Extreme X-DAC v3 mods by sandyK

[Deleted]

Brent
Apparently the first time round, in a hurry, I momentarily forgot the pinout of the 79L12 was different from the 78L12.The attached photo should however clearly show the correct orientation of the 79L12 , and the leads crossing over each other.
Alex

[Deleted]

Brent (and others)
For what it's worth, I have rechecked the photo, and I SHOULD HAVE insulated the centre lead so that it can't possibly contact the copper below. I didn't bother, as this lead was fairly rigid, and in my case, well above the copper below. It may be worthwhile replacing the 100uF electro added to the 79L12, purely as a precaution.
Alex

[Deleted]

mrarroyo
The photo is on page 1 of this thread , along with the other photos. Unfortunately, the +VE
sides of the 2 added electros don't clearly show the polarity, but the little red marks on the 2 electros are actually + signs, and quite clear in my original photo.Yes, I should have made the typing bigger !
Mike, you usually deal with much higher voltages than I do, and I bet that some of your slip ups have caused much longer sparks than mine have !
Alex

[ungawa]

Wooh ! Finally !

Sorry its taken a while to get back ... but its taken this long to get it cleared up !

Alex - I must have had an errant connection in place - it took olut the 79L12 .. Taking the unit back to its orig state allowed it to lock to an incoming dig signal, but no audio out .... and not having a backround in electronics is so frustrating doing this as I didn't know "where to look" so ended up getting a boffin aquaintance to check it out, and it turned out that a 7805 had failed ...
Once this was replaced.... I continued the mod today - very carefully!! ... No Smell, No crackles !!! I've completed the OpAmp/Vreg part (tho haven't replaced the IC Caps yet) .... and it sounds wonderful .. it feels so much more relaxed than what I was getting from the LM4562's...
I dunno if that was a state of mind thing or what - but I will roll them back in at some point to compare ..
In the meantime I'll let it settle down.. get the caps, and let you know how it goes ... er - and keep an eye on the smell !

Cheers - Brent

[Deleted]

Brent
Glad to see you have it sorted out.I don't know how you managed to blow up a 7805, as it isn't on the same supply rail. The 7805s etc are fed by the 2 diodes near where the bridge rectifier used to be.Are you using 2134s in all positions, but running off the new 12V regs ? If you are, the 8066 should give no problems. Do you have a PCB drill and .8mm drill bit ? Some modules pins can be a very tight fit, and the holes may need a little enlarging. If too tight, get your acquaintance to do it for you. You don't want to take any chances when you are this close.
If any questions email me, we are in fairly similar time zones. How would you rate the performance at present compared to standard with plugpack? Does it make up for the hiccup ?
Alex

[Deleted]

I.C. Converter modules. - GENERAL
If anyone is making up converter modules using their own I.C.s and BrownDog converter boards, don't try using .9mm matrix pins. They are too large, and enlarging the holes will most likely destroy any through plating. A friend of mine reminded me that gold plated square pin , 28 or 40pin header strips, which snap apart readily are a better choice. (I used them too)Normally there won't be too many problems with preassembled modules.
Another way would be to contact Mike and see if he stocks the I.C.s you want to use, and get him to make the modules for you.
SandyK

[XTRProf]

Oh boy, oh boy, too good to be true. Might as well bum it up for all MF X-Dac V3 Xtremists.

[Deleted]

In PART 2 Photo 3, I show generic 100uF electrolytic capacitors replacing the original capacitors. Yesterday, I replaced those with 100uF 35V Elna Starget capacitors. There was a further noticeable SQ improvement. However, this improvement may not be noticeable on all systems.
SandyK

[Deleted]

After the further small gains with the 100uF 35V Elna Starget capacitors, I recently replaced the 2 x 100uF capacitors at the output of the 78L12 and 79L12 regulators with 100uF 35V Elna Stargets as well. There was a further small improvement. Unfortunately, these capacitors are too bulky to fit under the PCB, so I scraped the PCB track on top and soldered the appropiate side of the capacitors to this earth . I was able to drill a couple of small holes through a nearby unclad part of the PCB and insert the other side of the capacitors through these holes, then solder to the supply rails underneath . This now means that the whole 4 capacitors are mounted on top of the PCB.

SandyK

[XTRProf]

Hi Alex,

Just a thinker, anyway to have a stiffer clock stuffed in the X-Dac V3?

[Deleted]

XTRProf
Unlike our highly experienced DAC modder. (the one with a pussycat for an avatar) I don't find the necessity to go to these extremes. I find that the use of a well dampened laser assembly and fast switching Switch Mode supply diodes in the source, results in a marked improvement to the SQ via SPDIF. Preventing interaction between the supplies of the DAC by the use of better/extra regulation and decoupling , when combined with the cleaner source SPDIF output, gives very good results. IMHO, in HI-FI , it is the combination of many small improvements in the whole chain, that yields the best results. Of course, this combination of many small improvements, makes it easier to further refine individual source components.
Alex

[XTRProf]

Hi Alex,

After having a preliminary look at your methodolgy, I will like ot ask is this the latest flavour of the month for your recommendation of BB OPA2134 & LM4562? Are these direct replacements for the NE5532? As this the easiest, I properly get this done with the latest favour to see how they sound. Or the BB OPA 627AP is much better sounding?

