I have a couple of PCM63 waiting a DHT output stage, I'm not very fan of NOS so I have to try what filter to use (welcome suggestions), I don't need SPDIF input because I use I2S, one point I'm not sure is about using pasive I/V resistor or transformer.
If you go with two pcm63, i recomend jrc 78 regulators, or texas instrument 78( date code 1994-1997) as floating supply. A simple opa LT1357/58 with 700pf -1nf on feedback resistrot. Digital filter dxf7001. Alteenative versions df1704 ( i dont like pmd100 with pcm63 too much ) passive resistor will be worse. (it makes sesnes when 10 or more pcm 63 per channel) transformer(sowter) also worse. Use mix of cerafine and kz capacitors- those " fix pcm63 colorations) dont use more than 20uf near chips. For smoothing panasonic tsha or ha . You may surprise how transparent and not too much coloured pcm 63 can sound.
Btw 8 of my
Dac have spdif or i2s inputs. As much as i want hear diferencies- I
Cant. Only on one dac (cs8414/df1704/pcm63 pass
Labs d1 output) i can hear. But diferencies between variuos usb spdif converters quite big. Either my spdif very good or i dont know what.
Dac have spdif or i2s inputs. As much as i want hear diferencies- I
Cant. Only on one dac (cs8414/df1704/pcm63 pass
Labs d1 output) i can hear. But diferencies between variuos usb spdif converters quite big. Either my spdif very good or i dont know what.
I have the DF1704 with two PCM1704 in the same PCB from Elektor if I remember correctly, the receiver in a separate PCB was CS8414 or CS8416.
I have also from Audionote DAC the 2 i/v tx and the 2 silver 1:1 tx
About powering the DAC surely I use R-core power tx & Salas Shunt regs. that I used before with great success.
I have also from Audionote DAC the 2 i/v tx and the 2 silver 1:1 tx
About powering the DAC surely I use R-core power tx & Salas Shunt regs. that I used before with great success.
Not sure what you mean by pcm63 coloration, but that chip is the least colored, most transparent multibit dac out there 😅 Those regulators you mention will make certain sound signature more than pcm63 itself.
@merlin el mago
Shunts are great for it. For filter, if budget allows go for https://www.kusy.audio/digital-interpolation-filter-fir/ not sure if anything on market is as good as.
As for dht output stage, still searching for a good solution 😁 Grounded grid d3a triode strapped is great for it.
For transformer i/v - output, don't go less than @bisesik work. @Dorin Bodea has great transformers too.
There is no difference it is all audio voodoooo blooo tooth sounds the same when 100%volume. I go optical, not taking any chances, feed is only 44, 48, or 88, no other way to minimize digital jitter/noise.
Maybe you could build a pcm58 dac, it has a way simplified architecture than tda/pcm63, the 1701 is possibly almost the same chip.
I succeeded at making this thing sing... but the truth hurts... records are very poor quality.
I did some tests with kids singing and their voices comes through as natural as it can be, but recordings are just recordings.
I found that the df1704 is OK, but it doesn't come close to the SM5847.
You need to perfectly adjust all 4 trimers, the type of chip doesn't matter in this setup, the cheapest from China is just perfect.
The more you multiply the chips/streams the closer you get to that modern sound.
for me its hard describe dac sonic siganture as dac chips, easier like it would be speakers. the most neutral chip from burbrown i heard is pcm56k closely folowed by pcm1702. hights are not refined on pcm56 but overall its prety neutral. pcm63 with almost any exotic IV or exotic power suplies ( subjectively )have elevated uper midrange 800-4000Hz, because of that soundstage seems smaller. of course vocals or some single intruments sounds vey lifelike. high frequencies also have tendency of being too smooth and rolled off comaparision with pcm1704 or pcm1794. on the othe hand pcm1704 have (subjectively) dip at 2-3khz - or increased 200-300hz which make sound darker and more juicy. its also colored. but pcm1704 colored as pcm1794, and pcm63 is totaly diferent coloration which makes dac sound "too analog" in bad meaning. for pcm1704 I can use exotic power suplies or exotic IV and sound does not become "too analog". if pcm1704 would have that musicality and rythm in music as pcm63 have i would never ever look at pcm63. but something special is about this chip how it conect listener with music. I could say its coloration, but many other chips have its coloration ir thei IV stages or power suplies have colaration and what does pcm63 is diferent.
Pcm63 uses multiple dacs averaging which can only augment the volume, it cannot substract errors.
If you // multiple pcm56 you could get similar results?
If you // multiple pcm56 you could get similar results?
What you're hearing definitely does not compare to what i hear 😅 you pretty much described tda1541a. It is musical, but troughout the range insanely clean, not hearing that elevation. Could be due to power supply, i/v? I prefer to use discrete shunt referenced and tps7 series regs. Older regulators, especially 78 79 have a fuzzy signature around that mid and drop soundstage.
1702 is the bassiest of them all, rolled off a bit in comparisson up top, sounds warmer, less attackish. 56 haven't listened too much.
1702 is the bassiest of them all, rolled off a bit in comparisson up top, sounds warmer, less attackish. 56 haven't listened too much.
Colinear DACs so dual DACs per channel. Not sure what is meant by "augment the volume, it can not subtract errors"?!
What sometimes is seen with old parts is some properties they never had in reality. This seems part of the "they don't make them like that anymore" nostalgia. Such DACs also seem to need "tube magic" and jittery receivers 🙂
What sometimes is seen with old parts is some properties they never had in reality. This seems part of the "they don't make them like that anymore" nostalgia. Such DACs also seem to need "tube magic" and jittery receivers 🙂
Oh I have heard them all but best one was "CS8412 may be jittery but so is music" (was about an AudioNote DAC).
