Recently I'm coming to conclusion that battery power is the best. I'm powering my current DAC setup from 6V battery and DAC (TDA1543), Tent clock and reclocker all have separate low drop regulators. I didn't try any cap comparison directly on DAC PS pin (but will do it very shortly), but what I noticed is that I liked 4,700 Elna Cerafine directly across battery, it makes the whole digital section to sound better. I also tried 22,000 Siemens Sikorel ans 2,400 Sprague Extralytic ( I it's their top grade) . Extralytic sounded very bright and Sikorel was good but not as smooth as Cerafine.
I'm using Rudolf's output stage (you should be familiar with this circuit) and I found that it sounds better with raw PS (only MUR860 and Technics/Panasonic 3,300 electrolytics) that with ZTX regulators he was suggestin. Still not bypassed battery power sounded better: smoother, glare free and very natural. I didn't noticed any cavities people are talking about (not dynamic and washed out sound). I'm using 12V Yuhasa 2Ah cells.
My post may be slight off topic, but I'm also didn't understand exactly what kind of circuits you had in mind taking about DAC's Vref.
I'm using Rudolf's output stage (you should be familiar with this circuit) and I found that it sounds better with raw PS (only MUR860 and Technics/Panasonic 3,300 electrolytics) that with ZTX regulators he was suggestin. Still not bypassed battery power sounded better: smoother, glare free and very natural. I didn't noticed any cavities people are talking about (not dynamic and washed out sound). I'm using 12V Yuhasa 2Ah cells.
My post may be slight off topic, but I'm also didn't understand exactly what kind of circuits you had in mind taking about DAC's Vref.
Surprisingly, an interesting thread didn't evolve too much.
As promised, I did some tests today and here are results. I attached an IC socket to the V+ and GRND pins of TDA1543, so I could easily change caps and compare. The PS consisted of 6V battery and there was no dirtectly connected cap before BB low drop regulator and battery. I attached SMD C0G caps (0.01 and 0.1 stacked) directly between power and ground pins.
I was not impressed with 150u OsCon. It provided rather soft and artificial signature, quite "organic" really
. I didn't like 1000u ZL Rubycon, an uninvolving sound. Also, 2200 STD BG wasn't impressive and somehow narrowed the soundstage and the highs were a bit too snappy.
I quite liked 1200u Panasonic HFQ, but comparing with the better caps, the highs were still too harsh and not silky enough, although not bad at all. Interestingly same type of cap but in 120u value sounded much worse. I liked Elna Cerafine pretty much and untill I tried BG N type i though it's good. The Cerafine had pretty balanced presentation, with smooth highs, but a little bit muffled and mellow. The best cap, in my test, was 4.7 N type BG. It simply made the music alive, the detail was much better, the sound seemed to unrestricted and very natural. There was a certain coherence to the musical notes, that other caps simply didn't provide, they sounded grey by comparison. BG was very fast and uncolored and sounded real. I got good result adding 470u Cerafine in parallel, there was more midrange energy in this way, but somehow it didn't sound as clear as BG alone.
The bass didn't suffer anything when using 4.7u cap only. Quite opposite, the bass was actually more clean, detailed and very fast. By adding Cerafine, I was getting more bloom, but to me it seemed like extra coloration. It was slightly more heavy, but not as tight. I think I prefer BG by itself, although it may seem a bit bright sometimes.
All those observations are pretty much in line, with what creator of GainCard is claiming. It seems to me too, that too much capacitance slows and muffles the sound, veiling all the natural characteristics that it should reflect. I also tried no cap at all, but it didn't sound right, it was missing something (a bit of air, weight and sparkle)
PS: Regarding OsCons I see people raving about them, but here I didn't observed anything special really. Maybe they are better in pure digital circuits, although it seemed to me that after changing Panasonic HFQs to OsCons, in KC7, I didn't observed any improvement but rather not so special sound. Didn't know what to blame that on, but maybe those OsCons are not that good sounding as people claim? Anybody has any "real" listening impressions?
As promised, I did some tests today and here are results. I attached an IC socket to the V+ and GRND pins of TDA1543, so I could easily change caps and compare. The PS consisted of 6V battery and there was no dirtectly connected cap before BB low drop regulator and battery. I attached SMD C0G caps (0.01 and 0.1 stacked) directly between power and ground pins.
