal distorsion and its control with silicon junction transistor or something like that. The man (Hephaistos) who wrote about it was the father of Lavardin IT amplifier whoch have a specific sound signature, not bad but not a revolution to.

But for listen hifi for 20 years (tube, vynil, transistor, CD) it's still difficult to make a hierarchy between technologies and mix of them in a audio 's chain reproduction. Like the old and now finish discussion about low or high db efficienty in speakers...

Regarless the price, it seems difficult to me to understand if discret design is better (and for what?) than aop, or tubes for low level signals after curent DAC like AD1865. But maybe an idiot question ? discret like transistor like SK30, SC170 are better here because of higher current use like tube, but aop's like BB627 AD797 seems to have lower distorsion...

Humm, I stop now, have to read big more thread to understand (not to design:D of course). I begin with your thread.

I think integration in the room with choices or design of filters (ISO curve,... ) is a part of solution : but need micro to check the listening room.
Understand you work in embeded electronic : we are all waiting for such solution in ours cars. I read some things about dithering than make the things more difficult to understand for the choice of a NOS 16 bits DAC in relation to 20, 24 or 32 bits (these last are not really more better if I read well the posts) : does it make sense to have in 21e century a truncate signal with a 16 NOS DAC. I doesn't understand yet, have to read more again.

I've to really stop know to not polute DIYA with my general considerations.

Thank you Abraxalito for you replies, maybe I will come back with more acurate questions.
cheers
 
Interesting point about going to more than 16bits - I rather ask the other question : does it really make sense to record 24bits when at least 4 and most probably up to 8 of those bits are noise? I see some value in a DAC with a better than 16bit noise floor, but not very much given the noise modulation on many of today's recordings.
 
Hi Eldam,
I bought a couple of moths ago SMSL SD1955 DAC (usd 80 in argentina), from a chineese supplier/manufacturer and compared it with my DAC Magic (Cambridge Audio usd 650 in argentina). It is not same quality, Dac magic is superior, but I tell you that SMSL for usd 80 is a great purchase. I listened to "very hard" jazz pieces and there were just moments that I noticed notable differences. From my experience I think china is working well, and improving fast too.
 
Hi Qunoaudio.

Thank-you for you reply.

Yes, hard to say if it's a problem of layout design (according Abraxalito) or chip quality.

I found by chance a Akai cd a-335 player in public garbadge. Just a mecanichal problem that I solve. It has a BB PCM63P - T. with little transformer, 2 aops low noise to finish with 2sk170 & it's complementary.
It' work greater than the Raindrop'hui clone of Andrea Ciuffolli. Pretty good for a zero cash investment:D.

I think the raindrop'hui has a problem : if it's the layout it will be difficult to solve, if it's the AD1865 a-k : i can buy maybe a N-J version or source a anaother A-K grade on an ebay's adress i saw in a groupbuy here which seems to be safe (used chips) : but it's more than 50 euros just for a doubt... it's not DIY friendly.... it's just doubt friendly:eek:


It's very difficult to be objectiv because cd-player i I2S feeded and my DAC is S/PDIF feeded by a squeezebox duet in direct on the CS8416 (pin 9+10). One CD-DVD-SACD philips 5005 players heavily tweaked (lundhal transformer, big independant powersupply for digital & analog stage) with a little AK dac chip wich is musical & something like... neutral (if it's word has sense). I'have another old Philips CD players sofetly tweaked (TDA1543/1545/1541), one pair of Lynfield 400 L, one pair of Kef 104/2 (that I love even if it's not transparent), one pair of Proac D15, all wired with Oyaide across 750 for asymetric modulation & idem with the most famous american brand between speakers and amplifier (Chord SPM 1000 B, or NAD 2100 Pro tweaked or the last full digitak NX320 from HifimyDIY feeded with my squeezbox): 20 years of financial effort for me. In 20 years few can answer about my question (if it has sense) about the best way to go between the DAC chip & the chinch plugs. Maybe the question is too general, I have to be more accurate, but laking of knowledge in relation to the excelent technical level of DIYA...
I'll have a look at your chineese DAC.
thanks
cheers
 
Hi Abraxalito,

A passive I/V with something near 1k homs like Bernardh and a LC filter ?

why not, just for pleasure because now I can't live without a squeezebox duet and a DAC.

the bass fall very too low (boomy, but upside: a very good something like 80-150 Hertz), need maybe to cut higher in the low-end near 20-30htz.
Treble are neutral but a little too metalic, easy too solve with another caps without tweaking the analog powerstage.
I'm surprise of the already good sound that this old and not expensive (low-midle position in the Akai range?) CD player have ! Happy to find it and save from destruction. It's not the first time, I found an old amplifier pioneer SA-610 : good look but not good sound (the olders sound better). Know i have a look when each time i throw wastes in public garbage... maybe one day an Accuphase... so what, it's christmass !:hohoho:

Abraxalito : is a 20 bits DAC like AD1862 a better way than a 16 one in relation to how is recorded Cds ?