Comparisons of such devices with other devices are practically always off (not just because they do not adhere to standards). Totally useless to bring otherwise excellent sounding/measuring DACs to such comparisons. I have heard devices sound clearly better overall than such oddball devices and still people claiming the opposite probably a bit blinded by the glow of tubes and the brand name 🙂
It is fun to bring a device and put it in a shoebox and then let people listen just with ears 😀 Or by putting completely unexpected electronics in a large very nice casing. I once "won" such a contest like listening session with a large casing containing 2 tiny 6 Euro Sanwu TPA3118 modules from china combined with my own regulated linear PSU. This was with a few expensive tube amplifiers also being in the contest. Or by having tubes glowing in a class D amplifier casing (doing nothing) and then people claiming it sounds so much better. There is much psychological stuff going on in audio but it takes guts to admit that.
Comparisons of such devices with other devices are practically always off (not just because they do not adhere to standards). Totally useless to bring otherwise excellent sounding/measuring DACs to such comparisons. I have heard devices sound clearly better overall than such oddball devices and still people claiming the opposite probably a bit blinded by the glow of tubes and the brand name 🙂
It is fun to bring a device and put it in a shoebox and then let people listen just with ears 😀 Or by putting completely unexpected electronics in a large very nice casing. I once "won" such a contest like listening session with a large casing containing 2 tiny 6 Euro Sanwu TPA3118 modules from china combined with my own regulated linear PSU. This was with a few expensive tube amplifiers also being in the contest. Or by having tubes glowing in a class D amplifier casing (doing nothing) and then people claiming it sounds so much better. There is much psychological stuff going on in audio but it takes guts to admit that.
Last edited:
Last edited:
Try applying a CS8412 low jitter replacement module (for instance one with WM8804) and check the reactions.
We often devote much time, money and energy in technical improvements (much more than actually using the equipment) but maybe 75% of our experience when listening is psychological/prejudiced/pre-opinioned. Bold statement but not far from your own truth I guess. So excellent stuff may just not appeal because of various reasons (sigma delta = 666) and stuff that is not OK on paper and instruments or that even manipulates sound may just suit you well. One of the both worst and best (shared number 1) discoveries is that even a device you definitely not like will be liked by you in time when you don't have anything else at hand.
It's just audio.
We often devote much time, money and energy in technical improvements (much more than actually using the equipment) but maybe 75% of our experience when listening is psychological/prejudiced/pre-opinioned. Bold statement but not far from your own truth I guess. So excellent stuff may just not appeal because of various reasons (sigma delta = 666) and stuff that is not OK on paper and instruments or that even manipulates sound may just suit you well. One of the both worst and best (shared number 1) discoveries is that even a device you definitely not like will be liked by you in time when you don't have anything else at hand.
It's just audio.
Last edited:
when I say there is diferencies in dac chips sound signature i mean SAME ACTUAL CIRCUIT, listened at same time (+-1minute, sometimes as short as 10-seconds- jsut to switch cables) for example the comaprision of pcm63 vs pcm1704 in same balanced dac where 2 stacked (or one- i have tried both versions when testing ) sit on 2 chanels and pcm63 in other two( balanced DAC) there is some adjustments sneeded in volume control (2mA vs 2.4mA, or 2mA vs 1.2mA) but resuts ALWAYS the same when changing output of dac to amp(no need to turn of dac - so no warm up or changes in sound because that) its very easy to tell which dac chip is playing in blind test- because vey big subjective diferencies in midrange are- pcm63 midrange is pushed up(or pcm1704 is recessed) we hae also saped palces of thips in dac to eliminto possible diferencies in IV circuits( which are identtical 4 but still) the digital section of all 4 phases is also identical( - XLR phae are NOT inverted digitaly- just 4 same polarity outputs)
I would not talk here about subjective or MAYBE (looks to me) and its not single person opinion, and also not based opinion of multiple listeners who knows other person findings. most of the times listeners( including me) does not know which chip playing. but simply noticing midrange qty makes easy to quess whhich chip is playing at the moment.
amplifiers gave digitaly contrled volume so output levels on both chips sets withing 0.1db(re checked with microphone of course)
the question is not IF THERE ANY DIFERENCIES in pcm1704 vs pcm63, i just experimented how to make pcm63 soundstage and high frequencies be same as pcm1704 and make midrange less shouty upfront and stll maintain that unique character of pcm63.
I cant comment about 78/79 regultors sound IN GENERAL as they are simply too much diferent in sound between each other. i just found two makers where 78 sounds better than modern regulators like lt1963. My,and my friends finding that Motorolla 78xx have rather weighty , and non sharp sound. it was amde same year as Natonal semiconductor 78xx but sound is SO MUCH DIFERENT.
My other surpises was that super complex discete power suplies from MARK LEVINSON, or more simpler DIY versions(like JUNG etc )always sound too refined ,too polite and with subdued dynamyc, and not razor sharp - when comapring to those exactly those 78xx regulaors. I want like dicrete regulators because of great reviews, and because they look cool and because I am audiophile. but once you start comapring in AB tests funny outcomes - not whats expected. i remember trying to figure out why two mark levinson units sounded diferent- diferent regualtor was fitted! and then why Mark Levinson no36 sounded so diferent from no36s- same schematics but.. diferent power suplies(simple 78xx/79xx vs very complex discrete)
generalising all 78xx regulators is like its generalising sound of signetic 5534 from 1992 with with 2020 TI 5534. the diferencies is hughe.