diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Digital Source (http://www.diyaudio.com/forums/digital-source/)
-   -   Upgrading & modding new Oppos, BDP-93 & BDP-95 (http://www.diyaudio.com/forums/digital-source/190033-upgrading-modding-new-oppos-bdp-93-bdp-95-a.html)

Joe Rasmussen 1st June 2011 11:32 AM

Upgrading & modding new Oppos, BDP-93 & BDP-95
 
Hi Guys

I mistakenly started this thread on Digital Line forum... oops, now I am in the right place. :)

Subject: Upgrading new Oppo Players, BDP-93 & BDP-95

Here in Australia there has been a bit of drama and delays to getting the new Oppos from the importer due to multi-region hacks causing freezing problems and nobody down here wanted them until fixed.

Many of the commercial "modders" in the US are already doing the players up. Just wondering if anybody on diyaudio has gotten their hands on '93 or '95 and done anything to them.

As far as I can tell it is a 27MHz(?) Master Clock on the main MediaTek chip.

On the stereo analog outputs, transformers with 1:1 ratio can be used on both the CS4382 DAC on the '93 and also on the Sabre DAC on the '95

I believe both these players are going to be SIGNIFICANT for most, if not all, of us !!!

This really excites me!

Players and DACs that can play non-compressed or lossless full resolution files... this is the FUTURE!

They play FLAC files, even Hi-Rez to full MCH, Wave files too. I have a bunch of 24 bit FLAC files, music that I have downloaded legally from B&W (yes, the speaker people), including the latest Peter Gabriel "Scratch My Back" album.

Actually not fussed about the video side, but believe it will be very good as Oppo usually is, but we are of course DIY Audio guys.

Over to you and comments you may have.

Cheers, Joe R.

Joe Rasmussen 16th June 2011 02:34 AM

Somebody is playing games!!!

The commercial modders are sitting on information that needs to get out. Previous Oppo players have always been 27Mhz.

It is taken some detective work to find this out, and you will need to know where to look, and it wouldn't be a bad thing if you knew a little bit of French.

It is NOT 27Mhz, but 25MHz

It seems the commercial modders who have worked this out is trying to hide this for as long as they can. For example the order number from Audiocom is SCLK-2500, yet it is not listed on the website when ordering their Superclock. Some French guy did order it and posted the following on a French tweaker forum:

http://customanalogue.com/diy/Oppo93...ck_Install.jpg

This is the X1 Xtal - 25MHz:

http://customanalogue.com/diy/Oppo9395/Before_25MHz.jpg

Here is a close-up image where I have attached some details:

http://customanalogue.com/diy/Oppo93...ve_Connect.jpg

You can that three components are removed, X1, C216 & R71

The Superclock is powered from here (note only "+" needs to be connected):

http://customanalogue.com/diy/Oppo9395/DC_Power.jpg

Clearly it is better to power from an entirely separate power supply with it's own transformer - this allows it to float and not share the same ground as the player.

Here is the actual connection:

http://customanalogue.com/diy/Oppo9395/Connect.jpg

Finally the plate-heatsink is re-fitted:

http://customanalogue.com/diy/Oppo9395/Top_View.jpg

Please note, the above is not my handiwork. I have not fitted a clock yet. Needless to say, I would fit a Terra Firma Clock.

Final comments: The change from 27MHz Master Clock to 25MHz is something that is not well known. There is another Xtal X2 under the PCB which is more difficult to access. The numbers on it gives no idea as to its frequency, the markings are "0153T" and "E 811" - the 25MHz Xtal X1 on the top of the PCB is much more easily accessed and can be worked on without removing the PCB, taking the usual precautions. The markings on the X1 is "2500P" and "E042A". A frequency meter on a '93 here confirmed that was 25MHz.

For those who can read French, this is the forum:

Les Tweakers BDP93/BDP95 (tout ou presque en page 1) - Page 23 - Les sources

Cheers, Joe R.

digilog5 22nd July 2011 04:52 PM

Oppo 93 and i2s
 
Joe,

Thanks for posting this information. In one of my systems I have an Oppo 93 which I use as transport to a W4S DAC2 for video and music. I've been primarily using the Asynch USB with a Windos7x64 and Jriver for playback but I was also interested in having the Oppo acting as my main music playback and get away from the PC method. So far my impression is that the stock oppo 93, even via coaxial pcm out, does a pretty decent job as a digital player. The jitter reduction of the W4S Sabre32 dac helps alot.

My thinking is to add an LVDS HDMI board to tap into the i2s lines and use this as output to the DAC. This can be done. I am only interested in three line feeds: DATA, bitclock and wordclock. My external w4S dac has its own very low jitter high freq. clock to run as master.

My question is, would a superclock with a clean supply (or any other low jitter clock) which feeds the main Mediatek CPU of the Oppo, have a significant impact in the three outgoing i2s lines after the Mediatek chip? I don't care really about the outgoing MCLK which i will not be using.