AD8066 is just only for buffering, right? So anyway, to bypass it for better sound as I'm having a X-10 V3 tube buffer already.

Just some thinker for me and will go deeper when I have more time on my side. But perhaps, start little and build up a storm - The Perfect Storm. Also, then can compare to see how much improvements to each stage of mode.

[Deleted]

XTRProf
The attached basic circuit will show that the AD8066 is not just a
buffer. My recommendation would be to fit quality turned pin
DIL I.C. sockets to all the analogue I.C.s positions , except those of the AD8066 and fit OPA2134. They seem to work better than the LM4562 in these locations. A combination of the LM4562 and AD8066 is likely to be ultra detailed, and fatigueing.
Why don't you try without the tube buffer ? You may be very surprised just how good the X-DAC V3 can be direct into a quality amplifier. In my case, I am able to use my Oppo DV981HD DVD player to digitally remotely control volume via the X-DAC V3, bypassing any preamplifier, for the ultimate in SQ, particularly with high resolution source material. The Oppo has to be set to output LPCM, however to do this.
Alex

[leo]

You ever tried removing the first op-amps and using passive I/V feeding that differential to line out op-amp Alex?

Or maybe even add something like this instead of the output op-amp removing the filtering on the input and maybe changing those J-fets to transistors home.quicknet.nl/qn/prive/ra.vdsteen/pdfs/CD_outputstage_PCB.pdf

[Deleted]

Leo
I prefer to think that the MF designers knew what they were doing, but built the X-DAC V3 to a price constraint. I try to avoid completely redesigning a unit like this where there isn't much additional room, and the PCB tracks are very fine. These modified units clearly outperform the Marantz SA11 at >AU$4,000,and I do not believe that worthwhile gains would be achieved over what we have already. All 3 modified units in Sydney are streets ahead of anything else we have heard elsewhere. You have to know when to leave well enough alone!
Combined with Class A amplification and the old QED speakers , and fed directly with high resolution material from the Oppo DV981HD player, the results are better than we have heard in any HI-FI showroom , even when using Quad ESLs. That Norah Jones 24 bit recording is sensational in this mode.
Alex

[leo]

sandyk said:

Leo
I prefer to think that the MF designers knew what they were doing, but built the X-DAC V3 to a price constraint.

Thats why they copied the datasheet ;D
Anyway your happy so thats fine, we all have our own way of doing things

[Deleted]

"Thats why they copied the datasheet"

What a good starting point ! ;D

Leo
Let's choose to disagree on this one ! You prefer NOS, I prefer OS. As this DAC oversamples to 24bit 192KHZ, the "after the DAC bit" is reasonably optimised for that. One friend has tried removing the final filtering caps, but there really is bugger all in the difference. Passive filtering as you described, often leads to lower O/P level too ? Let's see how good your Oppo sounds via your choice of DAC into that garage Xmas lighting substitute when completed. It will be interesting to see how digital remote control of volume using the Oppo into your choice of DAC , pans out. Sometimes you don't realise just how good a component really is, until you feed it into something with much higher resolution than normal. ( Try those TL speakers if you can,too.
You may find that the bottom end is more realistic. It's a shame they are too big for your new listening area.)
You will then be able to beat Simba and XTRProf at their own game ! . The little changes become even more obvious, whether for better or worse.
BTW, It's almost XMAS, so in a few days time, you will be able to do the SMPS mods etc. on your Oppo Xmas present ;D
Alex

[leo]

I wouldn't use adjustable SPDIF attenuation with a 16bit, this is where the newer one would work better

Anyway you know me well enough now Alex, I think its great to discuss this stuff, in a way I'm annoyed because in one hand I have one of the best spec dacs where I've thrown almost everything I can at it and the other hand I have something thats bigger, requires many more supplies, cost more, it uses out of date dac chip etc yet to my ears and others who has heard them side by side commented the oldie sounds better
BTW I did also try the classic op-amp I/V as detailed in the datasheet with the PCM1794, it is very good but for human voice and piano still sounds electronic
The PCM1794 has the better bass, massive dynamics when needed.
TBh soon as I can find something I feel betters the oldie then thats fine, I don't write anything off until trying it first no matter how crummy it looks

[Deleted]

