[Priidik]

  I'm running an old avm 20. It has audiophile grade burr brown op-amps and dacs throughout. The system was well designed in layout and components which is why it received many awards and steep price of over $3000 back in 2002. 
 
For the money I would think there are a lot of better dacs out there. It seems this one has received a bit of hype. 

One source says it uses AK4832 dac chip (s). 
BB op-amps. Looks decent, and potentially very good psu side. 

 

[Anglomilian]

Can any of DA8 owners help with this query?

I want to buy this device for driving my Audeze EL-8 Closed. So the question is: will an amp of the da8 give a some background hiss on my hi impedance (30 ohms) audezes?

Thanks!

 

[earfonia]

Can any of DA8 owners help with this query?

I want to buy this device for driving my Audeze EL-8 Closed. So the question is: will an amp of the da8 give a some background hiss on my hi impedance (30 ohms) audezes?

Thanks!

For almost 2 years, so far I never heard hiss noise when using headphones. I've tried it with more than 15 or 20 headphones, including Hifiman planars, but never tried Audeze with DA8 or DA8ii.

 

[Zashoomin]

Can any of DA8 owners help with this query?

I want to buy this device for driving my Audeze EL-8 Closed. So the question is: will an amp of the da8 give a some background hiss on my hi impedance (30 ohms) audezes?

Thanks!

No, you won't hear anything.  At least shouldn't hear anything.  

 

[teiki arii]

 
Summary: The sound, for whatever reason is better on my old preamp, and I see no reason to keep this dac. The da8 will be heading off to classified heaven to recoup some costs, fortunately, I got it on sale and will therefore lose a bit less than if I paid full price.

Hello infinity88,
I have been using this Dac for 22 months now with Discret AZP OPA. I never had any problem with. My Yulong DA8 has nothing to do with the original one in terms of rendition of course. If you have got a problem of 3D imaging, it is not because of The ES9018 ship. What you must do with this Dac according to my experience:
 
- just use it as a "pure Dac", turn the preamp off! (except for using it as a headphone Amp of course)
- Don't use the "jitter eliminator" (choose "bypass" mode)
- Don't use the "Sharp" mode for not DSD files,
- Don't upsampler the files on computer,
- Don't use asymetric outputs, it's a real symetric Dac.
 
If you have got a Mac and Audirvana plus ver. 2.2.5 software, try iZotope converter with MBIT+ Dithering. But best results are reached with an analog preamp, unless this one is not so good. I have an Audio Analog Bellini and a UGS preamp (on an original idea of Nelson Pass and Flat Labs) very close to Pass Labs X0.2, very transparent, far better than the new XP20 IMHO... Please, try RAMDisk Manager and create virtual disks: you improve autonomy on Macbook Pro (at least twice even with SSD) and you avoid Disk access. At last but not least, choose "DSD over PCM standard 1.1" (capable of DSD256) which is now compatible and avoid "El Capitan"...
I still am not ready to change. If I had no Dac, I would go for Metrum Pavane Dac or TotalDac Dual D1...
Best regards.

 

[ririeiman]

  DA8 review:
 
For the past few weeks, I've had the experience of using a da8 in my chain of components. My chain goes from laptop to da8 via usb, then to a tube amp (integrated), and off to a very good pair of speakers. Originally, it was a cheap usb/spdif converter, to a digital preamp, and then to the same tube amp and speakers.
 
Out of shear intrigue, I sat positing the question of whether or not the system (as it was) could be improved. It already sounded amazing, quite balanced and fairly neutral throughout the frequency range, a tad on the dark side, which is the way I prefer it. Well, since money wasn't a concern at the time, I let curiosity get the best of me and I ordered a da8, one of the last few they claim they have in stock since the da8 ii was released. 
 
