Upgrading Philips DVD 963 SA

In this thread I would like to share knowledge on upgrading the Philips DVD 963 SA (CD-upsampling / DVD / SACD) player.

This player is interesting because of its rich list of features and its low price.

This price has also led to some serious shortcomings, as observed in the other thread: http://www.diyaudio.com/forums/showthread.php?s=&threadid=10224

Here I will start with posting some links and some personal experience with this player.

Lucas
 
From the above sites there are a few things to observe:

The chinese site gives us an example of a very thorough upgrade. They have added a seperate linear PSU for the analogue circuit. They seem to have left the switching PSU without changes. They added new clocks, and replaced the opamp circuit with a discrete output.

The Audiocom site concentrates on adding new clocks, and (like most others) replacing the capacitors with Blackgates.

They had a list with CPU-parts that were being substituded, but that list was not very accurate. I wrote them the following email:


>"When checking the inside of the machine, I found a few errors
>your part-listing:

>Philips Partnumber:

>2414 I could not detect this one... Where is it located?
>2413 This is originally a 25 V 100 mfd and not a 16 V 470 mfd
>2628 This is originally a 25 V 47 mfd and not a 25 V 4.7mfd
>2213 This is originally a 50 V 4.7 mfd and not a 50 V 10 mfd. Is it substituted indeed by a 6.3 V 47 mfd?
>7201 This is not a capacitor but the AD 1895A . Maybe you mean the closeby partnumber 2206 which is indeed 50 V 10 mfd.

>The rest of the capacitors were correct.
>However, 2213 is located on the picture-pcb. Is this on purpose?

>Furthermore you mention partnumber 7801. I could not find it. >There is a 7401, which is a LM833 opamp, not a LM338.
>By the way there are two of them. One is close to the (inferior) analogue output (not the superior front-channels, that use AD 1895A opamps.
>The other LM833 sits far away on partnumber 7802. Do you leave this one untouched?

>Finally, I am surprised that you leave those two Nichicon Fine Gold 16 V 100 mfd output-capacitors (partnumbers 2421 and 2429) on the two frontchannels untouched....
>Are they that good that a BG could not bring improvement?

>One last question: You exchange the clock oscillator (partnumber 1102, frequency 33.8688) for the Superclock.
>This is the clock for the SACD part. Why don't you upgrade the other clock (frequency 24.576) for the CD part as well?
>Is improveing CD-playback not one of your goals?"


They anwered with the following email:


>"Today we have amended our website to include the revised lists for the Philips DVD963SA.
>It seems that the model we used for the parts listings was not a European version and many of the parts numbers differed.

>The revised listings will provide answers to most of your questions.

>According to our data the 33.8688MHz clock is the master clock oscillator for both CD & SACD. The 24.576MHz clock is for upsampling only."


What they didn't answer is why they don't upgrade the two output caps on the two frontchannels.
With regard to the clock, they seem to be fully right. However, how important is the 24.576 MHz clock for upsampled CD sound?
 
Last week I had decided to bring both my Philips DVD 963 SA and a Philips SACD 1000 (owned by a friend) to the Dutch audiofirm van Medevoort (www.audioart.nl)
They did an upgrade to both machines costing 200 Euro for my cheap player and 300 Euro for the expensive heavyweight.

They have info (albeit in Dutch) on their mods on this webpage: http://www.audioart.nl/nl/upgrade.html

I collected both players last friday and did some first listening tests.

The 963 SA sounds a little bit better now. There is a little more bass, but it has also become somewhate less controlled. The treble is bit more relaxed and softer.

The SACD 1000 sounds gorgeous now. It has an open, transparant sound, and yet it has a tonal balance that is much more convincing, because it seems much more complete (bass, mids and trebles are balanced in a single whole, without being teared apart). The bass is also extemely good. It is for me the first time that I am somewhat convinced by SACD. I already loved the quality of bass and midrange of SACD, but now I can also live with its treble.

The 963 still has a somewhat pronounced treble that seems to shine over a rather veiled midrange. It doesn't show the power and the guts of the music because of its thin lower midrange.