Anybody benchmark I/V stage behind the same DAC between Tubes, discrete and aop ? Which sound better ? Better is for me : more natural, less agressive but not smooth. Is it a bad question, does-it depend of the I DAC used : for example passive better with a BB PCM63, passive I/V with resistor better for AD1865, tubes for TDA1541??? Want to understand because I'm LOst season 7 between all these solutions and no time, money enough to test all of them !
 
No, I wouldn't go as high as 1k on any passive I/V (except maybe TDA1543 but my days with that chip are behind me :D). The LC filter I have up on here and over at WBF is designed to work with 100R source and load. I've never listened to AD1862, I did play with a sample on the bench many, many years ago and it looked a bit glitchy on the 'scope. But subjectively it gets good reports. I doubt that its better than a 16bit part on redbook material. The architecture of the DAC is what I believe makes the greatest difference - which is why TDA1541A is so revered, because its not R2R and hence not very glitchy.

My theory, for what its worth, is that its glitchiness that makes the biggest difference to the SQ of a DAC, hence TDA1541A's superiority. The more glitchy the DAC, the more strain put on the subsequent stage - so perhaps opamp with TDA1541A could match valves and PCM1704 in terms of SQ. The LC filter works its magic I think because it takes out a lot of the gliches without introducing its own distortion, meaning that the downstream active stage has an easier time. You need maximum HF linearity in the active stage, which I believe is provided by valves. That is until I explore results with some extremely HF transistors I currently have on order ;)
 
Interesting but very more complicated that I thought...:rolleyes:.

when you design your filter I don't understand if you have the same cut curve for different load : 20khoms, 50K homs or 100 K homs amplifier or do you talk about amplifier stage just behind the DAC ?. Most of source from DAC are superior than 100R ? Is it not very low ? Maybe I don't understand what you wrote:wchair:. Is it a complicated filter ? first order ?
About glitch, I understand know why some people try to make discrete DAC from scratch.
I have 3 old breaked philips cd player with tda 1541 (and a Marantz cd-40 with 14 bits TDA1540)... maybe i have to try 1541 DAC board with these genuine parts.:confused: Don't like pcm1704 i heard... maybe because of the analog stages even with tubes. Clinical is not musical:sax:
 
I'm just here talking about the dedicated amp stage right after the DAC chip, not something in the next box. We need to have a low value resistor to do the I/V conversion because many DACs do not have much output compliance, so its normally best to keep the swing under a few hundred mV, depending on the DAC chip. TDA1545 has a fair compliance but I don't use much of it. TDA1387 has even more.

Yes the filter is fairly complicated, the schematic is shown on that thread I linked to earlier - in single ended it has 7 inductors and 3 caps. In differential (which I'm using now) it has 11 inductors, all hand-wound :eek:

Making discrete DAC from scratch I believe is the recipe for more glitching, not less. At least parasitic capacitances are small when all on one small die. So I'm not going there, I'll buy ready made DAC chips thank you :D If you didn't like PCM1704 I'd guess its because the amp stage straight after it was upset by the glitching, they didn't use a passive filter perhaps? I don't go for PCM1704 either, that's mainly because of the price - I use multiple chips in my designs so it would get jolly expensive really, really quickly :)
 
It's make me think about a german brand which use 2 dac in their CD players : one for bass region , one for the rest of the audio band.

Maybe I have to risk my life with tubes:eek: in front of a passive I/V (heard positives talks about the little russian 6n16) with good resistor, takman or whatever. It seems to be the simpliest way for people like me. Not to much defendor here for a "classic low noise resistors" discrete I/V. But I like the easy looking of these discrete active stage like the Raindrop'hui DAC I have... easy looking but easier sounding ? I feel free of the way here, unresrtainedly of model like full transistor but limited money and knowledge. I can't design but I can soldering (no CMS but good old chip) on an IC. Or maybe I deal with the dust on my Rega Planar 3... like it but prefer gaz for cooking that silex, not sure it's better to look behind. One hand with red wine and one hand with squuezebox remote (work greater with a smartphone for remote) is a better way for modern gentlemen.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.