Chameleon DAC listening and modifications
Nov 23, 2010 at 2:06 PM Post #1,113 of 1,158
I don't think it's appropriate to ask here, but I think it's the only place that I have a good chance of getting help.  My Chameleon DAC is having a power issue in that the power connector port on the DAC (where the umbilical cord connects, and also this is on the DAC only not the DC30W) is extremely loose and it wobbles in its place on the chassis, whereas everything else such as the RCA connectors are rock solid.  I didn't think it to be too much of a problem, but today I hooked up the DAC and it didn't power on.  The DC30W shows its on but the DAC doesn't, which I'm assuming is due to the wobbly power connector.  Can I commission anyone to look at it and make the appropriate repairs?  Seems like it should be a simple job but I've never soldered before, and I don't feel like starting out on a $500 DAC =\
 
Jul 18, 2011 at 5:03 PM Post #1,114 of 1,158
Don't know it this is an active thread anymore, but a question....: 
 
The chameleon boards --- as they are shipped these days---  have multiple pads for fitting replacement and / or bypass components.... and in the case of the output caps, various choices for variable size and position ...   it would seem.
 
 
As marked by the white arrows below, it would appear that any of the pads around the Wima output 10uf caps will connect to the leg of the wima you are bypassing.   So it would appear, anyway.    The two 'close by' terminals I've marked with arrows would be perfect for a vishay 1837 and its narrow base.
 
Anyone know if that's how it goes, whether you could just mount a bypass cap on the auxiliary pads and leave the wimas connected ?    Asking because that would simplify disassembly and reassembly quite a bit if you could just solder from above. 
 
While we're at it, if we are, do the boxed segments near the chip caps --right of photo--  indicate the same thing, i.e. a secondary position for a bypass component already traced to the original ?   (and which of the smd caps do they connect to ?)
 
Thanks for any responses ....  (here in this ghost town)......  
 
 

 
Jul 19, 2011 at 4:07 PM Post #1,116 of 1,158
Thanks very much.  Seems logical, just wanted some confirmation.  (btw I did read this whole thread maybe a year back before buying a Tdak Chameleon, but can't remember which parts are where at this point...  wish it was indexed or something...)
 
On the subject of the SMD caps, each 'station' looks like it comprises, left to right :  a 0.1uf smd cap, then a 10.0uf smd cap, then a set of contact pads.  (so I guess it's 32 total chip-caps as is)....    Do I understand correctly that you're saying the two smd caps are already connected in parallel, and the vacant set of pads adjacent is available for a third parallel or bypass component ?      (or is it just that only the 10.0uf, next to the open pads, has this option...  ) 
 
I suppose it's just that I've carefully looked, without removing board or chassis base-- under the board for the Chameleon and all 16 chips have that white thermal conductivity glop ensuring that they make contact with the bottom of the chassis--  I'm assuming as heatsink insurance.  
And if I don't have to disturb that with dissassembly and then try to re-establish it again --  I'm happy to do any new soldering from the top.
If that makes sense. 
 
thanks for your help
 
 
Jul 20, 2011 at 4:52 AM Post #1,117 of 1,158
The usual procedure is to remove the Wima caps and replace with your choice; I now use a pair of 6uF Russian KBG PIO caps outside the box (they are BIG) and Teflon bypass caps inside, but you can leave the Wimas; long burn in time for some caps.  Remove the large SMD caps and replace with Blackgates or Elna Silmic (easier to find).  Pull the chips before soldering near them, the thermal compound stays on them.  Or buy a small pot from Maplin.  Have fun.
 
Sep 18, 2011 at 7:58 AM Post #1,118 of 1,158
I am looking to purchase the Chameleon, can I ask a quick question please?
 
If the Chameleon is based on the 16 bit Philips TDA1543 chip how is it managing to handle 24 bit files from either the SPDIF / USB inputs?
 
Is it truncating the 24 bit files down to 16 bit resolution, therefore eliminating the eight extra bits, or is there some fancy magic going on in the DAC's design where the parallel processing is able to handle the full 24 bit resolution files being thrown at it?
 
Many thanks in advance of any responses 
 
Russ
 
Sep 18, 2011 at 10:30 AM Post #1,119 of 1,158


Quote:
Is it truncating the 24 bit files down to 16 bit resolution, therefore eliminating the eight extra bits, or is there some fancy magic going on in the DAC's design where the parallel processing is able to handle the full 24 bit resolution files being thrown at it?
 