I guess the two main reasons for the 963 sounding less musical and with less coherence, is the lack of a good linear powersupply for the analogue section and the use of a poorly fed opamp, instead of the well made discrete output of the SACD 1000.

I think both items need to be addressed in order to transform it from a detailed, clean, but rather technical sounding player into something sounding more musical and involving...
 
Yesterday, I took the time to check what Audioart had actually modded to the players.

They have done the following:

1. They bypassed most of the caps on the audio-pcb with ERO 0.33 (63V) caps (MKP's I guess). In the 963 I counted 34 of them.

2. On the output caps they exchanged the standard 100 mFd Nichicon Fine Gold with 2 pieces of ordinary Nichicon 470 mFd, combined with a 470 mFd cap that looks like a tantalum, and again bypassed with the little ERO caps mentioned before.
The total capacitance was therefore raised from 100 to some 1440 mFd.... Can anyone comment on the advantages / disadvantages of such a huge raise?

3. They added about 5 strips of a kind of bitumen, glued on the surfaces of the enclosure, to damp vibrations.

4. They glued a sticky white stuff on the existing clocks to damp vibrations.

5. On the SACD 1000 they exchanged the diodes in the linear PSU with other diodes.


All in all a rather surprising mod!

Of course one cannot expect the use of too many expensive parts with a mod that costs only 200 or 300 Euro.

Yet, I cannot understand the use of such inferior Electrolitic caps in the output of the frontchannels. They should have chosen at least a good Panasonic FC if they find a blackgate too expensive.

I guess that they have spent about 10 Euro for the material in my 962SA and about 15 Euro for the SACD 1000 mod. That is a meagre affair indeed (at least for the customer). I will ask Audioart why they choose this path...

The bypassing of bad caps with little MKP's does bring some improvements, of course, but the real important issues are not touched at all by this mod... All in all a mod that leaves the customers with some feeling of dissatisfaction!

So, I expect that I have to do a lot of work here by myself, and it seems a good thing to exchange experiences with others.

One could also decide to built a very good DAC, like Guido Tent and others did. However, the ability to play SACD as well is a worthwile bonus, which makes upgrading the 963 all the more attractive. The other advantage is the lack of need to address the timing-problems one has to confront with a seperate drive and DAC.

My next step will be exchanging the output caps with BGs.
Then I will try the upgrade of the large cap in the switching PSU.
See this posting: http://www.audioasylum.com/audio/tweaks/messages/87613.html

I am quite curious whether this mod will really improve the SPSU dramatically?

Then I will consider building a seperate Linear PSU.
For this, I will have to know what voltages are required for the analoque output stage.

Finally I would like to know whether a 230 Volt filter box at the mains inlet would help. (At least it would prevent the nasty SPSU to send HF rubbish to my other gear...)

When I have new findings I will post them.

Bye for now,

Lucas
 
Philips DVD963SA

Hi Lucas,
Woooow, you spare no effort to reach Audio Nirvana!
Pesonally I think van Meedevoorts mod is far too expensive. The SACD1000 is a different player, also uses a different DAC (Crystal I believe) and Hidam modules.
I tried to translate the Chinese site with Altavista-Babelfish but this gave a laughable result:

"Philips Dvd963sa Modified
with Tubefilter and M2XO

By Hatasa

What ¡èW¡õ~¡õ~ ¤¤ Tau the knack ¡õ| §ï ¡õl¡õx field ¡õ¡õPhilips DVD962SA¡õA step ¡õ writing brush { does the step base peep ¡õ the navigation T¡õW¡õ takes A §Ú ¡õt¡è@¡õI¡õK¡õR¡èF¡õI¡õu¡õO ¥¿ satirizes §Ú ¡õQ the active ¤â ¤§ regret ¡õA¡õs¡è@¡õN¡õ Han ¡õ abdomen 963SA¡èS §Ö ¡õn collapses ¡õx¡èf¡õa ¤ß ¡õQ the Lai Jianxi Egypt ¡èFVideo ¤§ ¡õ~¡õAAudio ¤è ¡õ collapses pulls sparkling jade-like stone S Tau ¤° or ¡èj¡õ startles ï° loses a¡õI¡õ¡õ¡õb¡õ@¡õ scratches with the finger U¡èjForum¡èW stomach ¡õ ¹¾ wifeless S Tau ¡õX¡ìQ¡õ`¡õA¡èSupsampling traces ¡õK ¤£ ¡Ào ¤£ the list ¡èf¡õf ¤ñ ¡õ ì© w¡õ is cold ] ð¥ cX¡èf¡õ¡õ963sa¡õ the full @¡è@ ¨è ¡õa¡õt¡õo Tau ¡è@ bureau tau §Ö ¤§ ¡õ navigation Mo province A¡õi¡õO anchors ¨£ ¡õw¡õq¡õ the Dan Hantan B ¤Í ¡è@¡õ infant S¡è@¡õ camp harms bForum¡èW Er Er ¡õt¡õt¡õA The arc does obeisance ¡õD¡õp¡õ óª harms h¡õ the umbrella A §Ú ¤ß citrus reticulata Tau ¡õI ¥¢ distinguishes ¤§ ¡õl¡õA¡õu¡õn¡èS the list ¡èf¡õi" etc.


Well at least we now know Hatasa wrote it. Maybe it is even the same Hatasa that posts on this forum. Any comment Hatasa? I am very interested in the part "stomach-wifeless":clown:
Helas, I have no experience with the Philips but have some with a Sony SCD-1. I tried three different clocks: KC, LCaudio and Guido's in Allen Wright implementation but despite much better definition in the sound with a good clock I don't like the sound SACD at all. It's simply is no music to my ears.
I just acquired a Philips DSR2000 digital satellite receiver. With my DAC [AD8561-CS8412-TDA1543 (NON-OS) I get MUSIC ]
You understand I am extremely puzzled by this result!:confused: :confused: :confused:
It's even MPEG !!:bawling: :bawling: :bawling:
 
:eek: What happen there????

:wave2: Becoz my artcle not writing in Chinese... It is with the Cantonese.... Funny!!! Hahahahahahahahahahahaa!!!!!

:cannotbe: Ok! I will have the new review on my own 963SA this week. This time also will perpare the english version at the hidden link for yours! Becoz many of you sent me the e-mail for the request on it....

:wave2: Please give me few days there.... I've just got the toothache and need to meet the dentist tomorrow! :bawling:

Anyway, thanks for the attention on my review!! :clown:
 
Re: Philips DVD963SA

Elso Kwak said:

Helas, I have no experience with the Philips but have some with a Sony SCD-1. I tried three different clocks: KC, LCaudio and Guido's in Allen Wright implementation but despite much better definition in the sound with a good clock I don't like the sound SACD at all. It's simply is no music to my ears.
I just acquired a Philips DSR2000 digital satellite receiver. With my DAC [AD8561-CS8412-TDA1543 (NON-OS) I get MUSIC ]
You understand I am extremely puzzled by this result!:confused: :confused: :confused:
It's even MPEG !!:bawling: :bawling: :bawling:

Hi Elso,

I agree that the beautifully built Sony SCD-1 does NOT produce music. Its treble is unacceptably harsh and rude. Its midrange sounds somewhat dull and technical...

Until now, I thought that the problems with treble were inevitable with the DSD-coding system. I thought that only the lower frequencies were the strenght of DSD...

As I have discovered now, the market situation is dictating sound-quality of SACD.
The marketing people are dissatisfied with SACD sales. They want to earn back all their investments in SACD. The only way they can force mass-market acceptance, is to dump prices.
That is where we are now. Cheap SACD players use DSD-PCM conversions and low quality analogue outputs. These players then sound worse than a well implemented cd-player.

The Sony SACD players of the first generation were built by engineers only. I guess noone ever listened to the musical side.

The modified SACD 1000 is the first player that showed me how good DSD in SACD can sound. Its bass and midrange is spectacularly good. In the upper midrange you can indeed hear a quality of room-reverberance unheard in PCM! The instrument is surrounded by air (not fake air) and stays absolutely stable in the 3D positioning (no need for multichannel!).

The treble is nearly as good as with 24bit 96 kHz (in which I record myself), and much, much better than with all the other SACD players that I have heard.