My first thoughts were that it's a well built unit, with a nice design and quality feel. I wasn't thrilled about the lack of remote control but sound quality is the forefront of my priority list and these things can be worked around in any case. 
Here are my thoughts on the sound:
 
Pros:
 
The sound itself is quite full, it has a good balance of frequencies, detail, music sounds alive and crisp, great bottom end, the output of the dac has plenty of gain (I don't have to turn up my amp as high as I did when using the digital preamp). 
This dac supports DSD, but since the industry doesn't...this isn't much of a consolation. Just try finding a DSD file online :/  
 
Cons:
 
The first thing I noticed when I fired up the da8 is my soundstage morphed from amazing 3D in my previous setup, to a unison kind of soundstage. The difference is astonishing, literally night and day. I did A/B comparisons on multiple music sources switching from the da8 to the onboard burr brown dac on my preamp. Where the vocals would previously be directly in front of me, or to the left or right (depending on the source) and slightly behind the speakers, and the acoustic guitar slightly to her right with the old dac, with the da8, the vocals and guitar now come from the left and right speakers only. The 3D imaging has been completely obliterated. 
 
 
It seems people here have had good experiences with this dac. I wonder though if anyone else has experienced the break up of 3D imaging, or if the imaging could possibly fix itself over time. Someone posted that it took nearly a year for their da8 to break in and sound better than the sound card in their computer. It's not worth it to wait for the sound to magically acquire sonic benefits that my system already has, and without the convenience of a remote. 
 
Summary: The sound, for whatever reason is better on my old preamp, and I see no reason to keep this dac. The da8 will be heading off to classified heaven to recoup some costs, fortunately, I got it on sale and will therefore lose a bit less than if I paid full price.

theres no way a DAC can destroy 3D image , it can be happened only if you have reverse polarity or u need to recalibrate your speaker position (this is the most important thing if u change your setup) . 

 

[teiki arii]

I'd like to add some thoughts about Yulong DA8 vs DA8II. According to the specfications of the ESS9018 DAC:
 
- Jitter is "less" induced by the audio crystal oscillator but
- Jitter is directly induced by the quality of the Vref
- Jitter is induced by the quality of the power supply and his schematic rather than thanks to an extraordinary oscillator.
- Jitter is directly induced by specifications of OPA
 
Besides, according to the difference of prices between the "old" oscillator and the new one ("Low phase noise Crystek CCHD-950-25-100"), I do not understand the difference of prices between DA8 and DA8 II Dacs. Actually, I am sure that the sonic difference between the two Dacs is very small, and exchanging the oscillator didn't make a significant change. It is all a matter of marketing, no more...
To improve seriously the sound of the Dac, you have to drasticly improve power supply, and so the Vref and so the OPA. What my friend have done in my DAC. So, don't throw your DA8 for DA8 II, you should be very disappointed! Nothing has really changed.
Best regards.
 
PS: comments are welcome.

 

[project86]

From what I hear it makes a small but worthwhile difference. And, at least in my area, the price is the same as the original model.  

 

[teiki arii]

Hi Project86,
You are right, the price was the same at the beginning between the two versions but dropped very quickly because of competition at less than 900US$ for the first version. According to me, Yulong implemented the new clock and reconsidered the price for marketing, no less. You should hear more differences (not to say more) with Classe A discrete OPAs as I have got.
 
Hera are the specifications of the Dicrete OPAs implemented in my DA8:
http://www.homecinema-fr.com/forum/post178070956.html#p178070956

 
 
To summarize between Discret OPAs and the famous LM49722 (try google for translation):

Bande passante 1Hz - 1MHz
DIY : courbe confondue avec le générateur
LME : coupe en suiveur avant 1MHz et perturbe le générateur (qui coupe aussi), coupe beaucoup plus tôt en gain par rapport au DIY
Vainqueur : DIY

Bande passante audio
DIY : Parfait
LME : Parfait
Vainqueur : aucun

Slew Rate
DIY : entre 36.8 et 38.6V/µs
LME : entre 21.1 et 24V/µs
Vainqueur : DIY

THD+N
DIY : -102dB suiveur, -101.7dB Gain
LME : -97.7dB suiveur, -102.2dB Gain
Vainqueur : DIY en suiveur (+4.3dB), LME en gain (+0.5dB, presque équivalent)

