Replace caps of Arcam Alpha 9 CDP?

Application : pulse power 🙂 ! Very good ESR for the 2.5 F cap and values after ! 2.5 Farad is huge ! The life time is not good 1000 hours at 65° ! For the digital rail maybe if below 5.5 V

Should be excellent after a battery ! You should ask to the guys of PS section if valuable for such use ! Have a doubt as smoothing cap because the low temperature and life performances ! You will need to putt several units in serie in relation to the traffo choosed !
 
Here are two other small EPCOS UltraCaps™. I also can't find prices however..

B49200-L1105-Q000 (100F, 2.3V, 9 mOhm ESR, 90000 hrs @ 25°C, D 17mm, W 34 mm, L 55mm)
B49200-F1125-Q000 (120F, 2.3V, 12 mOhm ESR, 90000 hrs @ 25°C, D 17mm, W 34 mm, L 55mm)

images


http://www.angliac.com/epcos/literature/EPC690037600.pdf

B49200-F1125-Q000 Datasheet | DatasheetLib.com
 
So I’ve done the bypass of C47 & C48 of the Alpha 9 (see attached picture) and it definitely is an improvement! The Alpha 9 and the CD23T are now closer in terms of detail/resolution and soundstage. However, something is still lacking, most notably in the bass which seems more linear and which is deeper and better defined in the CD23T..
 

Attachments

  • BypassC47C48.jpg
    BypassC47C48.jpg
    92.8 KB · Views: 182
Yes I believe some caps don't change the layout and I surmise the CD23T to have a better one and different implementation and active parts; maybe too a different digital filter !

But hey; is it not "formidable" you improve a lot the result with just 2 euros of caps 🙂 !

Are they FM or FC ? It looks like FM ? Could you give me the serie/brand of the main caps which are bypassed (c47/48 on the other side of the pcb) ! You could also swap them by Panasonic FC ! (if your bypass is FM and you believe bass or mid-bass are missing yet : go for the swap of the main C47/48 with FC and/or try too smooth the result with bypassing with FM or add some devill by bypassing with bigger value : around 330 to 470 uf with FC serie !). If your bypass is FC like I advised go further :

If you already bought c44/45 for the swap with FC : try this before modifyng what you just made : but desolder the two bypass to understand what difference c44/c45 are with Panasonic FC !

Step by step is the key, you have to go back to understand what is the better equilibrium ! After understanding if FC is better as well for C44/45 or not : you will again bypass c47/c48 to listen to how it is this time with C44/45 FC !

Only after all of this, try to swap the main c47/48 by FC (and keeping their bypass with FC also or FM if not smooth enough (at the ears) : but FM will rmove some bass and toght mid-bass : choose youir equilibrium by tests !
 
Last edited:
Those type of sonic differences could well come down to different DAC implementations. Is one bitstream and one multi bit by any chance ?
The Arcam Alpha 9 and CD23(T) are essentially the same player. Both have the (multi bit) dCS RingDac (algorithmically the same as the dCS Elgar DAC) which Arcam developed in collaboration with dCS and which was first applied in the Arcam Alpha9 CDP and subsequently in the CD92, CD23 and CD23T. The CD23T has the following upgrades relative to the Alpha9:

• A separate toroidal transformer to supply the DAC
• A different pickup assembly (KSS-213C instead of KSS-240A)
• New dual layer motherboard, with improved power supply grounding and buffering of clocks and audio data
• New Ring DAC board which reduces distortion in the audio band
• Improved master clock and clock output distribution circuitry on the Ring DAC board

However, the sonic difference between the Alpha9 and CD23T should be subtle, which it currently isn’t..
 
Last edited:
However, the sonic difference between the Alpha9 and CD23T should be subtle, which it currently isn’t..

Thanks. I was just curious what the technology was. If there are real and major differences in the audio then it might be time for some basic measurements to check LF extension (although I don't think you have changed any coupling caps)
 
Yes I believe some caps don't change the layout and I surmise the CD23T to have a better one and different implementation and active parts; maybe too a different digital filter 🙂 !
But hey; is it not "formidable" you improve a lot the result with just 2 euros of caps !
Yes, the CD23T should be a better CDP. It seems the only thing that was changed in the Alpha9 through the bypasses is the (DC) ERS (impedance?) seen by subsequent circuits. How does that improve detail (treble?) and sound stage?


Are they FM or FC ? It looks like FM ? Could you give me the serie/brand of the main caps which are bypassed (c47/48 on the other side of the pcb) !
I used Panasonic FC for the bypasses, not FM (see my post #52). The main caps bypassed are Rubicons, there’s no series information on them. It might be a badge especially made for Arcam..


I have the good hope to resolve your bass lacking problems if you follow my advices !
The main problem will be not to waste too much the trebles by working on the bass !
try what I wrote on post 67 !
There’s no lack of bass in the Alpha9, it’s just not as linear and well defined as in the CD23T. Overall, the sound of the Alpha 9 is (relatively) muddy..

I will follow your advice.. 🙂
 
Well I replaced C44, C45 and C36 with Panasonic FC and didn’t remove the bypasses of C47 and C48. Overall sound seems to be slightly improved in terms of resolution/definition, but there’s still a big difference with the CD23T. The CD23T is supposed to be one of the best CDPs out there, so maybe this is just the difference between a good CDP and a top player..

I think I will arrange another CDP, a (single bit) Carver SD/A 360, which I know well and which I like very much, but which used to be outperformed by the Arcam Alhpa 9. If the Alpha 9 is now outperformed by the Carver, I’ll know something is wrong with the Alpha 9..
 

Attachments

  • C36C44C45Replaced.JPG
    C36C44C45Replaced.JPG
    100.9 KB · Views: 148
That's a good news for few bucks your cdplayer is now safe with the swap of the smoothing caps (your first goal)...and you have a little improvement thanks to the 150 uF bypass as well!

