[skibum]

Quote:

Originally Posted by marcelnl /img/forum/go_quote.gif Skibum; good to see things worked out, could you pls share what the problem was?

Well I had so many problems I am not sure which one you are talking about...

My last problem was my DAC's output distorting on some tracks. The reason for this was getting ripped CD's from dodgy places.... Once I started just ripping my own CD's using dbpoweramp on my PC or XLD on my Mac all my problems went away. I have a large collection of music that I did not personally rip. Most of it sounds horrrible. What I have learned is you can not trust anyone else to rip a CD properly.

 

[Bill Allen]

Skibum- glad to hear you had success with your very first Mod outing. Feels empowering doesn't it! Now you are free to tackle the rest of your system, one component at a time, imagine the possibilities! Keep it simple ...Keep it real.

Myself I modded everything "but" a DAC up until last August when my first Valab DAC came along. I knew virtually nothing about the inner working of a DAC, the whole thing was black art to me. However once I delved into the Valab I soon discovered most of the black art is actually inside the digital to analog chip itself and DAC designers can do virtually nothing about it. They are digital interface and analog output engineers, my goal as always is to paint them out of the picture. Modern DAC chips just give them lots of choices on how to implement the in's and out's. Think of these choices like the many inputs to the Chameleon DAC, IMO its best to focus on perfecting only the one. Multiple inputs only offer the chance for multiple issues.

As everyone knows the Phillips TDA1543 is a legacy chip from 1990. The beauty is in its large size, simplicity, and I2S input with no fancy options ... none. Perfect for the DIY guy to mess around with. As the chips became surplus in the mid 90's a man named Ryohei Kusunoki with some clever inspiration discovered how to implement post DAC filtering using only our ears. The NOS DAC was born.
TNT-Audio inter.views Kusunoki San [English]

Just like the SET amp, using current measurement technology, the NOS DAC shouldn't sound superior ... yet as we all have discovered ... they do. As the old adage goes."If it measures bad and sounds great ... you are measuring the wrong thing" I think of it like the current crop of 3D movies coming out, I have no idea how it works but boy is it more realistic. Ironic then that making the Chameleon an I2S input only DAC actually adds a complexity. I have found this theme to be present in all ultra simple audio applications, the last hurdles are always the highest. What's cool about the the M2Tech HiFace is it removes the high hurdles when using computer audio at a very attractive price. I myself would prefer if JKenny would offer a galvanic isolated HiFace with 5V I2S output and battery power supply. That way you could eliminate the entire main circuit board and its limitations. (By the way I have yet to hear a pop with the I2S direct connection.) What remains is figuring out how best to direct terminate the I2S wiring, its obviously that the current factory set up is also far from ideal. I still find it valid for the brazen modder to try I2S direct to the DAC board ... be it transport or HiFace. Who's with me?

Having said that I hooked up my fully modded Chameleon DAC with Spdif input and my partially modded Chameleon I2S and put them head to head. This was easy as the Pace Car can simultaneously output both I2S and Spdif. It was just a matter of selecting the source select on my Silverstone Ultimate TVC. While the differences were not large there where meaningful observations. First it's clear I need to finish installing the Blackgates & film bypass caps on all 16 DAC chips of the I2S version. It lacks global dynamics. However the Spdif version lacks what I call Valoomph, a word I have coined to describe hanging musical nuances with dynamic slam. To get crude it’s like taking a massive dump that comes out all at once with no need to wipe afterwards. Every note hangs in the air for just a millisecond longer and decays naturally into the ether. Subtle yet deluxe.

Could I live with the Spdif version as modded, yes you bet. Could the Spdif input sound better with simplification and better external clocks, yes I already proved that. Would there be lower jitter with a better clock on the upsampling board, yes Rhodes & Tony have already pioneered that trail. Will these mods reveal Valoomp in my system, you bet. Understand however there are levels of Valoomph ... will you still need to wipe after you mod. Depends!!!

The question is ... Choices.

 

[Newk Yuler]

Valoomph! From now on that's what I'm calling a massive... what you said. My ex called it "poopie nice". As opposed to poopie gross that needed a lot of wiping. How's that for digressing?

I'm not giving up on Chameleon so it's not really a lost sale, as if one sale was going to make much difference. I substituted for the time being after discovering a worthy alternative and its delightfully surprising features made me valoomph a brick. I expect the A-GD DAC19DSP to sound magnificent behind the Off Ramp. Meanwhile I'm going to let time pass and see what develops with Michael and the evolution of his TDA1543 designs.

