DIYHiFi.org

All else being equal, I am assume (correct me if I'm wrong) that a reasonably-implemented pulse transformer (PT) "circuit" on a stand-alone DAC's input is better than not including one.

Given this, I found something *seemingly* odd (to me, anyway) on the Musical Fidelity A324, a stand-alone DAC: A PT on its S/PDIF OUT but *none* on S/PDIF IN...



Someone suggested that the topology for this model may have been scaled down from one of MF's higher-end units...and economies-of-scale did not "allow" the use of better/more components. And/or the manuf didn't want the unit to sound too good (IIRC, MF used to have beefy trade-up policy to get you to buy their more-$ units).

W/o having access to/experience with MF's other DACs, I can't anything for sure. What I can say is that some early experimentation with a SC PT did not make this unit sound that much better than stock. Maybe it's time for a Nuwava?

No DC blocking cap on the primary, secondary grounded and a shield in between, and an RCA jack (of course!). All make for rotten performance.

Might be one reason you don't like it.

Jocko

Thx for your reply.

When I added the SC PT, I basically "dropped" it in...between the input jack and C35 (0.1uF). I wan't too clear on this in the orig post, so I hope this is what you meant.

As far as the mod, I think I'm going the route you apparently endorsed a while back on diyaudio.com: the final design of the S/PDIF input is, TTBOMK, here.

Hi,
There is another potential problem with this DAC. I remember seing some photos in Hi-Fi Choice a few years back. The SPDIF circuitry is somewhere past the middle of the pcb, isn't it? I mean quite far from the back panel...

Can you post a photo?

Hope this is clear enough...



(see this page if above image is not clear enough)

georgehifi recently posted part of this unit's schematic here:

http://www.diyhifi.org/forums/viewtopic ... sc&start=0

Other portions of the unit's schematic are here:

http://www.diyaudio.com/forums/showthre ... adid=60414
(see georgehifi 's posts about half-way down this page)

Thx for your help.

No need to comment on how bad this design is.

Jocko

Ahem, that's the one! Best under 1000 coins...something. Well you know what to do next...

Actually, I don't :)

FWIW, the MF A324, from 2002-2004, rec'd Stereophile's Class A recommendation in the category Digital Processors. It was one of the least-$ DACs to do so. I do think that there was more politics than usual (even for Stereophile!) behind that decision because the rag's editor and MF's founder were buddy buddies.

As a result of the glowing -- but biased -- Stereophile review, lots of these models were sold. And, hence, lots are still in "circulation" in the used marketplace.

It would be great to squeeze as much as possible out of this unit (design) -- mod-wise -- other than what was posted here at diyaudiosome time back. (The schematics are here)

The DAC used in this model is the PCM1738, a delta-sigma chip which carlosfm didn't seem to care for in this post. Some other chips used are: CS8420 (sample-rate conv.), PIC16C71X, Philips 74HC4094, NPC SM58498F (ASRC), two SN54HC244 (one before and one after ASRC), and input RX ICs are National DS26LS32AC and Philips 74HC131N (8-input multiplexer).

I guess the chassis and PS are salvageable. But given the parts-list and topological info above, is further-modding this unit worth the trouble? Or is one better off ditching the main board altogether and putting in something else?

Hollow, years ago I worked on a Sony SCD-333ES.
That thing had PCM1738 dacs inside.
The mags gave it good reviews, the internal pics impressed the audiofools, but really, it sounded dreadful.
I've put a good clock, with S/PDIF reclocking (Tent XO3), solved lots of problems of improper decoupling (huge Silmic caps miles away from the chips... typical Sony), analog stage, PSU, etc.
It certainly improved vastly but the end result was that it was much better as a transport, connected to a (very modified) Audio Alchemy DDE v3.0 dac.
Yes, with S/PDIF.
My conclusions back then were that the PCM1738, although used in some high(er) end players, is no much different, sonically, from most delta-sigmas.
Then again, a 6 channel player (like the SCD-333ES) is always too compromized for high end stereo, 2 channel playback.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

Carlos,

You have anticipated me
Just wanted to point out that the 1738 is an early advanced segment dac, exactly like the pcm1794, the cs4396-97-98-122, the WM8740 the AD1955 etc.. That is, like all the modern good sounding dacs?

So it's the implementation, again?

