Chameleon DAC listening and modifications

I received this DAC a few weeks ago and am posting to share
some early experiences with it. Mine is a pre commercial production
model but is a fair shot at what i think the commercial model will sound like.

This next generation of the 8x NOS DAC this time using 16 DACs, an input Pulse
transformer for use with the Coax and Toslink inputs, and an upsampling / de-jitter
board. It also comes with the DC30 PSU. I am obviously not fair in naming these
components and others are far better at nomenclature than I am but here we go ...



After ~200 hours fo burnin the sound is somewhat similar to the previous 8x NOS DAC
in but with far more refinement. By refinement I mean more inner detail and smooth
presentation of across the board detail.

This detail makes itself more felt in the midrange. There is a vibrant sound overall
with a good bass presentation. Modding does improve the output.

I change the output caps to 1uF Obbligatos bypassed with .015uF Teflon and the bass
became more enjoyable, the midrange fleshed up, and the topend took on some sparkle
and shimmer.

My machine suffers from a bit of noise from the input board and a new one is on the
way. Michael is again very prompt with support and all his efforts honestly seem to
go to trying to get the best sound out of a DAC at a very reasonable price.

Here is a view of my modifications so far:

I have one on the way. The main reason being that although the original Valab is superb value, as the quality of my upgrades improved I began to reach its limitations.

I am hoping there will be improvements in the noise floor,doubling the TDA like this should push it down some 3db, and increase the resolution, inching the sound closer to the ideal full 16 bit. Loss of detail and higher noise floor are generic for this type of NOS dac, the Valab goes some way to overcome this, so perhaps more of the same might be better still?

This is theory, but it seems that your description of differences in the sound pat, 'increased refinement and detail' seem to indicate this is the case.

Anyway I've pulled the trigger.

Apparently Michael says the optimum iv value should be 175r, not a lot of difference to 180r but there we have it. I anyway ordered 2x 175r Texas TX2575, and have some 180r in other types on the way to do a mini compo.

Very nice start!!

Like Pat, I have one of the pre-production units and as mentioned before out of the box it sounded quite good when I first compared it with my modded valab V2.1 with DC-30W in tow. My first impression when I first turned it on was it has more sound dynamic and is quieter.
I waited about 72 hours before opening the covers and placed the DAC on the bench - I simply couldn't wait for 200 hour burning time seemed like infinity...

Since I was already familiar with what mods affected what sound in the previous DAC, I was more tactical this time around. Below, is a list of changes I've made and I tried to remember the effects it had to the sound, again your observations may be different from mine. One thing I can say is with my current mod I can listen to this Chameleon for 4+ hours without fatigue which tells me I have got the thing in the right direction.

List of mods:
1] replaced 10uF WIMA with 3.3 uF Mundorf Supreme output cap (couldn't past wife with Supreme silver oil which is what I really wanted). I plan to couple with Rusian teflon later.
2] replaced output 0.01uF silver micas with Vishay MKP 1837 of same value. This mod brought out more upper frequency acuracy without fatigue. "I am hearing things I never heard before" I suggest that you first listen without the silver micas and then listen with MKPs in. You will want to keep the MKPs in or at least I have.
3] changed I/V resistor to around 172R (two 330 in parallel after manual matching). Stock is 160R, but PCB screen shows 180R. Michael confirmed that 175R is ideal, but hard to get. I first changed it to 180R and did hear occasional clipping.
4] I currently have the 7.6V for the DAC chips set to 7.85V and did notice more bass coming out of the DAC. Be carefull with this one because TD1543s are only rated at 8V.

With these mods, I now have transparency, depth, quietness in my music. I think the road to nirvana just changed gears

PS:My previous group-buy for this DAC is dead when the thread disappered.... Long story, please don't ask on this thread.

Norm

I seem to have missed this one, cab't find it at Ebay either.....anyone care to share the price tag ?

(I'm a bit miffed as this one seems a better modding platform)

hifi2001, see that you just could not resist eh, just had to open that lid!

Bt do have any plans on the 10uf TDA coupling caps?

Quote:



Funny you ask, every time I am looking at the DAC board I see all those empty holes beside TD1543 that TeraDak have already set aside for possible 10uF cap mod.... Based on the 8 x TD1543 board, I think the options are 10uF BG, or Muse caps, 0.1uF Vishay MKP caps, or leave it alone... I am looking for suggestions

Being the lazy guy I am, it did occur to me why solder 16 x10uF if I can just solder 4 x 47uF and call it a day... well it turned out that those caps need to charge/discharge very quickly so individual caps is the way to go, otherwise I might get a very lazy sound which is not what I want..