I have to wonder how many Chameleons he's sold. I imagine the various versions of the original that became the Valab (and others) have sold a lot more. I wonder how he decides what's worthwhile to pursue given the development and cost considerations. It would be great to see Chameleon evolve like the Valab did, but what drives the evolution? In that respect, does Chameleon have what the Valab had?

Chameleon was designed with daughter cards for versatility, so they could be replaced with different daughter cards. That was my understanding, anyway. So what the heck will make Michael decide to utilize the versatility? There will be more buyer interest if the options are available at time of purchase, not just a few guys at HeadFi with dreams and hot solder irons.

 

[marcelnl]

I HAD the schematic, mind you that there may be differences we are not aware of but it should serve the purpose of finding the location for the incision

Could also do with a V3 schematic, had one but apparently misplaced it in my web mail after my recent HDD crash

Bill. I'm with you, if Philips could do it in the 80ies we can do it now

 

[Wood]

With this one simple word Bill has wiped out the subjective limitations from audiophile journalism, for ever. Valoomph gets to the true bottom of it all, as we arrive at a new passage in reviewing. Valoomph reaches deep within us all, regardless of sex, race or musical preferences as there is not one of us who has not experienced Valoomph in all its various manifestations.

Sorry Bill, just couldn’t resist, but I will certainly be listening out for Klingons on jkeny’s bespoke mod. Beam me up Scotty!



Newk Yuler, I guess because the Chameleon only recently went into mass production, Teradak does not yet have a large enough owner base to risk production of alternative daughter boards at this stage.

I notice they have been modifying some basic aspects of the Chameleon from feedback from early adopters. But those are things that will upgrade the quality of the stock Chameleon for all users.

The worry for them, as Michael has indicated, is that a stock simplified board will degrade the sound for users with average CDP as rhodes showed us with his scope pics.

Other input options like I2S have only, very recently, been discussed, kicked of by Bills bespoke mod. So we can not expect anything soon here.

We have a different time perspective, for the contributors to this thread, Chameleon has been kicking around for 4 months, but for the rest of the world it is, as it says on Teradaks site;

“New Chameleon DAC for mass production is ready to ship 3/29/2010”

Its frustrating, but a consequence of Teradak working on such small profit margins, it keeps the price down for us, but limits the options on initial release.

I’m certain as the owner base grows so will daughter board options.

 

[mmerrill99]

Quote:

Originally Posted by Bill Allen /img/forum/go_quote.gif .............. What's cool about the the M2Tech HiFace is it removes the high hurdles when using computer audio at a very attractive price. I myself would prefer if JKenny would offer a galvanic isolated HiFace with 5V I2S output and battery power supply. That way you could eliminate the entire main circuit board and its limitations. (By the way I have yet to hear a pop with the I2S direct connection.) What remains is figuring out how best to direct terminate the I2S wiring, its obviously that the current factory set up is also far from ideal. I still find it valid for the brazen modder to try I2S direct to the DAC board ... be it transport or HiFace. Who's with me? .................... The question is ... Choices.

One problem here is that I2S is a board level connection not a box to box connection. A problem that is being stated here is that the TDA1543 DAC uses 5V signalling & so the I2S would have to be 5V. This is not strictly correct - the 3.3V may well work into a 5V I2S receivers on the TDA1543 DAC. It probably wouldn't work the other way around, however (5V I2S into 3.3V DAC & would potentially fry the chip)

Can I suggest a couple of things for you "brazen modders" - there are two approaches to this & it might be that option one is a stepping stone to option 2.

Option 1 is you tap into the I2S output from the Tenor chip on the plug-in daughter board as I'm suggesting for the HiFace. You then don't need to consider level shifting the I2S signal form 3.3V to 5V (its' already done in the 74HC157 Mux chip which selects between SPDIF input & USB input). On the plus side you retain all the functionality of the device but it has many connections along the way to I2S feeds from the different parts of the device. This is the transmission line effects that Steve & Bill are talking about, I think. It may not sound as good as the "purist" approach of option 2 but who knows until it's tried. I don't believe in predicting how something will sound.

Option 2 is the favoured & already implemented configuration of Bill/Steve - you bypass all the input boards & tap I2S into the TDA1453 DAC board. This is the most direct & probably best sounding - by how much, I don't know. What you need to do in this approach is ensure your I2S transmission line effects are minimised or at least better than the board level ones. This is a challenge, particularly with box to box cabling. The best thing is you can remove the board from the Chameleon, as Bill has done, & test it separately without too much hacking & this can be put back together too.

