leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 12:17:41 GMT
Hi guys, Seeing as though my AYA and the PCM1794 dac projects may be a little too diy lets try something different. A good friend of mine picked me up one of these on his travels minus the psu to play with (cheers mate, you know who you are ) Anyway if you search ebay and type in item 140204504125 its basically one of those but minus the PSU The price just for the dac pcb was just over £20 UKP's , (can't remember exactly) you need to supply a PSU, case wiring etc Its got both coaxial and optical Inputs, receiver is a CS8416, Dac is a CS4397, output op-amp is 5532 on a socket. Something to think about, the dac chip is also pin compatible with other CS chips including their top chip, maybe something to look at later ;D Ok first thing I'm going to do is try this thing as total standard just powering it from a simple supply, the next step will be to try the very basic tweaks like op-amp rolling, after that I'm going to build some better PSU's with nice regulation, before SandyK asks yes it'll probably be JLH for the analogue then I'm going to build something discrete to replace the cheesy op-amp, remove the DC coupling caps on the dacs differential outputs and work from there
|
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Feb 11, 2008 12:46:35 GMT
Hi guys, Seeing as though the AYA and the PCM1794 dac projects may be a little too diy lets try something different. A good friend of mine picked me up one of these on his travels minus the psu to play with (cheers mate, you know who you are ) Anyway if you search ebay and type in item 140204504125 its basically one of those but minus the PSU The price just for the dac pcb was just over £20 UKP's , (can't remember exactly) you need to supply a PSU, case wiring etc Its got both coaxial and optical Inputs, receiver is a CS8416, Dac is a CS4397, output op-amp is 5532 on a socket. Something to think about, the dac chip is also pin compatible with other CS chips including their top chip, maybe something to look at later ;D Ok first thing I'm going to do is try this thing as total standard just powering it from a simple supply, the next step will be to try the very basic tweaks like op-amp rolling, after that I'm going to build some better PSU's with nice regulation, before SandyK asks yes it'll probably be JLH for the analogue then I'm going to build something discrete to replace the cheesy op-amp, remove the DC coupling caps on the dacs differential outputs and work from there Hi Leo, What brand is this? NOS? Thanks.
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 13:00:27 GMT
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 15:32:15 GMT
or better yet something like this "forever" solution if adjusted for +5VDC and only used to power the digital section (more on that next ) www.raleighaudio.com/lv_power_supply.htmhas all the AC & DC filtration you will ever need (AC filter are ESSENTIAL for difital supplies because any line harmonics WILL get into the digital signal ) Or [ u]not[/u].[/i] ! This type of DAC chip already has op-amps in the output and worse they are CMOS type and that is the major problem with time slicing DA converters because anything you follow it with already has been comprimised PLUS adding stages just to do it never a really good idea. If you want to transcend this "El Cheapo" DAC into a system keeper then all you need do is couple the output like this www.jensen-transformers.com/as/as093.pdfexpensive at first glance yes but in the end you will leave all that "fiddling" with POS Op Amps behind not to mention the audio grade passive parts that go with the "opamp rolling",not have a need for another "expensive" and likely comprimised power supply for the analog section,will have total galvanic isolation of ALL digital ground noise and a more natural analog domain anti-aliasing filter that in the end comes out a wash on the financial level but whips major ass on the SQ level all done but casing that sucker up if you follow the above suggestions with the only place you can go left is looking at the digital bypasses and the SPDIF input which should have a 75 ohm pulse trafos on the input
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 15:32:58 GMT
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 16:14:10 GMT
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 16:23:18 GMT
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 16:28:16 GMT
sure is dude but two points to consider here 1-the caps are on the [ b]output [/b][/u] of the regulator so there is no real start up charging stress on the rectifiers 2-this is a DIGITAL ONLY power supply so the normal downsides of that amount of capacitance (music sounds slowed and LESS rather than what you would think with all those current reserves MORE dynamic as we mostly all know) do not apply with the upside being a very low supply output impedance and a damn lot of 60/120 hz hum suppression (would be bettered with a CLC before the V-Reg and CRC after but that would make it truly an expensive five volts ) as for SQ it gets about as good as can be expected for such a DAC chip type (read pretty damn good actually all things considered,good enough you won't reach for another DAC for most listening )rather than further degraded by the analog section as most well meaning hobbyists and far too many "experts" do.Simple is damn near always best if it performs the same job as the complicated solution and we far too often over-think and over-compensate for perceived problem areas that many times only require the REMOVAL of parts and not adding in more
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 16:40:13 GMT
