Just one question ...in the 2 middle circles should I replace blue caps or blue and red ones as well? ... and 470uF ? is that what was already on the board or what you had in supplies ?
Cheers
Just replace the electrolytics with the oscons. In the kit they give you 10uf.
I suggest you put 470uf oscon. But I guess any value oscon you have will be OK.
Be sure to give it 22000uf on all rails!
Last edited:
Couple of my friends have Buffalo II DAC that uses ESS chip ... although I havent heard it myself Ive heard good things about it... but it depends on what system you plug it in.
I also prefer analog/less detail to digititis/fireworks detail , anytime.
Will do, i have 400.000uF in my amp ... Im sure it wont mind if I borrow some
Peace!
I also prefer analog/less detail to digititis/fireworks detail , anytime.
Will do, i have 400.000uF in my amp ... Im sure it wont mind if I borrow some
Peace!
Couple of my friends have Buffalo II DAC that uses ESS chip ... although I havent heard it myself Ive heard good things about it... but it depends on what system you plug it in.
I also prefer analog/less detail to digititis/fireworks detail , anytime.
Will do, i have 400.000uF in my amp ... Im sure it wont mind if I borrow some
Peace!
You will like this AK4393 DAC.
400,000uf!! ROCK ON
Are these as good as Oscons:
http://cgi.ebay.co.uk/5PCS-470uF-16...907?pt=LH_DefaultDomain_0&hash=item51946f57c3
Thanks.
http://cgi.ebay.co.uk/5PCS-470uF-16...907?pt=LH_DefaultDomain_0&hash=item51946f57c3
Thanks.
Last edited:
Sanyo makes a few lines of OS-CON capacitor. Ranging from average ones to the top of the line SEPC variety.
NCC PSA or PSC are quite good, and if you are on a budget, they would be my second choice of capacitor. With OS-CON SEPC being my first choice.
The NCC PSA gives a bit of a fuzzy sound in the mid bass region, and a slightly slower overall sound. The SEPC is lighning fast and crystal clear across all frequencies.
The NCC PSA is still a large step up in performance from a regular electrolytic capacitor.
Many thanks for that Erin - very helpful indeed!
These look like the correct items (much cheaper than I thought too... hope they're not fake!):
4pcs 16V 470UF SANYO SEPC Aluminum solid Capacitor | eBay UK
- John
These look like the correct items (much cheaper than I thought too... hope they're not fake!):
4pcs 16V 470UF SANYO SEPC Aluminum solid Capacitor | eBay UK
- John
Corret ones, but those ones have very short legs and will be a pest to install. you need to get ones with long legs.
Sanyo OS-CON is slightly larger and heavier than NCC PSA and has a different angle bend in the bottom of the outer can, and has a darker shade of grey/silver paint on it.
It appears that the ones being sold on ebay are older stock from 2008 and becuse they have some age on them, manufactures or suppliers are selling on the surplus market. I have no reason to suspect their originality.
In case anyone was wondering;
NCC PSA or PSC 470uf 16V take about 24 hours and a few power cycles to run in. They sound a bit fuzzy when you first put them in and there is little bass, however that all changes once they are run in.
Osocon SEPC 470uf 16V takes about 4-8 hours to run in, they also suffer from a little fuzzyness when first installed and slightly lacking in bass, but not as bad as NCC PSA.
I'm currently trying some 1500uF 6V Oscon around the DAC chip and so far the treble and mids are nicer, but the bass is a bit wooly and I'm hoping this will change tomorrow after 24 hours of run in.
NCC PSA or PSC 470uf 16V take about 24 hours and a few power cycles to run in. They sound a bit fuzzy when you first put them in and there is little bass, however that all changes once they are run in.
Osocon SEPC 470uf 16V takes about 4-8 hours to run in, they also suffer from a little fuzzyness when first installed and slightly lacking in bass, but not as bad as NCC PSA.
I'm currently trying some 1500uF 6V Oscon around the DAC chip and so far the treble and mids are nicer, but the bass is a bit wooly and I'm hoping this will change tomorrow after 24 hours of run in.
Last edited:
I'm waiting too this dac from Hong Kong.
Meanwhile I've started studying it and made a mod hypothesis.
Attached are the possibile BOM and a nice photo edited with PS and signal paths...
All bypasses are experimental, too much times I've found that no bypass could be better...
Erin, you used big value Oscons everywhere but while C28, C30 and C33 are good canditates for Oscons, all analog positions (C32, C33,C35) would benefit, IMHO, from audio parts like Silmic, Cerafine or Muse FG/KZ.
C35 (VCOM decoupling) should not be much bigger than 10uF, its value is critical for internal compensation of AK439x.
Try to remove bypasses where you used Oscons, they don't like to be bypassed...nasty resonances are certain...
Meanwhile I've started studying it and made a mod hypothesis.
Attached are the possibile BOM and a nice photo edited with PS and signal paths...
All bypasses are experimental, too much times I've found that no bypass could be better...
Erin, you used big value Oscons everywhere but while C28, C30 and C33 are good canditates for Oscons, all analog positions (C32, C33,C35) would benefit, IMHO, from audio parts like Silmic, Cerafine or Muse FG/KZ.
C35 (VCOM decoupling) should not be much bigger than 10uF, its value is critical for internal compensation of AK439x.
Try to remove bypasses where you used Oscons, they don't like to be bypassed...nasty resonances are certain...
Attachments
Hello Clave, thank you for your nice work on this. It must have taken you some time.
Can you tell me, are your comments regarding (C32, C33,C35) using silmic, based on your own listening results? Or are you speculating that becuse its the analog part of the DAC that it will sound better?