My understanding is that this is a complex CPU which is doing many things at the same time (video and audio decoding) and generates the desired PCM i2s going to the DACs. I am not sure exactly how it handles/generates the different sample rates for 44.1 or 48 k (and their derivative freqs) for audio.

Theoretically it is possible that by externally feeding a low jitter master clock to the decoding CPU, the outgoing i2s lines will be cleaned up to some extent. This external clock somehow in synergy with the CPU, must have an impact to the other clocks and data too. At least this is my understanding...can anyone comment on this?

I've read some comments for other chips for example the popular Xilinx FPGAs for audio decoding, that their output is affected by an external master clock.

Thanks!

Coris 11th August 2011 11:08 PM

Oppo BDP95 mods...
 
I just started a modification process on my new BDP95.

First I have to make known some of my observations about what is inside of this newest Oppo machine.
I could be helpfully if somebody else can confirm some of my observations.

I was very surprised to see/feel that the analogue/serial power supply dedicated to the sound processing PCB, gets very hot. There is not only the regulators witch gets hot (this is quite normal) but the filter capacitors get hots too. This is very bad! This ting revel that a serious design issue occur in this case. At Oppo designers could design a such power supply for their hi end product is very surprisingly, and maybe a shame too....
Capacitors gets hot in a power supply only because they have to endure a big ripple current. This is the first thing an electronics designer have to care about in a design process.
More than this, they who designed this serial power supply in BDP95 has chosen to regulate +/-15v from a transformer/bridge with an output of 22v. There are 7v witch produce here only thermal energy. Oppo state that is used an custom made toroid transformer from Rotel. I have to say that this transformer is one with a quite standard output: 18v. This mean that in DC (after bridge/filter capacitors) will have to be 24v. If the transformer is right loaded, it have to maintain this output level. It is not happen here in Oppo`s power supply. The output in this case is +/-22v. witch will be later regulated to +/-15v. For the regulators Oppo used in this power supply, is no need more than 3v over the regulated tension, for that the regulation occur and be effective. This mean that is no need more than 18vDC (after the bridge) to get it a regulated 15v. If one have more than 18v, mean that only thermal energy witch dissipate useless inside the player...
Oppo`s designers chose it to use +15v rail to create/regulate 9v, used on the DACs board. On the board is is also printed 12v on the input pin where one measure 9v from the power supply... The +15v rail is overload in this way and produce ripple current in the filter capacitors. This have an negative impact in the digital conversion process, and in the rest of the electronics.
I`m more disappointed to see that on the audio board it self, Oppo choose it to create the 3,3v/1,8v needed of Ess9018 chips, from +15v rail. This is really bad idea! A huge tension difference (12v) goes in thermal energy in this case too. It is create unnecessary overloads for both components and the power supply.
All these first my observations mean only that many modifications will be necessary to repair the serious design issues in this Oppo product.
But it sounds well at least, once one used the player as it is from the producer. Why? Because the main components Ess9018 are too goods and help to remove many of the design fails. The same is the case of the digital process witch correct many of the imperfections in the Oppo`s design. But anyway...

About the modification presented here (replacing the main clock), I`m quite critical about the final result in audio stage. If one want to get up in quality with this player (audio stage), is most rational to change the clock witch is involved directly in this audio process. This clock is placed (quite bad) on the audio board, and is an 54 Mhz oscillator. Personally I have not yet modified this one, but I have a plan for this as soon as the problem with the power supply of the player will be fix it.
I think that are quite many things witch have to be replaced on the audio/DAC board. I just started with the most of the capacitors on this board, and I have to say that the sound quality already get up remarkable.
I will come back later with necessary details, as I can go forwards with the modifications....

Coris 11th August 2011 11:20 PM

I just removed Oppo`s originally serial power supply, and replaced with two +/-15v and +6v shunt regulators. It is not the final modification this about power supply, but I have to use something better anyway... The result in audio quality come at once!

Coris 12th August 2011 07:38 AM

unfortunatelly there was many grammar faults in my previous text. I`ve corrected some
 
I just started a modification process on my new BDP95.

First I have to make known some observations about what is inside of this newest Oppo machine.
It could be helpfully if somebody else can confirm some of these observations.

