Forum breadcrumbs - You are here:CommunityGeneral discussionsAYA psu mods
You must be logged in to post to the forum.

AYA psu mods

12
Quote from Guest on 09/05/2019, 23:24

(...)

Re: Pedja's response, if it is okay to ask the 'tweaking of reservoir capacitor cap value and type' for alleviation of 'un-forgiving' sound, does this follow closely enough with Mr. Eldams recipie with FM 330uF + FC 1200uF combo?

(...)

 

Yes, but the final success with local bypass cap removal will still depend on your system.

In my own AYA II sample that I was using for these trials, an FC 1200 uF remained ultimately the only reservoir cap there. So there is no FM 330 uF. But the local bypass was not ultimately omitted either, and it is BG NX Hi-Q 0.47 uF there.

 

Thanks Pedja.  Any one with 0.47uF BG NX HiQ to swap for 47uF or 22uF 6v3 types?

Quote from Guest on 03/06/2019, 07:48

Thanks Pedja.  Any one with 0.47uF BG NX HiQ to swap for 47uF or 22uF 6v3 types?

New old stock

The best results that I got is to parallel the 1200uf with nichicon KZ 22uf 50 v caps. Fc alone has a more forward sound & crash of cymbals or percussion instruments not so realistic & does not bloom. Beware though with KZ never use 25v version, they don't sound good.  The beauty of AyA is every part of PS does contribute to SQ. If your like me a compulsive tweaker, it is a delight & a learning process as well.

Cheers

Yes, that is interesting, different voltage caps can sound different. I am however not sure about any rule here. For instance, Elna Silmic II usually work better if they are lower voltage.

There really is no rule cause looking a spec sheets alone will not tell you the results Pedja. The only way to know is to try. The Kz 22uf 50v that I mentioned is the sweet spot of the range going higher value sound changes becomes less natural & dynamic. Silmic I only use ARS don't like Silmic ll.  Speaking of which yes lower voltage Silmic sounds more dynamic but dryier sounding & highs not so good. On my AyA there's at least 6 different types/brands of caps. All there to create the sound that Im looking for. Though I cannot design, I've applied lots of wacky tweaking technics on AyA that has been proven to work very well. Probably when you see it in detail you'll scream heresy Pedja. Lol

 

Cheers

Hello all,

I hat not received notifications but connected here to shop some news about the AYA and tweaks.

I noted some questions and inputs about my -15V tweak and I' happy to see many adapted it according his own hifi. So Eldam there to add some precisions or answers.

Yes the main improvment is about this particular FC cap first as noticed Pedja Rogic and this precise one as I tested dozen & dozen to find what could suit to my hifi & tastes without waste the musicality of the stock AYA 2 2014 edition I have.

Sorry in advance to speak about capacitors sound, it is just for convenience and shortcuts in the discussion as we all know caps have no sound of their own but have "colors" in a circuit according their technical specs, uses and constructors  production - dielectric formula, comodities, etc.. A Rubycon is not a Nichicon, nore a Panasonic and vice versa, I play with dozen and dozen of brands and the game is to try to the one that works fine for you. And frankly no polemic here, it is a part of the design that is long to add the last flavors and a possibility to adapt to his own hifi, and I just post in this context as of course what is good to me it's maybe not to you and the opposit may als be true, halas one has to try to know : different speakers and devices and room... et tout ça, make difficult this last necessary changes, at least it never solve a flawed design or a bad hifi/room ! Mine for the discussion and permitt to understand/adapt what you have at home, is: speakers Boston Acoustic Lynnfield 400L (sealed design, very nervous and fast and clear/informative) slighty modded to get off an excentricity of this high end Boston Acoustic aluminium dome in the 2500 hz-4000 hz. Amp is a Chord, pre is a Yamaha C2 -full dicrete made- that make an happy combo with the very fast and powerfull neutral but detailed amp (this amp has really). wires are Oyaide for the Amp/pre & Aya and speakers are Cardas. If I compare with my Kef 104/2 ref, I will say the Bostons are way above, I can copare because both are bass band pass cabinet and needs tons of "ggod" watts to sing. Sorry for the disgress.

The Panasonic FC cap I like for its snap I already noticed as a decoupling power supply in some amps is precisly the FC 1200/50V large diameter and short heigth. Cause there are two 120/50v FC models in the Panasonic catalog and yes they gives not the same result at ears at home when I tried both, second is, you guessed, body less width diameter but higher length.

I keeped a FM for its smooth in upper bass lower mid despite I never liked the mid-highs of FM caps : most and in most designs give to my ears an harshness in this focussed band - often lower treble-. The combo was good in my hifi though. I didn't liked the Nichicon KZ combo of Jaffrie mate, but have no doubt it gives for him a better result with his hifi : it's always a particular situation. For me the supress of the 100 uF Rubycon gives all the snap and clarity of the main  FC/FM, just needed as HK noted a work to choose between 0.1 uf and 0.4 uF at home. I'm not surprised anyone has his preference there, I have mine. I tried : BG NX 0.47 uF, BG N 33 uF/16V with no better result than without but keeped 0.2 uF on the smd pads at the end. I still have to try the post reg BG NX cap without the main // FM 330 uF Pedja is liking some posts above. Anyway first time of my life getting off a Black Gate gave better resut... certainly speciic to my hifi, I just don't know ! But the last cap decoupling was also a tricky one to find and certainly not universal, so I will not testimony on it.