Noise Floor
DIY : -103.4dBu (4.89µV RMS) suiveur, -85.4dBu (39.0µV RMS) gain
LME : -99.0dBu (7.98µV RMS) suiveur, -80.7dBu (70.0µV RMS) gain
Vainqueur : DIY

Signal Noise Ratio
DIY : 115.8dB suiveur, 111.7dB gain
LME : 111.5dB suiveur, 107dB gain
Vainqueur : DIY

THD+N vs Frequency
DIY : Identique générateur en suiveur, moins bon de ~2dB jusqu'à 3kHz en gain, avec petite perte de 1dB supplémentaire entre 3 et 8kHz
LME : Moins bon (que générateur) de ~2.5dB en suiveur jusqu'à 7kHz, moins bon de ~2dB en gain
Vainqueur : équivalents

THD+N vs Amplitude
DIY : Moins bon (que générateur) de ~5dB jusqu'à 0dBu en suiveur. Ecrête à ~+12dBu en gain
LME : Moins bon (que générateur) de ~20dB (!) jusqu'à +5dBu. Ecrête à ~+15dBu en gain
Vainqueur : DIY pour la THD, LME pour l'écrêtage plus tardif

Consommation (AOP double)
DIY : 70mA constants
LME : entre 9 et 12mA selon signal
Vainqueur : LME
 

 

[teiki arii]

Design:

 - Paire différentielle d'entrée sur JFET faible bruit 2SK170 et charge active (miroir de courant) avec bipolaires faible bruit
 - Etage de gain et abaisseur d'impédance en pure classe A, transistors bipolaires faible bruit
 - Sources de courant à fort CMRR (une par étage) compensée en température (+0.6µA/°C) et compensées en tension d'alim (+6µA/V)
 - Découplage d'alim local de chaque étage par capa MLCC faible ESR
 - Filtrage d'alim global avec ferrite + capa "3-terminal" (capas de dernière technologie ultra faible ESR en HF)
 - Toutes résistance Vishay Dale CMF55
 - Seulement 4 transistors sur le trajet du signal
 - Circuit imprimé double face professionnel, cuivre 35µm plaqué or, vernis épargne rouge, dim. 26x40mm
 - Pinout compatible AOP simple et AOP double avec le même circuit imprimé
 - Soudure étain/argent
 
https://lh4.googleusercontent.com/-x_exaSo0sz0/UjQ1JpF5kxI/AAAAAAAAFTo/fteioXNgIT0/s500/IMG_8451.jpg
https://lh6.googleusercontent.com/-GfF83gT_RoA/UjQ1IeCoqmI/AAAAAAAAFTc/xjrfvcFxkfk/s500/IMG_8450.jpg
 
Charateristics:
 
Alimentation max : +/-23V
Tension de sortie max en fonction de la charge :
- 11.5V crête @ Rc = 1k ohms
- 8V crête @ Rc = 630 ohms
- améliorable par simple réglage du bias du dernier étage (dans la limite de la dissipation max)
Tension de sortie max en fonction de l'alim :
- 12.8V crête @ +/-15V
- 17.8V crête @ +/-20V
- 20V crête @ +/-23V
Slew-Rate : 38V/µs, amortissement rapide sans trainage
https://lh4.googleusercontent.com/-A6J_wvD0KXA/UhDvbSl20UI/AAAAAAAAFRQ/r-wEVleKonM/s674/SlewRate_G2.jpg
Offset : +/-0.5mV typique, ajustable par trimmer 25 tours.
Bande passante (suiveur) : de 1Hz à 1MHz dans +/-0.1dB (coupure a 750kHz avec un gain de 11 en non inverseur)
THD :
- inverseur gain unitaire : THD -100dB (limite absolue du système de mesure), SNR 136dB (dynamique max du système de mesure)
- inverseur gain 2.2 : THD -99.6dB, SNR 134dB.
- inverseur gain 10 : THD -94.5dB, SNR 133dB.
Stable sur charge capacitive.
Stable au gain unitaire.
Tous transistors appairés sur un outil conçu et développé spécialement pour ce projet. Transistors appairés sous leur Vce et Ic effectifs dans les AOPs

 

[project86]

Any time you have a replacement version come to market, there will be remaining stock of the original which must sell for a discount. People know the replacement is out there so of course they won't pay full price for the first version. That's normal. 
 