Regards,

Lucas.

PS. What do you think of the van Medevoort raise on the output caps from 100 mFd up to 1440 mFd? Do you see disadvantages?
 
hatasa said:
:eek: What happen there????

:wave2: Becoz my artcle not writing in Chinese... It is with the Cantonese.... Funny!!! Hahahahahahahahahahahaa!!!!!

:cannotbe: Ok! I will have the new review on my own 963SA this week. This time also will perpare the english version at the hidden link for yours! Becoz many of you sent me the e-mail for the request on it....

:wave2: Please give me few days there.... I've just got the toothache and need to meet the dentist tomorrow! :bawling:

Anyway, thanks for the attention on my review!! :clown:


Hi Hatasa,

You did a wonderful job on the 963!
I look very much forward to you translated posting.
Keep us informed here...

I hope your dentist will help you soon!

Regards,

Lucas
 
Re: Re: Philips DVD963SA

Lucas_G said:


Hi Elso,
PS. What do you think of the van Medevoort raise on the output caps from 100 mFd up to 1440 mFd? Do you see disadvantages?

Hi Lucas,
I don't have any idea why van Medevoort is using such high capacitances as output caps. I have no schematic of the 963SA. Maybe the output impedance is a bit on the high side?
I am a little disturbed learning that the discontinued SACD1000 is much better than the 963SA........
:bawling:
 
Re: Re: Re: Philips DVD963SA

Elso Kwak said:

I am a little disturbed learning that the discontinued SACD1000 is much better than the 963SA........
:bawling:

Hi Elso,

Yes, I am too!
Sometimes you see them second-hand for about 500 Euro...

A disadvantage of the SACD 1000 is that it does not offer any PCM digital out (only Mpeg and Dolby digital out).

It also lacks progressive scan (not important to me, since I don't have a beamer...)

And of course it does not offer upsampling... (Why bother?)

I think however that if we provide the 963 with a better PSU and output section, it will sound just as good as the SACD 1000...

Regards,

Lucas.

P.S. I just found your Kwak-Clock mentioned. Could you please send me the details? My email is: l.guitink@zonnet.nl
 
Hi Lucas_G,

replacing the power cap 100uf/400v in the smps with a BG VK 150uf/350v does make a noticable improvement, at least in my case.

Also, most of the E Caps are bypass with a smd ceramic, including the opamp caps. If you already have film caps for bypass, you might want to remove those smd caps used for the opamps. I believe ceramic are not very good for analog. I would also replace 4 opamp E caps 470uf/10v (I believe its 2442~2445) with some better caps. Anything above 8V will be fine. Since space are tight, for the same value, I only find the Rubycon ZL or ZA to fit.
Also, definitely replace the digital decoupling caps, next to the dac and the upsampling chip, with BG-NX or OsCon. This will make the upsampling sound much more listenable. I recall reading some post that film caps are not really great for digital.

Certainly replace the cap next to the 4 regulators labelled something like 'LD33' (3.3v), LF50 (5V), LF80 (8V), and MC??80 (-8V). I think I am missing a regulator somewhere.
Can anybody recommend better regulator in the Dpak package?


I found that in my power supply, there is a large diode label "SB360" next to the larger transformer that get very very hot, does anybody have this in there 963?
 
Hi all.....

Today just met the Dentist and he took me out 2 there!! :bawling:

Sorry that I dont want to show up the price here.... Becoz I dont want that hard-sell my work! I will list it at the article also.... Can tell that it wasnt cheap!

3 master clock with 1ppm TCXO and tubefilter.....

I must be notice yours when the article ready! Thanks again for yours attention. Thanks again! :)
 
sony scd-1 no good ?

Elso Kwak was not satisfied with the Allen Wright modified SCD-1.

I have a SCD-555 (european version) done bij the Dutch agent of Allen Wright. He told me that the modified 555 sounded better than the modified SCD-1 (more musical were his words). I am interested in opinions about this subject.

I am very pleased with the result on the 555, but I admit that Ihave no real reference. It is certainly sounding better that a regular old Philips CD 650 in which I implanted a Kwak Clock. That is meant as a compliment for mr.Wright because I am still satisfied with the upgraded 650, but in a A/B the 555 wins clearly.