Maybe the best approach would be for Michael to put a jumper-able I2S connection right at the I2S entry to the TDA1543 board thus isolating the DAC board from the input boards - the purists option?

Hope this helps!

Oh, yes, I meant to address Bill's suggestion - I haven't yet tested galvanic isolation & it's effect on the sound & I'm reluctant to limit the speed of a 24/192 transport until I'm sure that it's worth doing. 24/96 really does sound better, you know - this is realised when you get the resolution of the modified Hiface - see slim's comments about real 24/96 Vs upsampled 24/96

 

[Wood]

Re-clockers, the latest sacred cow of the hi-end audio world.

The latest offering.

Audiophilleo1 - advanced ultra low jitter USB to SPDIF transport

 

[skibum]

Quote:

Originally Posted by marcelnl /img/forum/go_quote.gif I HAD the schematic, mind you that there may be differences we are not aware of but it should serve the purpose of finding the location for the incision Could also do with a V3 schematic, had one but apparently misplaced it in my web mail after my recent HDD crash

If you lost the schematic where did you get it from? I asked Teradak and was told they do not have the 3.0 schematic. They did give me the 2.5 schematic but that does not have the SPDIF reclocker on it.

 

[marcelnl]

i did also get the 2.5 schematic...

 

[Bill Allen]

Sorry Skibum I never played with a 3.0 version nor have the schematic. Perhaps you have taken your 3.0 to its logical conclusion.

Speaking of LC's its "appears" that's where I have already taken the Chameleon DAC. There is nothing left but 16 DAC chips laid flat on a circuit board inside an aluminum case. Somebody stop me!

Jkeny offers up a logical conclusion on where to stop with implementing his current USB Hiface I2S interface ... unless he can (wants to) figure out a simple way to up the I2S output voltage to 5V. JK any chance at using something like this. 3.3V to 5V DC-DC converter. « Circuits@Home

Building a dedicated I2S output transport like 47Labs has been done at the DIY level but is costly and time consuming. With I2S termination and galvanic isolation from the noisy transport still a major issue here too.

Steve Nugent says the TDA1543 DAC chip is limited to 96kHz although the PDF brazenly state 192khz. When I play a 176.4 file I can hear the music but it crackles, that says to me that its possible with perhaps even more careful termination. Not that I care to pursue that avenue. My Chameleon I2S does play 96/24 file with aplomb, but so does the stock Chameleon via its USB or Spdif input for you ... Inquiring Minds.

If we logically set the limit of the TDA1543 chip to 96/24 then the current USB isolation chip would also work just fine. The question is is economically viable for Michael or Jkeny to pursue these tangents. As I said the last hurdles are always the highest.

After seeing and hearing the short lived Kondo Ongaku DAC with 16 DAC chips per channel brings to mind the premium edition of the Chameleon I would like to see. (Of coarse I would call it the Madman DAC.) One Chassis design with dual battery/supercap power supply and dual 16 chip DAC board. Isolated Hiface I2S USB interface with some sort of chip to separate the I2S L/R digital channels and increase the voltage/current to 5 Volts. Analog output section consisting of a pair of Z-foil I/V's & silver wound output transformers.

Ok I am done daydreaming.

 

[marcelnl]

Bill, you're dreaming some of my dreams, I'll not share the other ones.....

I'm dreaming of an I2S Valab or Chameleon board integrated in a decent CDP......I don not care too much about 96/24 or more as not too much software in that format is available anyway. I'd be perfectly happy with CDs repro that comes close to my TDA 1541 gold crown ...

 

[mmerrill99]

Quote:

Originally Posted by Bill Allen /img/forum/go_quote.gif Sorry Skibum I never played with a 3.0 version nor have the schematic. Perhaps you have taken your 3.0 to its logical conclusion. Speaking of LC's its "appears" that's where I have already taken the Chameleon DAC. There is nothing left but 16 DAC chips laid flat on a circuit board inside an aluminum case. Somebody stop me! Jkeny offers up a logical conclusion on where to stop with implementing his current USB Hiface I2S interface ... unless he can (wants to) figure out a simple way to up the I2S output voltage to 5V. JK any chance at using something like this. 3.3V to 5V DC-DC converter. « Circuits@Home

That won't work it's just a DC to DC converter. There are ways of doing this but as I said already, the level shifting may not be needed when feeding 3.3V I2S into a 5V DAC. However driving into 16 parallel 5V DACs may pose a problem - has Steve actually tried this & found it to be a problem?