I was not impressed with 150u OsCon. It provided rather soft and artificial signature, quite "organic" really
. I didn't like 1000u ZL Rubycon, an uninvolving sound. Also, 2200 STD BG wasn't impressive and somehow narrowed the soundstage and the highs were a bit too snappy.I quite liked 1200u Panasonic HFQ, but comparing with the better caps, the highs were still too harsh and not silky enough, although not bad at all. Interestingly same type of cap but in 120u value sounded much worse. I liked Elna Cerafine pretty much and untill I tried BG N type i though it's good. The Cerafine had pretty balanced presentation, with smooth highs, but a little bit muffled and mellow. The best cap, in my test, was 4.7 N type BG. It simply made the music alive, the detail was much better, the sound seemed to unrestricted and very natural. There was a certain coherence to the musical notes, that other caps simply didn't provide, they sounded grey by comparison. BG was very fast and uncolored and sounded real. I got good result adding 470u Cerafine in parallel, there was more midrange energy in this way, but somehow it didn't sound as clear as BG alone.
The bass didn't suffer anything when using 4.7u cap only. Quite opposite, the bass was actually more clean, detailed and very fast. By adding Cerafine, I was getting more bloom, but to me it seemed like extra coloration. It was slightly more heavy, but not as tight. I think I prefer BG by itself, although it may seem a bit bright sometimes.
All those observations are pretty much in line, with what creator of GainCard is claiming. It seems to me too, that too much capacitance slows and muffles the sound, veiling all the natural characteristics that it should reflect. I also tried no cap at all, but it didn't sound right, it was missing something (a bit of air, weight and sparkle)
PS: Regarding OsCons I see people raving about them, but here I didn't observed anything special really. Maybe they are better in pure digital circuits, although it seemed to me that after changing Panasonic HFQs to OsCons, in KC7, I didn't observed any improvement but rather not so special sound. Didn't know what to blame that on, but maybe those OsCons are not that good sounding as people claim?
Anybody has any "real" listening impressions?Attachments
Great work Peter
I settled on a combo of:
1000u Rubycon ZL,
220u Oscon,
100n, 10n ceramics
till recently when I had the "super-pair" up and running.
I replaced the 220u Oscon by another Rubycon 1000u ZL.
The Oscon "muddied" the sound and there was some low mid "honking"; I realize now I used it to mask some brightness/harshness.
Bass is great now, but there the highs are quite bright. Transparancy is better than using the Oscon though.
I just received a pair of BG-N 220u/6.3V I want to try in anti-parallel config. I might do the same (again) with the 4.7u/50V BG-N I have here somewhere.
Peter you never removed the smd's, did you ?
I settled on a combo of:
1000u Rubycon ZL,
220u Oscon,
100n, 10n ceramics
till recently when I had the "super-pair" up and running.
I replaced the 220u Oscon by another Rubycon 1000u ZL.
The Oscon "muddied" the sound and there was some low mid "honking"; I realize now I used it to mask some brightness/harshness.
Bass is great now, but there the highs are quite bright. Transparancy is better than using the Oscon though.
I just received a pair of BG-N 220u/6.3V I want to try in anti-parallel config. I might do the same (again) with the 4.7u/50V BG-N I have here somewhere.
Peter you never removed the smd's, did you ?
Hi,
My experiences are that the somewhat sharp or harsh sound caused by some electrolytics can frequently be improved by special bypass solutions. Small MKP caps like Vishay Röderstein MKP1837 10 or 22nF (before ERO 1830) clean this effects almost perfect. So the costs of the expensive BGs can be investigated at other spots…
My personal best buy solution is a Panasonic FC + 0,1uF BG NX or M-Cap Supreme + 10nF MKP1837. The very fast low ESD FC gives fundament, the BG it cleanness and the MKP overlays all with color and coherence. I use it for V_reff and any supply at any analog stage. The 220uF FC costs 50€cent, the BG about 2,50€ and the MKP about 1€ (€ almost 1:1 to US$).
Many DAC ICs are much better than thought – with this kind of cleaned supply! All SMD ceramic gremlins I had killed any natural sound! Their pads offer solder spots for the small bypass stuff...
