AYA 2 list mods
Hi,
Dug out the AYA 2 and here's mods so far , I wondered if ive missed anything or should change anything especially caps for TDA1541A
Signal and I/V Allen Bradley, the I/V measures 1k7 maybe too high ? I may try TX2575
C407,408 1nf polystyrene
C401,402,403,404 changed to 0.1uf BG NX Hi Q
C302 ,305,308 changed to 0.1uf BG NX HiQ
C306,309 unpopulated
C405,406 unpopulated
C303 original 0.1uf ceramic
C3 1200uf 35v FC bypassed with 220uf FM
Output taken before C409,410
Regulators for OPA861 , TDA1541 +5v and -15v replaced with Paul Hynes shunts (-5v left as original)
I would expect the 1k7 I/V resistors to be fine. The swing will be +/-3.4V (nominal, although it is good to consider it a bit higher in practice), and the OPA861 nominal voltage compliance, with +/-5V supply, is +/-4.2V. If you still worry about this, you can increase the OPA861 supply voltage to +/-6V.
Polystyrene for C407/C408 is definitely fine.
C401-C404 value can be tweaked to suit your regulators. Ditto for C302 and C308.
For C305, a 0.1 uF may be also fine, and Eldam’s C3 mod is still nice, but taking into account all the other changes, I would suggest checking and ultimately tweaking these parts for the best overall results (tonal balance in the first place).
Yes, C405 and C406 are best omitted.
Direct output coupling is generally possible, but I would advise not to make a short on the output in this case.
Thanks Pedja , this is still a great sounding dac , the type of decoupling caps for TDA1541A seem to have big influence so I'm hoping to find best suited in those positions . Do you still use 100nf ceramic fitted to +/-5v and -15v beneath chip ? Do you have any preference ?
With WM8741 most caps tended to sound harsh mainly electrolytic with smaller ceramic bypasses ,I preferred 10uf smd ceramic on chip pins for 3x +5v AVDD and 1 x DVDD , 47uf BG NX High-Q on VMID L&R
The AYA 5 and S5 do use local 100nF bypass at TDA1541A supplies, and the parts used are Panasonic polyphenylene sulfide, after the Panasonic OS-Con main after-regulators caps.
Regarding the WM8741, it has been a long time since I worked with it, and designed the DAC for another manufacturer. The WM8741 did not ultimately sound harsh to my ears, but the oversampling filter type was the major contributor to its sonic performance (and some filters can sound harsh, indeed), probably more significant than the supply capacitors. As for the caps, I used Panasonic FM and PPS. Once you set the oversampling filter to your preferences, I guess the capacitors will be less critical.
If I had to spot WM8741's weaknesses, it would be a certain lack of the attack and natural sonority, but for me it is thus with the majority of sigma/delta D/A chips.
AYA 2 already sounds more realistic compared to the WM8741 based sac to my ears .
I have added SMD PPS 220nf onto the +/-5v and -15v pads bottom of board , no pre caps
I've ordered a few more 1k5 ,910R and 150R Allen Bradley to try and match them better , the only ones I could find was 0.25w CB1525R military grade , not sure if these are as good as the others ? I have also ordered a pair of 1k5 TX2575 to try , never used those before but can easily swap back to the AB's if no good.
I have also ordered a couple of 1000pf LCR Components EXFS/HR Polystyrene to try instead of the axial ones (these was recommended) , I can spread the legs maybe extend one if needed .
I use CB1525R, they are fine for me.
Polystyrene caps became scarce these days, but it is not necessary to use exactly 1 nF here. You can safely use lower values , or a bit higher too.
If the CB1525R are fine for you they will certainly be fine for me ! The ones I currently use are 0.5w non mil spec but not as well matched as I’d prefer . I’ll update how I find those TX2575 , at least I have the mil spec AB’s to go back to if needed .
Yes it’s a shame the old good parts are becoming hard to find now .
I've fitted the TX2575 1k5's , these have made quite a dramatic difference ! I was expecting them to sound bright but they are uncoloured , very detailed with lots of air and space . Are they better than the AB's I had in ? well they are very different I wouldn't say better . I then removed the 150R and 910R standard AB's and fitted the smaller 0.25w mil spec AB's. The sound is different again , very smooth a little coloured and warm. Do the mil spec AB's take time to burn in ? The original AB's fitted had many hours runtime but these newer ones values are better matched . I'll leave it powered up running test tracks for a few days . I also have some new mil spec 1k5 AB's incase I don't like the TX2575
To be honest, it was many years ago since I settled on the AB, and I can not remember if I heard any sound change over time that was related specifically to these parts. My experience is that it is the capacitors, and partially semiconductors, that take burn-in. Resistors not that much.
As for the TX2575, IIRC our Gramofone days, Malcolm liked these resistors, and he did like pleasant-sounding stuff. These are of course way higher-priced parts, but the Gramofone was a sort of cost-no-object design. We did plan to use these at the signal path in a later revision of the amplifier GF01, but it's regretfully never been released.
I've modded my AyA almost to death. Lol, they key to maxing AyA's capabilitise arent at the decoupling stage.
Main PS caps is the key. I do not use decoupling at all but am unable to get away with it on Op amp stage so had
to use a combi of nichicon KZ & BG NX. For i/v I've tested many including Rhopoint but the best sound to be had are
2W Shingkoh Tantalum. Many ways to skin a cat depends on individuals & listening taste, Oh I do not use output
caps but use a 600-600 ohm line opt, this too brought the sound to another level, best part only cost me $20 from ebay,
trafo came out of a sounds analyser of sorts cant recall & it could handle 60ma of dc current
Cheers