Your choice man but remember,active stages mean noise WILL be added from both the power supply and the signal circuitry with all you are gaining being ultimate voltage output that to be honest is usually in need of padding down at the preamp input anyway. If you must go active I think a better choice would be a zero gain buffer circuit (think of it as an active transformer replacement without the galvanic ground isolation ) such as a Borbely or Broskie JFET follower or even an LT1010 buffer run heavy into Class-A with either an AD744 with output stage bypass or the above followers fronting it so the buffer "sees" the low driving impedance it must have (don't even consider other monolithic buffers unless you can get your mitts on "NOS" LH0033's.Most others may spec fine but sound like shit ;D ). "KISS" Bubba.Don't add what you don't need and in DACKage that is more gain (remeber there is already a built in analog gain stage ) unless you are pushing studio length cables and must muscle up
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 16:47:58 GMT
Theres only 1 x 5v input socket to the pcb, the dacs analogue supplies (VA and VREF) are on the same supply line, this then feeds a SMD 3.3v regulator that supplies the dacs digital supply and input receiver. When you say the above regulation is only suitable for digital supply I presume the above reg is suitable for supplying the dac chips VD, VA and VREF pins? I'm just having a look at some of that available iron for the passive outputs
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 16:50:06 GMT
Your choice man but remember,active stages mean noise WILL be added from both the power supply and the signal circuitry with all you are gaining being ultimate voltage output that to be honest is usually in need of padding down at the preamp input anyway. If you must go active I think a better choice would be a zero gain buffer circuit (think of it as an active transformer replacement without the galvanic ground isolation ) such as a Borbely or Broskie JFET follower or even an LT1010 buffer run heavy into Class-A with either an AD744 with output stage bypass or the above followers fronting it so the buffer "sees" the low driving impedance it must have (don't even consider other monolithic buffers unless you can get your mitts on "NOS" LH0033's.Most others may spec fine but sound like shit ;D ). "KISS" Bubba.Don't add what you don't need and in DACKage that is more gain (remeber there is already a built in analog gain stage ) unless you are pushing studio length cables and must muscle up Agreed, I'm just thinking of a tight arse way to get the differential to line out
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 16:58:27 GMT
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 18:49:10 GMT
Excellent stuff, well done to the Rick monster I only got the dac this morning so I've quickly lashed it up to makes sure it works etc, I don't have enough of those caps for the supply you posted earlier so will get those later. I have a spare modified ALW super reg with pre reg replaced with a VBE, this ones set to output 5v so I've used it as temp. I've also swapped the op-amp to a pair of LT1028's. I'm now going to check those output stage idea's you posted
|
|
|
Post by PinkFloyd on Feb 11, 2008 19:56:41 GMT
What an interesting thread Jeez leo, you're quick out of the blocks man... can I have a bottle of what you're on?
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 20:47:54 GMT
make mine a double Thanks man.One does what one can and while there are many ways to come at every aspect of audio I try to only post what has worked for me after a damn lot of years screwing around with #1 on my list being to "simplify when possible and if not possible ? Simplify anyway then MAKE it work " Seems we have gotten into a plug-and-play computer attitude with audio where we look at the picture and miss all the details in a rush to get to the next picture so what i try to do is massage a thing and make it whatever it can be when it grows up before moving on to "the next big thing". Piss poor juggler so one ball in the air at a time is my way when tackling a project (probably why i have a damn trunk full of unfinished projects ) Anyway.when it comes to a digital audio device most attack te problem from the output side of things,the analog stage,instead of the INPUT side of things-feeding power to the chips and feeding the data stream in-as they should before even THINKIMNG about the output because the ancient GIGO (garbage in/garbage out) holds more true for audio work than ever it did in data processing. You need to think of the digital section as a single block first,functional blocks second then when you have all that right think about letting the music out through the analog section which feeds the downsteam part of the music chain. so where do you start ? At the beginning of course compadre ! 1-AC Mains feed.There can be NO RFI riding in or you will have nothing but crap later no matter what you do so filter the AC section of the digital supply with a common mode filter 2-Clean DC with enough current to do the deed.Unlike analog section work you don't need "headroom" as such just a clean imperturbable feed to the actual device pins and this means a SINGLE DC regulator to either individual pin shunts at the extreme or at the minimum LCR filters right at each and every V+ and V- supply pin.That last no easy thing in an already populated boar but can be done if you use your imagination which means under the pcb is just as good as on top and even better (though it looks sloppy and we CAN'T have that in this day of our gear having to LOOK good even IF it sounds like shit ;D ) directly across the digital chips pins (carefully ! Don't wanna fry the little bitches ). We not only want the DC feed to be clean BUT we don't want the various sections talking to each other in disharmony. Why is this disharmony ? Because it is ALL RF and if an RF signal travels to a place then back again and mixes with the original we get "ghosts".Yeah these ghosts may be so fast as to seem non existing but they be there just the same so lock up thos digital chip pins whenever there is an option to do it 3-Proper load matching.Again we are dealing with RF so in reality more akin to antenna theory than audio (and i suck at both so bear with me ) so in short you want a constant impedance with zero reflections.A tall order but one easily dealt with within the limits of our limited knowledge and experience so first you want a TRUE 75 ohm connector on the coax section and that means BNC at both ends of the coax cable (RCA can't get there because the hot pin is too close to the ground shell along with constant impedance 75 ohm load matching transformer (with proper drive at the "send" end and proper bypassing/isolation at the "receive" end) Once you do all the above you have in effect ISOLATED the digital circuits from the outside world leaving what ? looking to the anlog section to get the data BACK OUT where it can be enjoyed as a musical event instead of a set of test tones.... at least that is my story and I'm sticking to it