I wonder if the SCD-1 produces something that is no music what we have to invest te get something that comes near to it. Is it really that bad ?

regards,

Jaap:confused:
 
Hmmm... I thought this thread is about tweaking the 963sa.

Anyways, I found that part of the 5V use by the dac is pre-regulated by the 8v regulator use by the opamp. This doesn't sound too good! The digital ground and analog ground is connected on the board, not in the power supply.

And here's a pic of "Somebody" stack TWO :bigeyes: AD1955 on top of existing AD1955, so 3 dac in total, and only one cap is changed in the digital section. The cap used is Cerafine.

The upsampler AD1895 is also changed to AD1896. Again no change in decoupling cap.
The original caps are Rubycon YXF and a few ZL.
 

Attachments

  • 963sa_dacmod.jpg
    963sa_dacmod.jpg
    25.8 KB · Views: 3,841
Spec

hatasa said:
Hi all.....
3 master clock with 1ppm TCXO and tubefilter.....

I must be notice yours when the article ready! Thanks again for yours attention. Thanks again! :)

Hi Hatasa,
I presume the 1ppm is not a jitter spec but a frequency deviation...
Why do you feel this is of such importance? Are you able to hear a 1ppm frequency deviation????
:confused:
 
Re: sony scd-1 no good ?

Jaap said:
Elso Kwak was not satisfied with the Allen Wright modified SCD-1.

I have a SCD-555 (european version) done bij the Dutch agent of Allen Wright. He told me that the modified 555 sounded better than the modified SCD-1 (more musical were his words). I am interested in opinions about this subject.

I am very pleased with the result on the 555, but I admit that Ihave no real reference. It is certainly sounding better that a regular old Philips CD 650 in which I implanted a Kwak Clock. That is meant as a compliment for mr.Wright because I am still satisfied with the upgraded 650, but in a A/B the 555 wins clearly.

I wonder if the SCD-1 produces something that is no music what we have to invest te get something that comes near to it. Is it really that bad ?

regards,

Jaap:confused:
Hi Jaap, We only compared/exchanged clocks in the SCD-1 thus far. Later this week I will install the complete Allen Wright mod.

If we ask nicely the moderator can make a new thread : tweaking the Sony SCD-1.
;)
 
Re: sony scd-1 no good ?

Jaap said:

I have a SCD-555 (european version) done bij the Dutch agent of Allen Wright. He told me that the modified 555 sounded better than the modified SCD-1 (more musical were his words). I am interested in opinions about this subject.

I wonder if the SCD-1 produces something that is no music what we have to invest te get something that comes near to it. Is it really that bad ?

Jaap:confused:

Hi Jaap,

I have no idea why the SCD-1 doesn't sound so nice.
It is built like a Leopard Tank, but it was Sony's first SACD design. Maybe the DSD decoding chip belongs to an earlier generation?

If you would like to go deep into it, it is indeed better to start a seperate thread...

Regards,

Lucas.
 
ChuckT said:

Anyways, I found that part of the 5V use by the dac is pre-regulated by the 8v regulator use by the opamp. This doesn't sound too good! The digital ground and analog ground is connected on the board, not in the power supply.

And here's a pic of "Somebody" stack TWO :bigeyes: AD1955 on top of existing AD1955, so 3 dac in total, and only one cap is changed in the digital section. The cap used is Cerafine.

The upsampler AD1895 is also changed to AD1896. Again no change in decoupling cap.
The original caps are Rubycon YXF and a few ZL.

Hi Chuck,

How could one easily change the 5V supply of the DAC?
Is the grounding a real issue?

What is the benefit of two other AD 1955 on top of the existing one? How does this translate in sound?
I saw that in the SACD 1000 Philips also uses Elna Cerafine caps. How good are these? Are they good for both analogue and digital?

Is the AD 1896 better than the AD 1895?
Is the Rubycon considered to be of good quality?

Many questions....
I hope that also forum-members who do not own a DVD 963 SA will share their knowledge of parts-quality with us...

Regards,

Lucas.