It's also normal for a MKII version to cost a bit more, due to various improvements. But in this case, at least in my region, Yulong has not increased the price. MKII sells for what the original always sold for. Which means we get an upgraded version for the same price as the original. That's a good thing right?
 
I'm sure Yulong could have gone crazy and made it even better but the price would surely not remain the same. 

 

[teiki arii]

  Any time you have a replacement version come to market, there will be remaining stock of the original which must sell for a discount. People know the replacement is out there so of course they won't pay full price for the first version. That's normal. 
 
It's also normal for a MKII version to cost a bit more, due to various improvements. But in this case, at least in my region, Yulong has not increased the price. MKII sells for what the original always sold for. Which means we get an upgraded version for the same price as the original. That's a good thing right?
 
I'm sure Yulong could have gone crazy and made it even better but the price would surely not remain the same. 

Yes, you are partially right since I paid it 900US$ 22 months ago before Yulong DA8 II came on the market. Tweaking Opas cost me 120US$ at that time, it actually means 1020US$. How much doest a clock cost? What have other improvements been made, what do you mean by "various improvements"? I can't see anything convincing on the new board... That's why I say updating clock seems to justify keeping original price but it's all marketing, don't you think so? 
My two cents: as long as there will be IC OPAs (good but not enough) instead of nice real class A OPAs in Delta-Sigma Dac, considering that power supply (Vref)  is beyond reproach, you can change the clock by another better one as much as you wish, I cannot call it a real improvement. Actually, changing the clock in ESS9018 architecture doesn't really matter according to the datasheet...
About "crazy", Yulong made the Yulong D8 but with the same OPAs.. I would be very interested by a D8 review compared to Yulong DA8/DA8II.
Best regards.

 

[fishyee]

  Yes, you are partially right since I paid it 900US$ 22 months ago before Yulong DA8 II came on the market. Tweaking Opas cost me 120US$ at that time, it actually means 1020US$. How much doest a clock cost? What have other improvements been made, what do you mean by "various improvements"? I can't see anything convincing on the new board... That's why I say updating clock seems to justify keeping original price but it's all marketing, don't you think so? 
My two cents: as long as there will be IC OPAs (good but not enough) instead of nice real class A OPAs in Delta-Sigma Dac, considering that power supply (Vref)  is beyond reproach, you can change the clock by another better one as much as you wish, I cannot call it a real improvement. Actually, changing the clock in ESS9018 architecture doesn't really matter according to the datasheet...
About "crazy", Yulong made the Yulong D8 but with the same OPAs.. I would be very interested by a D8 review compared to Yulong DA8/DA8II.
Best regards.

 
I wish Yulong had a discrete OPA option for the DA8.  And of course, tested and measured by Yulong to prove real world benefits.

 

[teiki arii]

Hi fishyee,
thank you very much your "thumb up". I actually do not look for polemics about this Dac which is honest "in its own juice". But, everytime there is an upgrade, I'm interested of what's going on, I am just curious. ESS9018 is a mixt of conventional 6 bits Dac and Delta-Sigma modulation and his peculiarity is that he is free-jitter clock reliant (I mean the clock has not to be a fantastic one, a <1ns one is sufficient!). The actual jitter is coming  from power supply and OPAs. The less is the residual noise, better is the Vref. Better is the Vref, better is the jitter. As The noise is twice less in AZP (dual-)OPA than in LME49722, I let you imagine the benefice of that kind of Opa in DA8.
As the Dac is a real symetric, I give you this advice: don't use it as a SE Dac since analog RCA outputs cannot compete to XLR outputs.
Amanero USB link is a very good one, but I do not understand why I prefer Optical Toslink Input on file less or egual to 96KHz. I have no explanation about it.
 
PS: I would say more: I would be interested in comparison between DA8/DA8II/D8 and my tweaked DA8 with class A discret AZP OPAs.

 

[fishyee]

Agreed!  I only do balanced XLR.