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 20:49:10 GMT
What an interesting thread Jeez leo, you're quick out of the blocks man... can I have a bottle of what you're on? Its about time somebody else started building something, I bet folks are sick of seeing my stuff
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 21:05:18 GMT
Just like you to know your advice is much appreciated Rick everythings being digested too its also good for others wanting to try something like this cheapo dac and wondering how they can improve things My main co-axial leads for the other dacs are BNC, I just knocked a quick BNC to phono mainly for testing with this cheapo, now I know its ok I can fit a proper 75R BNC plug and proper loaded Newava pulse transformer, theres just a simple SMD cap and 75R resistor as standard for CS8416 input
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 21:09:13 GMT
now HERE folks is a man with his priorities in life in the proper order Cool
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 11, 2008 21:10:27 GMT
Leo While there is undoubtedly merit in many of Rick's suggestions, don't forget that old saying "You can't make a Silk Purse out of a Sow's ear " Also,it is best to stick to the "KISS" principle if you are encouraging others to have a go, with a simple project that sounds good. I do know that this is your objective ! We really need more members to try inexpensive projects, and then gain confidence to tackle more ambitious projects. Perhaps with all the ongoing interest in interconnects, this project could be added to with a DIY interconnect project, perhaps even one from the archives with Pink credentials ? Alex
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 21:17:37 GMT
Your right, tbh I think the other two dacs may have scared a few folks off, one wasn't cheap and the other was 50% surface mount parts, both of those other dacs needed to be built from bare pcbs and came with no parts. This one comes 90% built and is cheap too, ideal for those wanting a go at something easy. Main reason I wanted to try it as standard and then see what can be squeezed out of it
|
|
|
Post by PinkFloyd on Feb 11, 2008 21:46:53 GMT
What an interesting thread Jeez leo, you're quick out of the blocks man... can I have a bottle of what you're on? Its about time somebody else started building something, I bet folks are sick of seeing my stuff I was sick of looking at all the crap I was building (at a fast rate of knots) when I was on the same juice you're on now so I have been taking a well earned rest for a good while now. To be honest it became extremely boring putting headphone amps together in an afternoon..... rush inside, feed it with some music and "yawn"...... "now what will I go and do.... oh, I know, I'll go build another headphone amp!" Like bashing my head off a brick wall leo. The next one I build will have to be something SPECTACULAR, I just can't get the juices flowing at the moment what with one thing and another. Bloody Dacs "yawn", speakers "yawn", amps "yawn"..... all I want to do at the moment is get the 21st February out of the way (court trial) and then try and get relatively fit again. I suppose it's the same old story.... "too much of a good thing" and you lose interest and become numb to things others would be creaming their pants over.... I have most definitely had too much of a good thing when it comes to all things headphone amp / headphones related (and that includes interactive time spent on Head-Fi) I have been spoilt by the whole experience. Don't get me wrong, it was fun at the time and I had such passion for it all..... "PinkFloyd - Head-Fi" quite a buzz while it lasted. Look at this forum as one of those bars that ex- boxers buy where they reminisce on their past glory.... they draw in a small crowd and are happy to be out of the limelight...... that's basically where I'm at at the moment but, believe me, I am thoroughly enjoying watching and learning from you talented guys...... keep up the good work! I'll definitely enter the ring again one of these days but, in the meantime, am happy pottering about with my Little Pinkie power supplies and generally growing old disgracefully ;D 46 years old and facing possible Jail on the 21st I must be an idiot to have gotten myself into that mess Stanley! The "Social worker" who is preparing a report on me for the court ( a 30 year old lesbian wearing doc martin boots and flowery skirt) reckons I "blame others" for the situation I find myself in and that I would benefit from "self realisation therapy"........ BOLLOCKS! stupid bitch, I blame her for my existance, I blame her for my drinking to excess, I blame her for my inability to gain wood without the assistance of viagra...... Would you adam and eve it? A flamin' Lesbian cow with attitude writing a report on the pinkster? Not looking good..... not looking good at all, I've got a VERY bad feeling about that bitch.... she didn't pass the pink "you're ok" test.... she's deffo going to report me as a "bad toy" to the sheriff FunKadelic CoWpat. I've gone off on a tangent here, apologies Leo.... erm..... what was this thread about again? Off to BLAME THE WORLD.... I've run out of fag papers (Rizla to you YANKS.... I blame YOU TOO!) Night!!