"yet to my ears and others who has heard them side by side commented the oldie sounds better"
Leo
I know the feeling. I would have loved to ugraded the caps as well in the old Marantz CD65 with it's TDA1541A,extra power supplies, and it's home brew plugin buffer stage etc, but there comes a point when you spend more time keeping these old machines going, than listening to music ! (mainly IDP plugs and sockets)
It's a bit different when building a new design DAC using these components, but due to manufacturing techniques when these now very old chips were first made, they may not have as much life left in them as you would hope for.
Even premium digital chips fabricated for Telecommunications use , started to fail after a certain number of years. I forget the exact reason, but I remember it being mentioned that if the chips had been assembled out in space, these failures wouldn't have occurred. Today's strictly controlled manufacturing environment is much better.
Alex

[leo]

Luckily for me the old S2 was unused and still packaged up when I got it, its true that there is a few duff TDA's floating about, these chips get warmer (I fit heatsinks) and a lot of manufacturers run the supplies a bit too high!
You fit closely matched PPS caps for the MSB/LSB decoupling and these old things sound extremely good! wish I'd never tried one tbh

In the new year I'm hoping to try a fresh build with the PCM1794, I'll transfer the best bits from the one I use now and keep this one as the Guinnee pig

[Deleted]

"You fit closely matched PPS caps for the MSB/LSB decoupling and these old things sound extremely good!"
Leo
I found the SM capacitors around the TDA1541A in the CD65 were very poor tolerance. Some of those replaced were closer to the next/previous value in the range !
Replacing them with decent through hole types made a world of difference.
Alex

[leo]

sandyk said:

"You fit closely matched PPS caps for the MSB/LSB decoupling and these old things sound extremely good!"
Leo
I found the SM capacitors around the TDA1541A in the CD65 were very poor tolerance. Some of those replaced were closer to the next/previous value in the range !
Replacing them with decent through hole types made a world of difference.
Alex

I have an old CD65 boxed up underneath the bed, I seem to remember it using SMD 20% tol ceramics on the bottom of the board, I like the SMD PPS because of the tight tol and low inductance, infact these things are supposed to be the nearest to polystyrene and available in higher values.

Anyway! enough of this vintage chip and sorry to take your thread way off topic

[XTRProf]

sandyk said:

XTRProf
The attached basic circuit will show that the AD8066 is not just a
buffer. My recommendation would be to fit quality turned pin
DIL I.C. sockets to all the analogue I.C.s positions , except those of the AD8066 and fit OPA2134. They seem to work better than the LM4562 in these locations. A combination of the LM4562 and AD8066 is likely to be ultra detailed, and fatigueing.
Why don't you try without the tube buffer ? You may be very surprised just how good the X-DAC V3 can be direct into a quality amplifier. In my case, I am able to use my Oppo DV981HD DVD player to digitally remotely control volume via the X-DAC V3, bypassing any preamplifier, for the ultimate in SQ, particularly with high resolution source material. The Oppo has to be set to output LPCM, however to do this.
Alex

Hi Alex,

Noted. So you feel OPA 2134 is better than LM 4562. But with the tube buffer X-10 V3 that can change since with the LM 4562 you mentioned it's going to be extra clinical. I agreed technically that wouldn't be ideal since there will be an extra stage into the signal path. This deviates from KISS that I also firmly believed in. But zenly that could be a different story. There is also the 2604 consideration as there are many who had gone this route as well for some reasons that may be technically and zenly correct or righter. I'm just not sure until I have some feedbacks on the 2604 sound.

I do like to take out the X-10 V3 from the line in keeping with the KISS philosophy. However, I had only put in the X-10 V3 now since I don't have the chance to evaluate the differences as to whether it is favourably for me zenly or not before the X-PSU V3 commited suicide. I need some time to be sure that's for me or not as I don't believe in condemning sometime until I have lived with it for some time to know thoroughly its characteristics and synergy in my system. Tachnically, there is no argument as to whether it should be in or not. But I'm ZenTech so do alow for that.

Ya, I know the universal player of any brand has to be set to output digitally. That's what I'm doing now or I couldn't possibly have use the X-Dac V3, right? Wow, your Oppo has an inbuild digital volume control? For me, I'm just using a passive pre. Perhaps, I may put the X-10 V3 as a buffer for the varying passive pre vol pot for better impedance matching if not favourably used at the X-Dac V3 output. Technically, this is also correct as well if not zenly.

Anything can happen and that's what Hifi or AV tweaking is all about.

[Deleted]

XTRProf
The Oppo DV981HD has a few tricks up it's sleeve. The remote volume control via SPDIF when set to output LPCM is an undocumented feature.It also improves markedly with bridge rectifier diode replacement etc. in the SMPS. These simple mods are posted here somewhere.
One lucky member had his wife give him an Oppo for an early Xmas present !
For fear of retribution, I won't comment further about people who put valve type buffers after high quality sources !
Alex

[XTRProf]

Hi Alex,

You mention only the analog stages are not regulated. What about the digital stages? Are they regulated? Thanks.