But that twisted pair.. Think of that shortwave pollution generated all around..

Ciao, George

First you'll need to replace the RCA with a BNC 75 ohm and use a piece of 75 ohm coax down to the input transformer. In case you don't really need the digital output and the optical input then you can bypass those two ICs, reuse the output transformer as an input TX and go directly to CS8420 input(s)....and then of course ask Jocko a lot of questions.

I have to acknowledge my limited understanding of the 'RF land'. I whish I knew more...


btw the digital output is just the input looped out, there's no reclocking or upsampling done there... so why keep it?

First and foremost…super thx for everyone's high-quality & results-producing feedback. I'm glad I switched forums on this project!



Carlosfm: If I may be so bold… what are you using currently for a DAC and xport/player?

BTW: I'm new to this forum so it's archive's treasure troves are, for me, mostly unexplored. Some of my queries may, hence, be redundant … but that's the learning curve, I suppose. Anyway…

If I do ditch modding this DAC, what DIY DAC projects do you (anyone) recommend as being especially noteworthy? (Please filter out NOS and vacuum-tube designs, if possible ;) -- I've gone that route a number of different ways and never found the sound to my liking).

I've got a Tent clock on order for this unit, and will most likely (ultimately) integrate that with a Kwak Clock.



You may want to scope out this thread...

http://www.diyaudio.com/forums/showthre ... adid=94695

…and this message…

http://www.diyaudio.com/forums/showthre ... ost1114753

… at the other site for some more comparisons/contrasts between "popular", "modern" DAC chips.

As far as the A324, I believe the model up from (or the model that came after) the A324 used another TI/BB PCM17xx chip (1792??) Unfortunately, and all else being equal (i.e., if the rest of this unit's topology is decent and continuing to mod is efficacious) there is no easy way to pop in a better DAC chip -- say a 1796 -- into this circuit.



The image I posted earlier is several mos. old. I ditched the output circuit some time ago. But I haven't gone so far as to directly inputting SPDIF IN into CS8420, bypassing all that other junk, so thx for the push here!



Glad you brought up SPDIF and asking "Jocko a lot of questions" because…

… my transport (and SACD/DVD-Audio/-Video playback) unit is a Pioneer DV-59AVi. Potentially, as stock, a "6 channel player compromised design" as Carlosfm noted above.

Here is the Digital Out section of the Pioneer:



I want to upgrade its S/PDIF out. Opting for BNC, as many of you have suggested, is an obvious choice. But, what else? And especially with respect to my Pioneer. Scanning this circuit from a 2005 diyaudio thread…



… I *think* the best route for re-constructing the Pioneer's S/PDIF out is to integrate the pulse-transformer-based output Katapum/JH worked outa while back on diyhifi as well as diyaudio.

But I'm in the learning-curved/newbie-dark about elements of the circuit above.

First, what are the values for the two "2R???", just before the BNC jack for, say, my Pioneer's output.

Perhaps the answer lies in JH's response (below) but I can't "decode" it…

Specifically, in that late-2005 thread JH wrote:



Pad values…are these the "2R???" I queried about above?

Also, I'm going to *assume* -- correct me if I'm wrong -- that Katapum's SPDIF INPUT circuit is *generic* enough to be used in most DACs:



There was some further modifications to this input circuit here…
http://www.diyhifi.org/forums/viewtopic ... ight=#7196

…but I'm not sure what the final SPDIF INPUT circuit looked like?

Just wanted to clarify...



My plan is to bypass all the Pioneer's COAXIAL circuitry and drop in some form of the JH SPDIF input design: in the first (Pioneer) schematic above, add a wire or trace just after "DOUT", and then some form of JH's SPDIF input w/BNC.

I'm not sure if this is do'able as I don't fully understand what the COAXIAL circuit does other than, perhaps, load, "filter" and mute the output. (??)

I'm using at this moment in my main system a very modified and very old Philips CD160. (!)
At 4x OS, with that dreadful 7220. Which does sound very good, if you generate BCK to the TDA1541A by dividing the 11.x Mhz master clock (Tent XO) by 2. Disconnect BCK that comes from the 7220.
I also have a (very modified) Audio Alchemy DDE v.3.0 dac, which is connected to the CD160.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

you can connect J1 of transformer circuit to emitter resistor R607. I don't think that DOUT pin is capable to drive 75 Ohm