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 11, 2008 22:00:19 GMT
Mike (With apologies to Leo for interrupting his thread) Surely an unblemished record previously, as well as no longer drinking alcohol, must amount to something ? Good luck on the 21st Alex
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 22:07:15 GMT
No need for apologies mate, get it out of the system BTW I wasn't really referring to you when I said its about time somebody else started building something, I know your situation , theres more than just us into diy on here so was hoping to see some others have a go too You've mentioned Rizzla's I now want a fag I'm back on the old liquorice papers
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 11, 2008 22:08:18 GMT
Good luck on the 21st Alex Seconded
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Feb 11, 2008 23:44:02 GMT
Actually the heart of that DA Converter is not too shabby as far as Sigma-delta DACs go but only if what it is surrounded with is up to speed as well. a piss poor digital layout (RF again) layout and trying isolated grounds etc. is a recipe for SQ disaster but the analog section ? I have yet to see or hear any evidence that suggests adding a complicated or expensive active analog stage does anything to improve the sound since everything (including the IV stage is already done internall with a seriously bad if asked to do to much CMOS Op-Amp sitting between the internal analog filter and the output pins,as it is with any sigma-Delta DAC chip not multi-segment or with at least access to the signal path BEFORE all the internal garbage is added so the best you can ask to to not much shit up anymore than it is and that means (as stated earlier) isolating the digital from the outside world,foregoing ANY active circuitry on the analog out unless truly needed for a voltage increase or drive or a serious impedance mismatch THEN isolating the "outside" world from the godawful world of digital grounds through galvanic isolation,the line level audio transformer which to my ears can and does beat the lving crap out of most active solutions and if i DO have a need for active then it is usually DAC/Transformer/SS Buffer-Driver or Tube follower-Driver stage-NEVER gain because 1-1.5 VAC (2 volts marginally) works great in my system while 3-5 volts does not without padding down the digital input to my control center of jumping for the volume control/remote when changing sources As a side note-once you level match every single source input to your playback system it becomes clear just how "sterile" digital can be when compared to analog tape or vinyl without some assistance and that assistance intentionally "colored" electonics in the analog section and THAT in an age where tone controls and filters are frowned on which to be honest cracks me up. You can "pre-equalize" the dedicated electronics with all manner of fuzzy tube circuits and "romantic" sounding parts and be cool yet bust a move on a bass upper or a loudness control and the audio gods take away your audiophile membership card The "modding" outfits understand the game and have no problem adding triode stages or DOA's (discrete op-amp) to the analog output of digital players while charging a boatload of loot to do it yet you would be hard pressed to PROVE on paper with equipment plots that ANYTHING was improved in the electronic sense.Fact is many discrete circuits are not even close to "nuetral" and as for tubes can be but more often than not are intentionally NOT (read the ad copy for tube buffers and digital and they TELL you they are changing the sonics to make them sound more "analog" as if that term means "not nuetral" ) Weird but is what it is....................... anyway,the KISS rule does not mean "cheap" because more often than not you will spend FAR MORE for good audio iron than you will for any combination of active circuit and power supply so that dog won't hunt plus many modern active circuits are no more than synthesized "iron" (and dammit,they even SAY it by accident sometimes then bad mouth the old ways to justify charging you too much) so for example A balanced line receiver chip is a "less costly balanced transformer replacement" and an active filter a "less costly synthetic RCL (resistor/capacitor/inductor) filter" plus will usually say some shit like "even IF you could find the correct inductor values they will probably cost far too much" or a tone control or EQ band a "synthetic inductor" because again they tell us we have zero shot at finding good inductor and if we could well then they suck anyway The history of audio is NOT about better sound but cheaper constryction and so we went from a time when ALL amps were low powered single ended class-A to Class A-B push pull NOT because it was a step up in SQ but because you could get MORE POWER PER AMP POUND and the same with the move from tube gear to solid state-all about keeping the cost down,the biggest bang for the buck,same with pc builds over point to point (takes time and skills dammit !We don't have the time and we have no desire to train anyone to do anything other than stuff pc boards ! ) So all about the bottom line in many cases but progress ? Listen to a LARGE horn speaker system powered by a 300B SET amp using an open reel deck playing a quality tape some time then tell me about how far we have come so it isn't that KISS is to keep cost low and complexity low because it would otherwise be a pain in the ass but because it WORKS and has an eighty year proven track record long winded but hadda be said by someone dammit
|
|