Its terrific that there are solder holes supplied there, very diy friendly, I am planning on using BG Fk nice and fast. The holes mean the leads can be very short, eliminating rf noise!

But still have to remove the existing caps. Thanks hifi2001 for info about trimmer on the DC-30W,an easy way to experiment with the voltage supply to the chips. Think that means the ivs would have to be changed with the final setting if different though.


Playing around with my single TDA1543 test rig, I noticed lower voltage, say about 5-6 gives a thinner but airy sound, and the higher you go the denser and more authorative.


Like your choice of MKP 1837.

Mine is on the way too. I already have Mod parts on the way but I would like to listen to the stock dac first for a few weeks to provide michael feedback.

I noticed one thing on the new design, just by looking at the pictures, which need attention in my opinion (I already told Michael this and he thought it was a good idea):
The input switch for the digital sources. Why not use a (quality) relay? It has two advantages: The digital signal gets lesser affected and you can move the source switch to the front.

I do have one question about the new design:
Can anyone tell me what the 0,01 uF Mica caps doing on the board at the output? Are they meant to filter the output signal?

As for the mods I have these things in mind:

-16 Blackgate FK 47uF for decoupling of the TDAs (this really improved the sound of the V2.1)
- Replace those yellow 100 uF caps on the digital board to Sanyo Oscon.
- TX2575 175R resistors are ordered for I/V conversion (Oh, how much I love these resistors!)
- I've got two 5,6uF Mundorf Silver/Gold/Oil in mind as output caps (leftover from the filterupgrade in my Audio Physic Virgo III speakers).
- Splitting up the power supply: The DC-30 for the digital part (maybe upgraded with shunt regulators) and a Paul Hynes SR1-MR with a seperate power transformer providing 7.8V for the TDA dac chips (This was the biggest improvement after the output caps).
- Change the PE input transformer to Lundahl.
- Change those russian 0,01 uF caps to something better. I use MKP1837 in my current V2.1 and that is already an improvement but this might not be the best solution.

Quote:



In my case, I had to switch back to the Russians because I found the MKP brought on fatigueness to the sound... I initially thought it was great but the more I listened to my V2.1 I simply found it very tiring. So I guess it will be either teflon or better quality PIO.

hifi, Silver mica are pretty hard to beat, but whats the burn in time for those MKP's perhaps they needed more time, could try ERO MKP metalized polyester Audionote seem to like them?

rhodes54, I might try the oscon seps for the digital, should be even better then sp's Ive tried before, seps rival Bgs for digital and are certainly cheaper, also got a 75r texas on the way for spdif, same approach as pats on the Valab.

Whats your take on those Mundorf Silver/Gold/Oil?

Quote:



I've been also thinking about the Sanyo's SEPC's but they are hard to find in the correct value.

As for the coupling caps I will contact an engineer specialized in digital audio. He must have the correct answer for me.

The Silver/Gold/Oil are just the best caps i've ever heard together with the new Duelund CAST. Depending on your taste you might want to go for either Duelund or Mundorf. I personally prefer mundorf. The Silver/Gold are the second best. I don't like Silver/Oil and normal Supreme. They are not that transparent and don't give those microdetails and soundstage as the S/G and S/G/O. It's not that they are not good, it's just not my taste. They hide too much in my opinion.

Quote:



Maybe I didn't wait long enough for those silver micas to settle, with the MKPs I noticed it right away, well I can always put them back in when I have time...

rhodes54, yes the silver mica are output filtering, if you try bypassing them, as in use no caps there, you get a more open sound. I think pats already done that.

Those beige SMD look like Tantulum. But anybody know if there is any need to keep those 0.1uf smd bypass caps on the pins 4,5 (gnd vdd) if you change the 10uf Tants to say, Bg Fk? Bypassing Fk with cheapy SMD seems silly to me.

I found some Sanyo sepc on ebay 6.3 v 570uf, so will give them a whirl on the digital boards.

I just put a cap across the LM334 thats provides the 2.2v ref voltage on my Valab, I used a sanyo oscon SP 270uf, this is the same reg on the Chameleon definate improvement in noise floor sounds more fluid richer!

Whoops, get hiss at full volume with the cap just there!