You should remove the bypass to listen to the difference (because now you changed the smoothing caps) : step by step.
After replace the Rubycon with FC yet then listen
After replace the 150 uF with the double of this value: around 330 uF (here try FC or FM - both to choose !)
After... stop here !

Don't give up yet : the goal is not to beat the cd23t but here you listen by yourself how the reservoir caps can interact with the supply... Changing the tonal balance can sometimes improve a player by a subjective improvement of some aeras of the frequency curve (it's not about going lower or further, it's more about playing with the actual FR curve of the player): at least this the feeling it gives !

For the moment you just really remove the first smoothing caps and just bypass a little the Rubycon "reservoir" caps !

Of course some caps can't be more important than a different layout with some other active parts and such tweaks have some limits.... but you have a margin yet with the cd9... and this Learning will help you after on the Carver...

Good Luck...
 
Last edited:
So I removed the 150uF bypasses and it’s actually an improvement! 😕 The alpha9 sounds more balanced and pleasant to listen too. The 'non-linearity' in the (mid) bass is gone and there’s more space around individual voices and instruments. It might sound better than before the cap replacement.. Could it be that the new caps need to 'break in'?
 
🙂

That's a good lesson ! That's why I said : step by step, never more than one change and then a control by ears (always with the same discs!)

It just means the main Panasonic FC has changed the tonal balance, despite the Pi filter for C45/48. Don't forgett too the big 3300 uF ! : you swaped 3 caps !


If you have the courage to do it, you can now swap also the two Rubycon C47/48 by their Panasonic FC equivalent... then try a bypass (again) with a stronger value like I said above !

A Rubycon ZL here could be a candidate as well (darker than a FC and less snapy, a little like a Panasonic FM but less fatiguing most of the time !) for c47/48 ! You could try also a smd bypass with a Panasonic PPS (something around 0.22 to 1 uF on the leads of C47/48...)

This is just a try & errors technic : nothing scientific but sometimes you can get a very pleasant result ! It seems this is maybe what you have here ! The day where you will be happy with your tweak and if few caps are involved, you can go back to the original conf to benchmarck the best tweak tou keeped with the stock player (it will cost you some soldering and time, but you will be sure ! At least you have also your cd29t as a benchmark !).
 
Last edited:
So, according to the schema on page 9 of the service manual (Download service manual), I’ve replaced 3 main reservoir caps C36/44/45 of the DC power supply (not the pi-filter smoothing caps C47/48).

In fact, there are 2 separate power supply circuits on the main PCB, one apparently delivers 12V to the CD transport and laser unit and the other provides 5V to the DAC. C44/45/47/48 are part of the 12V supply. Only C36 is part of the 5V supply to the DAC.

What I don’t understand is how replacing and/or bypassing caps of the power supply to the CD transport can impact the sound. Shouldn’t C36 of the 5V supply to the DAC be much more important?
 
C36 is both a smoothing caps & a reservoir caps; c44/45 are smoothing caps; you putted résistors and cut traces before c47/48, those last are reservoir caps...

My understanding is c44/45 47/48 are involved in the output analog stages. I just have a quick look and just share with you 15 years of tweaking experience with caps on 2 or 3 dozen of CD players and DACs... I just want you to understand the subjective role of simple caps...on a modest way : tweaking by test & errors. (my experience is reading esl & esr on the datasheet is not enough ! experience and prove of concept stay important!)


But if a dac has problems in its supply, all the best analog stages can do nothing for you ! So yes , how is fed a dac chip and a digital layout is important, that's why so many chineese dac from eBay are sounding all the same : correct but not we are looking for !

Few weeks ago I improved an Aya 2 on the digital power and it cross a step sounding best to... demoniac in the mid bass with more transparence/clearness : PS and how the current return to the ground is an important thing ! Some more technician than I am could explain more than I can !
There is not on ething more important than an other, on a practical way you should ask, how to do it now ! The answer is step by step, carefully to not break your dac; one change should involve one listening test with concistence with the CD you listen to for test 🙂
 
Last edited:
C36 is both a smoothing caps & a reservoir caps; c44/45 are smoothing caps; you putted résistors and cut traces before c47/48, those last are reservoir caps...
Okay, but according to Wikipedia, the cap in a pi-filter that is closest to the (AC) source is called the reservoir cap and the cap closest to the (DC) load is the smoothing cap (see: Capacitor-input filter )..

My understanding is c44/45 47/48 are involved in the output analog stages.
Yes, that would explain the sound impact of bypassing C47/48. I’ll look into that..

I just have a quick look and just share with you 15 years of tweaking experience with caps on 2 or 3 dozen of CD players and DACs... I just want you to understand the subjective role of simple caps...on a modest way : tweaking by test & errors. (my experience is reading esl & esr on the datasheet is not enough ! experience and prove of concept stay important!)
I certainly respect your experience! And I thank you for letting me experience the importance of simple caps.. 🙂

But if a dac has problems in its supply, all the best analog stages can do nothing for you ! So yes , how is fed a dac chip and a digital layout is important, that's why so many chineese dac from eBay are sounding all the same : correct but not we are looking for !
So are you saying that if the analog output stages can be improved by improving their power supply, then there can’t be anything wrong with the DAC power supply?
 
Last edited:
No, I know my English is very bad but ask the guys here at the powersection,they are better than Wikipédia 🙂, or here of course Mooly ! the first cap smooths the current (peaks are flater) for a DC output ! the second is called reservoir because it feeds the load(s). Note with c36 it makes both jobs... but helped with local caps near the active devices....

I just said improving only an analog stage is not enough; the power section of the digital devices is important as well as this is what see the following stages see !


Ah sorry... time to eat !