diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Digital Line Level (http://www.diyaudio.com/forums/digital-line-level/)
-   -   8x TDA1543 & WM8805 DAC on Taobao (on Ebay soon ? ) (http://www.diyaudio.com/forums/digital-line-level/212416-8x-tda1543-wm8805-dac-taobao-ebay-soon.html)

KlipschKid 10th May 2012 02:26 AM

8x TDA1543 & WM8805 DAC on Taobao (on Ebay soon ? )
 
Hi,

EDIT : VERY BAD PCB AND COMPONENTS. The traces and through-plating are weak and thin and lift off easily. Routing is poor. Although the seller said I should try 16x tda, this created a current draw that the through-hole plating could not handle and the connection burned after a few seconds. I have spent more time just trying to get it to work reliably than I have actually improving it. It sounded mushy and distorted when I got it and I have managed to get it to sound vastly better, but the reliability problems persist so I'm giving up. It just isn't worth the time, trouble or expense. Other problems are in the second post.

I have deleted the changes I made - I don't want to encourage anyone to waste their money or time on this DAC.

KlipschKid 11th May 2012 02:58 PM

Hi,

Just a quick little warning to anyone thinking of getting one of these to mod it. It works as it should but is no good for desoldering etc and the parts are old.

The through plating is abysmal and often lifts off/out. I can see two places where this has happened to the builder too. Inexplicably bad. In one place, he has added a bridge across source-gate of the 2sk30 because it was open circuit and not shorted to create a constant current source.

One transistor died - 2SA970 - on a switch off/on - and that was part of the psu feedback so the damn thing passed through the full voltage of 2.5V zener + TL431. This became obvious because the selector board leds all lit up and the psu LED dimmed.

Thankfully, it hasn't killed anything partly because I switched off very promptly. I am now bypassing one psu section.

Good quality eh ? Well, it's a great way to improve my fault finding skills ;-)

Cheers,

Tom

50AE 2nd January 2013 07:58 PM

Anyone else ordered it? I did and maybe I would regret, but I saw this thread too late.

Anyone made it work and sound good without complains? It's on the way to be shipped for me.

Maybe I should make it work and sell it?

KlipschKid 3rd January 2013 01:13 AM

My board was the lowest quality one I've ever bought so if your board is the same, I'd advise against any de-soldering and if you must, then be very careful and patient. Personally, after trying, I felt it wasn't worth my time and it's now in a parts bin. It was a waste of time and money. If I could, I'd turn back the clock and not even bother, so my advice would be to sell it. But if you just want to have some fun with it before you throw it away there are 3 things to look at
1) the ground plane - two psu's share the same ground and this can create a background mains hum. This affects everything on the board so even if you can't hear it, you can be sure it's making a mess of the digital signals
2) the mclk - lots of jitter from poor routing and a low quality tcxo
3) output stage - electrolytic caps, wrong psu voltage/resistor values... do some reading about tda1543 i/v.

I'm not willing to offer any help - sorry - as above, my advice is SELL IT. It's junk.

Instead, buy this, it's so much better :

http://item.taobao.com/item.htm?spm=..._u=blptqfae743

Or if you want to do some diy, try this :

http://www.ebay.co.uk/itm/ES9018-32b...item4d082e3b58

paul_peraic 10th January 2013 03:43 AM

Quote:

Originally Posted by KlipschKid (Post 3308789)
My board was the lowest quality one I've ever bought so if your board is the same, I'd advise against any de-soldering and if you must, then be very careful and patient. Personally, after trying, I felt it wasn't worth my time and it's now in a parts bin. It was a waste of time and money. If I could, I'd turn back the clock and not even bother, so my advice would be to sell it. But if you just want to have some fun with it before you throw it away there are 3 things to look at
1) the ground plane - two psu's share the same ground and this can create a background mains hum. This affects everything on the board so even if you can't hear it, you can be sure it's making a mess of the digital signals
2) the mclk - lots of jitter from poor routing and a low quality tcxo
3) output stage - electrolytic caps, wrong psu voltage/resistor values... do some reading about tda1543 i/v.

I'm not willing to offer any help - sorry - as above, my advice is SELL IT. It's junk.

Instead, buy this, it's so much better :

USB SPDIF PCM2706˫AD1955 384kHz ASRC Ƶ DAC ׼-Ա

Or if you want to do some diy, try this :

ES9018 32bit Audio DAC PCB kit | eBay

Good to see that Raindrop Hui is being exposed as the charlatan that she is .
I have had nothing but grief with this seller and have learnt the hard way so would advise to any Diy enthusiasts to steer well clear of Raindrop Hui products :mad:
Cut and paste schematics and poor quality material/design along with surplus old stock components and not to mention consistent DOA's will just not do no matter what the price .

50AE 15th January 2013 04:35 PM

Ok, it arrived to me. It works fine for now, didn't have issues with desoldering, I changed a few caps and the Riv resistor. No pads lifted or something.

The 3D imaging is special about this DAC. But what I don't like about it, it sounds a little harsher and it has much background noise. Maybe I will play with the PSU and the bypass cap. It seems from now that it will be modded.

Can't understand something about the Vref pot. The value of the left channel is always 200mV higher than from the right channel. Is it related to IC matching?

DF96 15th January 2013 05:09 PM

Probably. In an ideal world each 1543 chip would have its own Vref resistor. You could check that the output resistors are matched. With 8 chips each producing a couple of mA you would only need a 10 ohm mismatch to get 200mV difference.

You probably need lower values anyway. 330R Rout and 240R Rref are too big, even for 8V supply. I would say Rref about 120, and Rout about 150. Otherwise you will get clipping on peaks.

50AE 15th January 2013 06:05 PM

Ok, I made the Riv resistor 150 ohm and somehow, Vref stabilized to 2,00V left and 2,03V right. What voltage is optimal and how do I choose Vref, by ear, by measurement of THD?

I made it sound much better when throwing the electrolytic coupling cap and put Mallory 2,2uF

DF96 15th January 2013 09:48 PM

From memory, the ideal quiescent point is around 2.3V i.e. equal to Vref. However, this might clash with the requirement to avoid peak clipping. Most people use resistor values which are too high so the best they can achieve is symmetric clipping. With values in the correct range you can adjust between +ve peak clipping and -ve peak clipping - both shown by a rise in 2nd harmonic on peaks. Between these two you should find a region where 2nd is low. You really need a test CD with full amplitude sine waves and either an oscilloscope or distortion meter. You have very little chance of getting it right with a music CD or adjusting by ear.

50AE 15th January 2013 11:12 PM

Alright, I got it by measuring THD, though I was close by ear. The hot spot showed by my DIMM was exactly 2,65V, but I made i t 2,70V to be sure, because a slight derivation below 2,65V and it gets serious all harmonics distortion. Above 2,65V, the 2nd harmonic rose.

By ear, I was between 2,6 and 2,8V and above I could hear the sound getting progressively "clogged", maybe because of the 2nd harmonic.

Thanks for the tips! :)


All times are GMT. The time now is 05:52 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2