Regards
Klaus
My experiences are that the somewhat sharp or harsh sound caused by some electrolytics can frequently be improved by special bypass solutions. Small MKP caps like Vishay Röderstein MKP1837 10 or 22nF (before ERO 1830) clean this effects almost perfect. So the costs of the expensive BGs can be investigated at other spots…
My personal best buy solution is a Panasonic FC + 0,1uF BG NX or M-Cap Supreme + 10nF MKP1837. The very fast low ESD FC gives fundament, the BG it cleanness and the MKP overlays all with color and coherence. I use it for V_reff and any supply at any analog stage. The 220uF FC costs 50€cent, the BG about 2,50€ and the MKP about 1€ (€ almost 1:1 to US$).
Many DAC ICs are much better than thought – with this kind of cleaned supply! All SMD ceramic gremlins I had killed any natural sound! Their pads offer solder spots for the small bypass stuff...
Regards
Klaus
rbroer said:
I did not remove them, I think they might be needed there, but who knows? After reading about Goudreau triplets, one might think that it's requirement, but again I tried 120 Panasonic HQ as he recommends and it was uninspiring. I'm using 4.7 BG N at the moment and the sound is spectacular, the best I have so far, I gues. This is with the new output stage. I hear music from the other room and even like that I can hear it's something special
I used 4.7 value as it was the only one I had available. It may well be that highr value would be better (but not necessarily). The resolution, clear and natural sound. On good recordings it sounds great. Few other may sound a bit bright, but it doesn't bother me much. I tried few times adding 470 Cerafine in parallel, but each time I did it I was loosing clarity and bass became more boomy with less resolution. Please also note that I'm using batteries, with a regular AC supply it may be different.I will continue my search, it never ends
I'm not so sure about bypassing anymore. Here is a pic of a silver foil cap I bought from Partsconnexion. When I tried it first time I was so impressed. The highs suddenly got much more detail and air that I started adding those caps everywhere. Later I did some "real" listening test with tweeter crossover (I spend at least 4 days for that) and came to conclusion that adding bypass here takes away the natural timbre from the sound and bypass cap always adds it's own signature that you hear. Initially it may sound better, but in a long run it seems artificial. So now, I always use a single cap. If chosen properly, it provides the best sound.
PS: Rudolf do you think I should remove OsCons from reclocker and clock circuits and put BG there?
Attachments
I mentioned previously that I was using 4,700u Cerafine across 6V battery terminal. Initially it seemed like it worked better this way, until a friend visited me (who knew my system from before) and his first remark was the chestiness to the sound, somehow it seemed very colored in a midrange region. So no caps across battery anymore. I tried to add the silver cap in this place, just out of curiosity and the sound changed in predictable way. Suddenly highs became center of attention, but I didn't like it.
---------------------------------------------------------------------------Peter Daniel said:Here's a pic of my testing setup. I installed an IC socket with all pins on ea site connected together and extention of the wires are soldered to V+ and GRND pins of TDA1543. In this way I can change the caps on a fly, without even switching the power off.
Peter
It looks as though you are suing SA or SC OSCONs and not the audio SG or newer SEP non magnetic.
Fred
I have expiriance in tube design against the sonic impessions :
one mistake continuing about the values of the power suply caps...
they are often too much higher then needed.And the sound is somehow slow
in the low mid and bass without
room beetwen the details...
probably the same thing repeats in the D
design...expecialy whwrw the speed is from vital importance...
For instance the QUAD II have only 2 x 16 micro F in PS...
one mistake continuing about the values of the power suply caps...
they are often too much higher then needed.And the sound is somehow slow
in the low mid and bass without
room beetwen the details...
probably the same thing repeats in the D
design...expecialy whwrw the speed is from vital importance...
For instance the QUAD II have only 2 x 16 micro F in PS...
------------------------------------------------------------Peter Daniel said:
The best source of OSCONs is Farnell. They do SGs (discontinued) and SEPs with OFC leads. The industrial types are magnetic.
Amazing, but you can buy OSCONs (SA and SC) in the Hong Kong electronic market (AP Liu St) for 10 to 20 cents US!
This is indeed an interesting thread but I am a bit confused.
I think fmak was asking about Vref* specifically and not about general ps decoupling caps!?
*( Vref like in Pin 19 on the PCM1704 or Pin 3 on the PCM63)
As most of you I've tried different cap types but have not done any tests on Vref only.
I think fmak was asking about Vref* specifically and not about general ps decoupling caps!?
*( Vref like in Pin 19 on the PCM1704 or Pin 3 on the PCM63)
As most of you I've tried different cap types but have not done any tests on Vref only.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.