Quote:

Building a dedicated I2S output transport like 47Labs has been done at the DIY level but is costly and time consuming. With I2S termination and galvanic isolation from the noisy transport still a major issue here too. Steve Nugent says the TDA1543 DAC chip is limited to 96kHz although the PDF brazenly state 192khz. When I play a 176.4 file I can hear the music but it crackles, that says to me that its possible with perhaps even more careful termination. Not that I care to pursue that avenue. My Chameleon I2S does play 96/24 file with aplomb, but so does the stock Chameleon via its USB or Spdif input for you ... Inquiring Minds. If we logically set the limit of the TDA1543 chip to 96/24 then the current USB isolation chip would also work just fine. The question is is economically viable for Michael or Jkeny to pursue these tangents. As I said the last hurdles are always the highest.

Why not get one of the Diyparadise ultravox & just stick it in the USB line in front of whatever transport you decide on - it's not strictly necessary to actually build the ADUM into an existing product to use it.

Quote:

After seeing and hearing the short lived Kondo Ongaku DAC with 16 DAC chips per channel brings to mind the premium edition of the Chameleon I would like to see. (Of coarse I would call it the Madman DAC.) One Chassis design with dual battery/supercap power supply and dual 16 chip DAC board. Isolated Hiface I2S USB interface with some sort of chip to separate the I2S L/R digital channels and increase the voltage/current to 5 Volts. Analog output section consisting of a pair of Z-foil I/V's & silver wound output transformers. Ok I am done daydreaming.

I've already done a balanced output from the Hiface into two DACs but can't say much more yet

Myabe this whets your appetite?

 

[skibum]

Quote:

Originally Posted by Bill Allen /img/forum/go_quote.gif Sorry Skibum I never played with a 3.0 version nor have the schematic. Perhaps you have taken your 3.0 to its logical conclusion. Steve Nugent says the TDA1543 DAC chip is limited to 96kHz although the PDF brazenly state 192khz. When I play a 176.4 file I can hear the music but it crackles, that says to me that its possible with perhaps even more careful termination. Not that I care to pursue that avenue. My Chameleon I2S does play 96/24 file with aplomb, but so does the stock Chameleon via its USB or Spdif input for you ... Inquiring Minds.

Have to say I am pretty content with my VALAB 3.0 even with the reclocked SPDIF input. My interest is also out of my curiosity about how reclocking circuitry works.

Why does the Chameleon play 96/24 were my VALAB will only get to 48Khz? Anything higher and lots of distortion, but I can hear the music. When you say "termination" what do you mean?

 

[taiphan]

Hi everyone..

firstly thank you your sharing your mods and ideals.. i've been following with keen interest over the last few months as i own both the original valab and the new chameleon.

can i please ask a couple of questions.

1) i used a Trend audio usb to Spdif converter then coxial into the chameleon.

The trend can be powered by usb or external 5v. my question is if i tap into the chameleon's 5v power line to power the Trend aswell, will this have an impact on the chameleon's performance or can the power supply handles it?

2)my Chameleon power supply unit (yellow wire mod) makes a buzzing noise under load (when the dac unit is plugged in), is that normal? the buzzing is audible if i am close enough to it. i emailed Kevin about it but he doesn't seem to care much.


thank you in advance.

ps. ordered the parts, can't wait to mod mine.

 

[Bill Allen]

Quote:

Originally Posted by jkeny /img/forum/go_quote.gif That won't work it's just a DC to DC converter. There are ways of doing this but as I said already, the level shifting may not be needed when feeding 3.3V I2S into a 5V DAC. However driving into 16 parallel 5V DACs may pose a problem - has Steve actually tried this & found it to be a problem?

Yes Steve had first tried the standard 3.3V output from the Pace Car, and while it worked, he found the sound quality lacking. Converting the Pace Car I2S output to 5V changed everything, I think its safe to say the Chameleon needs 5V for optimum operation.

Quote:

Originally Posted by jkeny /img/forum/go_quote.gif Why not get one of the Diyparadise ultravox & just stick it in the USB line in front of whatever transport you decide on - it's not strictly necessary to actually build the ADUM into an existing product to use it.

Agreed! But I for one like one stop shopping, are you going to offer an ADUM product?


Quote:

Originally Posted by jkeny /img/forum/go_quote.gif I've already done a balanced output from the Hiface into two DACs but can't say much more yet Myabe this whets your appetite?

Reeheeheally!

I will PM you.