[amb]

Not to dispute you, tomb, but my Alien DAC conclusions were not drawn as a result of the RMAA tests. In fact I did my comparison listening tests before I ran RMAA, because it was easy and a logical thing to do. Also, nowhere did I state that distortions and noise floor were the be-all and end-all. They are simply "glimpses" into the overall performance of the device under test. And what a glimpse it is. In particular, there is one thing about the PCM2702 that is, by its low-cost nature, flawed. The integration of everything on one chip means that there is plenty of opportunity for digital noise to pollute the analog section. And the common digital and analog ground does nothing to alleviate this. This is demonstrated in the spikey and rough noise floor. There is a reason why higher end DACs have more isolation between the sections and in general treat these matters with a lot more care.

Don't get me wrong, I like the Alien DAC very much. I think it's a cool little DAC and it's got extraordinary price/performance. Its frequency response is nice and flat, its distortions are not bad, and it sounds pretty good. In fact it handily outperforms some other DACs many times its price. I've heard and measured far worse. But the Alien DAC is not a super high-end source (and at that price, one wouldn't expect it to be). One comparison with, say, a Benchmark DAC1 will reveal the difference. But then a DAC1 is ~$1000 and it better be good...

Btw my Alien DAC has BG caps too.

 

[tomb]

I agree - and at no time was I trying to compare the Alien to a high-end DAC. My ears tell me something else when comparing it to the Transit, however. That's all I was trying to convey.

I have listened to the Transit every single night for more than a year, but obviously, it may be my ears deluding me.

 

[Yen]

Hehe, I think it is time for a Finnish ancient wisdom: Opinions are like @ssholes -- everybody has one.

Especially here at head-fi DIY section, Alien has become kind of a legend, even superior than it actually is. It is great and cheap little DAC, but it has its limitations. I havent heard the Transit, but I have heard X-fi, and I think Alien compares quite well with the regular X-fi.

How much of a difference can it be, if film caps, such as Mundorfs, are used in positions CL/CR? And what about Muse, Blackgates etc.? I have some ZNs waiting, but havent enought time yet for building a new Alien...

And amb: what kind of caps you used in your Alien that was measured?

 

[amb]

Quote:

Originally Posted by Yen /img/forum/go_quote.gif Hehe, I think it is time for a Finnish ancient wisdom: Opinions are like @ssholes -- everybody has one.

LOL, it's not just Finnish... it's fairly universal around here too, except with like this: "Opinions are like @ssholes, everyone has one, but no one like to look at the other guy's".

Quote:

Especially here at head-fi DIY section, Alien has become kind of a legend, even superior than it actually is.

Agree!

Quote:

And amb: what kind of caps you used in your Alien that was measured?

The output coupling caps are Blackgate N.

 

[dizzyorange]

I did a little comparo of the Alien with a Silverstone EB01, a cheap USB DAC at the same price level here.

For what it's worth, I thought the Silverstone had a slight edge over the Alien. Both were significantly below what the EMU0404USB and a DAC-AH had to offer.

 

[thebob]

Quote:

Originally Posted by amb /img/forum/go_quote.gif LOL, it's not just Finnish... it's fairly universal around here too, except with like this: "Opinions are like @ssholes, everyone has one, but no one like to look at the other guy's". Agree! The output coupling caps are Blackgate N.

Do you think that component choice can improve the Alien DAC?

More Blackgates? Back to back? How would you hotrod the Alien DAC?

 

[MASantos]

Quote:

Originally Posted by thebob /img/forum/go_quote.gif Do you think that component choice can improve the Alien DAC? More Blackgates? Back to back? How would you hotrod the Alien DAC?

The only place where it makes sence to use Blackgates or other boutique caps in the Alien is in the CL/CR positions. For the other caps use eithe Panasonic FC or FC or NIchicon UPW, for C13 a OSCON cap may be used, but if you can hear the difference only your ears will tell. I would say that anything above a BG in the CL/CR position is a waste of money, film caps are too expensive for something like the Alien. If you consider using film caps, just go for a better design, such as the OPUS DAC.

 

[tomb]

Manuel knows the Alien as good as anybody, so I hope he doesn't mind if I slightly disagree:

1. There are some film caps worth trying that are still a long way from the price of an OPUS. For instance, from reports, it's possible that a couple of Vitamin Q's may improve the BG's slightly. VitQ's, when purchased in volume, can be had for about $2-$3 each.

If you use something like the Muse ES for CL/CR, then the Sonicap GEN II's are an ideal bypass cap at $3.30 each.

2. Upsize C2 to at least 1000uF. Also, I've been having this discussion in another thread, but IMHO, you can't do much better on the Alien than with Panasonic FM's. You can almost approach OsCon territory if you simply use a larger size FM. You are only limited to the board footprint diameters of the caps and 15mm height if you are using the metal Hammond.

3. C13 - this is the one that needs lowest ESR. Manuel is correct that an OsCon is perfect for this position. Alf specifies the range as 10uf-47uf. However, there are some of us who've put in a 330uf, 6.3V FM with good results - it's only 6mm diameter vs. 11.2mm high. The OsCon would still be better, but you can approach it with a reasonable alternative.

4. C23, C33 - there are some who've used up to 100uf for these caps. Strictly speaking, these are not classic power caps on the regulator. They come after the reg's output. In the case of big stuff, we'd want a fast tantalum in this position, but the Alien's Regs are working under different principles. The REG101 data sheet states that "Noise can be further reduced by carefully choosing an output capacitor, COUT {C23 and C33, respectively}. Best overall noise performance is achieved with very low (< 0.22μF) or very high (> 2.2μF) values of COUT." In the data sheet graphs, however, they show values of 10uF for Cout.

Alf lists from 33uf to 100uf for C23/C33, so he apparently went high with the sizes rather than low. If the logic of the data sheet follows, then you may gain greater noise reduction all the way to 100uf. So using 100uf for C23/C33 may be a good choice. I've used 47uf's and can't report any ill effects, but I don't know if that tells you much.

Anyway - those are some things you can do ...