I don't have one of these DACs so I'm speaking just as someone who's read the thread and also designs DACs myself. The short answer is - truncation. Fancy magic would require a DSP or FPGA to do some signal processing and this DAC doesn't have one. To obtain better than 16bits from a bunch of paralleled DACs does require some fairly fancy footwork in terms of matching the DACs. If this were done precisely then 16 16 bit DACs could in theory result in the same performance as a single 20 bit one. In practice that's almost impossible because the matching between DACs doesn't get close enough. 24bit performance is in the realm of fantasy.
 
Sep 18, 2011 at 3:34 PM Post #1,120 of 1,158
Many thanks for the reply appreciated.
 
So if I am reading correctly a "true" 24 bit DAC then is not currently available, or is available at significant cost?
 
Why I asking is I am also looking at other DACS like the Audio-DG and other units at the same price range and ideally would like to have a DAC that is truely a 24 bit device that doesn't truncate.
 
Thanks again,

Russ
 
Sep 18, 2011 at 9:51 PM Post #1,121 of 1,158
If by 'true 24 bit' you mean something with 24 bit specifications (noise floor better than -140dB, distortion below that) then no, none is available to my knowledge. But then such a DAC would be way better than almost any imaginable pre and power amp downstream of it. There are DAC chips which accept 24bits as input, don't truncate but deliver considerably lower performance than this, they have some value. PCM1704 is one such but I understand its becoming less freely available. Other DACs presumably don't truncate from the input but (I'm guessing from the datasheets) have internal signal processing which does. Such as PCM1792 and AD1955. Probably the ESS DACs are in the same category.
 
May I ask - what advantage do you see from accepting all 24bits ?
 
Sep 19, 2011 at 2:41 AM Post #1,122 of 1,158
It's not an "advantage" per se...
 
More thinking along the lines of what's the point of having files at 24 bit resolution like SACD rips, DVD-A rips, HDTracks downloads, etc and then streaming them into a device that then reduces this resolution down to 16 bits?
 
It defies the whole point in having a 24 bit resolution files in the first instance, you might as well just stick with a 16 bit resolution file format if there is nothing in the DAC domain that can handle this resolution successfully?
 
(understandably at this moment in time)
 
I believe that some of the Audio-GD products do use the PCM1704 chips and this may further push me into looking in that direction as opposed to mod friendly Chameleon.
 
Russ
 
Sep 19, 2011 at 7:34 AM Post #1,123 of 1,158


Quote:
More thinking along the lines of what's the point of having files at 24 bit resolution like SACD rips, DVD-A rips, HDTracks downloads, etc and then streaming them into a device that then reduces this resolution down to 16 bits?
 
It defies the whole point in having a 24 bit resolution files in the first instance, you might as well just stick with a 16 bit resolution file format if there is nothing in the DAC domain that can handle this resolution successfully?

 
From a technology point of view I myself don't see any point. 16bits vs 24bits according to digital audio theory only gives us a noise floor reduction and no other benefit whatsoever. Whilst I don't think the theory's infallible, its the best formal treatment of the recording/replay process we've got and so until its falsified, I'm sticking with it. Seeing as I find almost all recordings offer me a higher noise level than my 16 bit DAC (the hiss increases noticeably when the music starts playing out) I have no reason to upgrade from 16bits. If you're a marketer though I can see plenty of money to be made from so called 'hi-res' formats. You've asked a very valid question that's perhaps best directed in the vicinity of the purveyors of hi-res (as opposed to this terminally skeptical engineer 
wink_face.gif
).
 
 
 
Jan 10, 2012 at 12:03 PM Post #1,124 of 1,158
Hi guys!
I've recently ordered one with stapped potentiometer and being thinking about parts upgrade. I've started to read this thread, but it's endless :) If I understand correctly, the most significant improvement will be if I change SMD caps to better audio grade caps. Am I right? What that caps for?
 
Jan 10, 2012 at 5:39 PM Post #1,125 of 1,158
hi gents just a short input from a beginner whos happened to pick up a relevant info. 
 
it's general consensus that bg caps do not like to be bypassed with anything else since they themselves already behave like non electrolyte cap.
so if you have them bypassed, you might notice an improvement if you take the bypass out. 
 

Users who are viewing this thread

Back
Top