Yes the caps voltage matter and I agree with Pedja's conclusion : no precise rule, try & error is your friend, halas time consuming to find and one must have tons of caps to play this game. To answer to HK's question : I didn't tried the FC 1200/50V short body on the two 5V rails of the TDA but on the analog opa861 rails and there they didn't stay ore than one hour of test ! The two caps I use for the TDA 5V rails are not FC, nore BG, nore FM and are from two different voltage : and that's strange : they are not switchable, a particular voltage cap gives a better result for each of the two 5V pins. That means the +5V & the -V has no the same exact purpose (decoupling or coupling?) inside the TDA.I eventually preffered C-Core traffo from the EI and the filter EI maybe because of this as well... I just don't now, my ears are the last judge at home and I know it is subjective.

Being not a specialist and how currents flows I eventually bring together that find with the one of the main -15V in the power supply tipology (serie reg follower, don't know the name?) of the AYA, Pedja improved year after year and that sounds best that any modern integrated reg for the TDA1541A. I believe the carbon resistor marry well here and play well with the fast flow of current and the particular clearness of the Panasonic FC serie, serious cap that has automotive grade & that's for the one who think Panasonic makes craps... just ridiculous assertion I read on forums. Too bad imho some in the AYA construction sapped them for metal type cause they think the Johnson noise is better... my baby ears just don't care ,comme dit Simone (Nina).

So yes I just lastly decouple the - 15V rail after having spent hours of tries with the three smd pads. My simplist view of it is the three votage rails inside the chip in my hifi context (my hifi devices + not big ground loops as I have no main Earth in my living room electricity supply... uH !). So no last decoupling cap for me for the two TDA 5V rails : the good Pedja Rogics regs play faster and with more clearness and no harsh there, at least at home, YMMV. I finally asking myself all is about the ground perhaps and sorry no technoligical background I have, just guess and 30 years of enthusiast only, playing with the TDA1541A chips. With all of that, the AYA 2 2014 DEM clock mode option was muddy on my board so I don't use it : at home it get off the mid snap I a found of and I don't find with Delta Sigma chips most of the time.

AYA 2 still playing right now and just regrett not to have tried AYA4 mode with integrated USB and the new 22/24 crystals Pedja just found I have been reading today 🙂 ! On Stereo AU net ezine, a guy said the Aya4 was better than the Aya 2 DS. My understanding is he had what I have : simultaneous mode, fifo and clock from Iancanada with WaveI/O usb before -though mine are tweaked on caps & power supply for a far better result and I concur with HK's testimonial than batteries, even LiPoFe, is not my cup of tea-

Long post and story and disgress but it mates with the long one to find those tips...and eventually explain all is relative to the context. Though what I'm sure is the FC 1200/50V without the post Rubycon 100V (tried both the old and newest version) on the -15V is where tweakers should start with and tweak the rest from there... best surprise for me is no Black Gate is really needed for me. Stock DIY AYA BOM version is certainly the good compromise for everyone average hifi and it's certainly a longer to find trade off ! I wisch I have had the old famous no made anymore main power supply blue cap Pedja used in the soonest AYA version... 🙂

cheers all, a blink eyes for the cool story we shared with Cheung T. from HK, Shane C. from Adelaid elsewhere on a forum and that was conclued by a letter to Pedja to remake a board with some of our wishes... uf-l was mine iirc, Shane and Cheung was already proud owner of Audial AYA dacs before 😉

another blink eyes to thanks Pedja to have imputed my tip on the manual and for all this story made possible, the first class technical advices and the caps exchange with Jaffrie from Singapour 😉

As we say in France, not better camembert than a shared one...

Thank you too, for your post.

When choosing passive parts, my nominal goal is "not to waste what is already there".

When it comes to the signal path parts, it is actually a logical step: you omit the part completely (if possible), and then look for the part that is sonically the most invisible.

With supply decoupling, it is a bit different, since there is no reference point to start from. So it may be a place to do some fine-tuning, but the bad part can really ruin the otherwise good device. A bit of intuition is needed here. For me, what makes things easier is (my) requirement to make the final sonic balance as neutral as possible, since the device should (hopefully) suit different systems. Admittedly, this is my way of thinking most of the time. 🙂

Speaking about this, I recall one of the most useful thoughts on the audio I ever read. It was decades ago, so I don't know where it was, and don't even know who wrote that, but it was something like this.
"One of the most important points of the good audio device is its ability to sound fine with a wide range of other equipment." 😉

12