C35 (Vcom Decoupling) is indeed stated in the AK4393 datasheet that this capacitor is used as a filter to remove high frequency switching noise. The datasheet recommends 10uf bypassed with 100nf. I think that there is no problem using a bigger cap, as it will still filter out the high frequency noise, though perhaps its a waste of a bigger cap?
Do you have any listening results to confirm that maximum size should be 10uf?
Yes all bypassing is experimental. What I recommended is based on listening results. It is a successfull recipie. There are no nasty surprises to the sound, in fact very nice sound. But I totally agree that the wrong bypass can totally wreck the sound. I have done much experimentation with this, on this dac and on others.
Personally, and no offence intended to anyone, but I do not like the sound of Muze KZ capacitors. I have never found a good location for them. to me they always sound cold and lifeless and have no bass, and never seem to "run in". I think that Panasonic FC is the most over rated capacitor in existance. To me it is not special in any way, it has a grainy and wooly sound and its HF is very restricted. I would prefer to use a generic Jamicon capacitor ahead of either of those. But Silmic is very nice across all frequencies.
I look forward to your answers
thanks
Erin
Can you tell me, are your comments regarding (C32, C33,C35) using silmic, based on your own listening results? Or are you speculating that becuse its the analog part of the DAC that it will sound better?
C35 (Vcom Decoupling) is indeed stated in the AK4393 datasheet that this capacitor is used as a filter to remove high frequency switching noise. The datasheet recommends 10uf bypassed with 100nf. I think that there is no problem using a bigger cap, as it will still filter out the high frequency noise, though perhaps its a waste of a bigger cap?
Do you have any listening results to confirm that maximum size should be 10uf?
Yes all bypassing is experimental. What I recommended is based on listening results. It is a successfull recipie. There are no nasty surprises to the sound, in fact very nice sound. But I totally agree that the wrong bypass can totally wreck the sound. I have done much experimentation with this, on this dac and on others.
Personally, and no offence intended to anyone, but I do not like the sound of Muze KZ capacitors. I have never found a good location for them. to me they always sound cold and lifeless and have no bass, and never seem to "run in". I think that Panasonic FC is the most over rated capacitor in existance. To me it is not special in any way, it has a grainy and wooly sound and its HF is very restricted. I would prefer to use a generic Jamicon capacitor ahead of either of those. But Silmic is very nice across all frequencies.
I look forward to your answers
thanks
Erin
Hi Erin, you're welcome
Yes it took me some hours but I needed it to understand which caps to swap and what can be improved.
No speculation, I've modded several CD/SACD players and I've learned that Oscons should be used with care (and without ANY bypass on them)...
For example I've had great results from Oscons on digital supplies but on my SCD-555ES they worked very well on VRef but on DVDD the original high value Nichicon FW was way better.
On analog supplies I've had better results from Silmics, Cerafines and even KZs
See:
http://www.diyaudio.com/forums/digi...-board-design-support-needed.html#post2532801
and:
http://www.diyaudio.com/forums/digital-source/25213-capcitor-type-vref.html#post300396
Here good candidates are, IMHO, 10uF FMs, Cerafines, BlackGates, Oscons
Not yet... but as I've done with MyRef I'll mount sockets everywhere for quick and safe component swap.
I don't doubt that as is your dac sounds great but I suspect that without Oscon's bypass it could sound even better.
Oscons without bypass is a must, resonance are proven:
(Oscon 100uF in green Oscon with ceramic bypass in blue)
This plot is from this great thread (sorry, in Italian but Google translate could help) which demonstrates that most caps don't like simple bypasses and that PPS bypass are very very good.
I completely agree but one exception: in dac's analog supplies KZs can sound pretty good.
Last edited:
I suspect that without Oscon's bypass it could sound even better.
Oscons without bypass is a must, resonance are proven:
(Oscon 100uF in green Oscon with ceramic bypass in blue)
This plot is from this great thread (sorry, in Italian but Google translate could help) which demonstrates that most caps don't like simple bypasses and that PPS bypass are very very good.
I completely agree but one exception: in dac's analog supplies KZs can sound pretty good.
Hi Clave,
Thanks for the link. I have looked over it and it was interesting.
In case you were wondering. I use PPS 220nf for all my bypass caps, except for bypassing the 470uf on the op amp, where I use (33nf // 33nf) 66nf KP1830.
I do have some KZ 100uf capacitors, but I tried Nichicon FW (which I like better in analog circuits [but could be wrong in this instance -perhaps I should have tried KZ??]) in the location of C32 and C33. Sadly it was a very bad sound -flat lifeless and slow (sorry!!
).I then tried Oscon 47uf in C32 and C33, and it was too bright, not enough bass and a bit normal and boring sounding.
I then went back to 470uF oscon SEPC and found the very nice sound again.
I have not yet tried changing the Vref capacitor. Because, to be honest 470uF sounds very good.
I can try a 47uf or 33uf oscon here. I don't have any 10uf oscons.
Last edited:
I might just add that I also bypass this 470uf Oscon with 220nf PPS capacitor.
Bypassing makes all the difference with much smoother HF and a more rich sounding midrange.
The most unfortunate thing is that these capacitors are very expensive, you can get them at RS and Farnel, but they cost too much
I might just add that I also bypass this 470uf Oscon with 220nf PPS capacitor.
Bypassing makes all the difference with much smoother HF and a more rich sounding midrange.
The most unfortunate thing is that these capacitors are very expensive, you can get them at RS and Farnel, but they cost too much