I was very surprised to see/feel that the analogue/serial power supply dedicated to the audio stage processing PCB, gets very hot. There is not only the regulators witch gets hot here (this is quite normal), but the filter capacitors get hots too. This is very bad and wrong! This ting revel that a serious design issue occur in this case. At Oppo designers could design a such power supply for their hi end product is very surprisingly, and maybe a shameful too....
Capacitors gets hot in a power supply only because they have to endure a big ripple current. If they are rated to work in quite high temperature environment is something. When they get hot themselves because wrong use, is very different! This is the first thing an electronics designer have to care about in a power supply design process.
More than this, they who designed this serial power supply in BDP95 has chosen to regulate +/-15v from a the transformer/bridge/filter witch output an +/-22v. There are here 7v witch produce only thermal energy. Oppo state that is used an custom made toroid transformer by Rotel. I have to say that this transformer is one with a quite standard output: 2x18v. This mean that in DC (after bridge/filter capacitors) will have to be 25v. If the transformer is right loaded, it have to maintain all the time this output level. It is not happen here in Oppo`s power supply. It looks like this transformer it has not eniough power to sustain the needs of the regulated power supply and the audio board too. The output in this case is measured +/-22v. This will be later regulated to +/-15v. For the regulators Oppo used in this power supply, is no need more than 3v over the need it regulated tension, for that effective regulation occur. This mean that is no need more than 18vDC (after the bridge/filter) to get it a good regulated 15v. If one have more than 18v, that mean only thermal energy witch dissipate useless inside the player, and gets hot the regulators components...
Oppo`s designers chose it also to use +15v rail to create/regulate 9v, used on the DACs board. By the way, on the board is is also printed 12v on that input pin where one can measure 9v from the power supply... The +15v rail is overload in this way and produce ripple current in the filter capacitors. This have an negative impact in the digital conversion process, and in the rest of the electronics.
I`m more disappointed to see that on the audio board it self, Oppo choose it to create the 3,3v/1,8v needed by the Ess9018 chips, from the +15v rail. This is really very bad idea! A huge tension difference (12v) goes in thermal energy and is dissipate by the regulator elements.... It is create unnecessary overloads for both components and the power supply, and all this only shorter the life of these components.
All these first observations mean that many modifications will be necessary to repair the serious design issues in this Oppo product.
But the player it sounds enough well at least, once one used the player as it is from the producer. Why? Because the main components Ess9018 are too good and help a lot to remove many of the design faults. The same is the case of the digital process witch correct many of the imperfections in the Oppo`s design. But anyway...

About the modification presented here (replacing the main clock), I`m quite critical about the final result in audio stage. If one want to get up in quality with this player (audio stage), is most rational to change the clock witch is involved directly in this audio process. This clock is placed (quite bad) on the audio board, and is an 54 Mhz oscillator. Personally I have not yet modified this one, but I have a plan for this as soon as the problem with the power supply of the player will be fix it.
I think that are quite many things witch have to be replaced on the audio/DAC board. I just started with the most of the capacitors on this board, and I have to say that the sound quality already get up remarkable.
I will come back later with necessary details, as I can go forwards with the modifications....[/QUOTE]

johnm 14th August 2011 02:18 PM

I own an Oppo BDP-83 and mine gets pretty hot too. I'd like to use it as a CD transport but the fan tends to come on after approximately 15-20mins and it is audible.

I'm wondering if some passive (extra heatsinking?) cooling on the regs would alleviate the need for the fan?

Interestingly, my stock BDP-83 sounds so-so with CD, and yet sounds fantastic with SACD. Strange as they both use the same output stage...

Whilst using a new clock can improve the sound, I'd guess that building an external power supply (i.e like Naim) would offer greater sound quality gains? The PSU is the foundation of any player of course...

SoNic_real_one 14th August 2011 04:58 PM

Quote:

Originally Posted by Coris (Post 2669172)
I`m more disappointed to see that on the audio board it self, Oppo choose it to create the 3,3v/1,8v needed by the Ess9018 chips, from the +15v rail. This is really very bad idea! A huge tension difference (12v) goes in thermal energy and is dissipate by the regulator elements....

[/QUOTE]

Only from this statement I can tell that you have no clue of what those power supplies are and still you make comments about how bad the design is.
I won't waste my time to point out the other usual urban legends that you pick up from around this forum.
BTW, the 25MHz is NOT the same clock with the 27Mhz, that is not a "conspiration" of the modders. Those are two separate clocks.
And also there are two separate paths for PCM and for DSD (until they hit the DAC's).

Coris 14th August 2011 07:29 PM

Only from this statement I can tell that you have no clue of what those power supplies are and still you make comments about how bad the design is.
I won't waste my time to point out the other usual urban legends that you pick up from around this forum.
BTW, the 25MHz is NOT the same clock with the 27Mhz, that is not a "conspiration" of the modders. Those are two separate clocks.
And also there are two separate paths for PCM and for DSD (until they hit the DAC's).[/QUOTE]

Well SoNic_real_one. Have you a clue about how these power supplies are? Just great! Maybe you can share your clue with us ...
Anyway, a power supply witch run their filter capacitors hot because of an internal hi ripple current, caused by a unprofessional design, is very wrong! And a shame too for a such name as Oppo...
BTW, what I state in my comment is supported by facts. Just measure an see your self on the audio board. I know well what I`m saying/writing!
Else, the statement you refer about the 25/27 Mhz clock is not mine... There is true that on the boards are many clocks. I do not care about these. I care about the clock witch is involved in audio stage. This is I`m talking about.
I`m not a hobby modder who want to mod only for fun. But I want that things (witch I pay quite much money for...) works correctly. I get frustrated when I see that a sold hi end product suffer of primary/elementary design faults. Oppo could run self a research, consult some professionals in design of electronics, before lunch their last player. Is not enough that a player sounds well enough. One have to get the most of that product witch incorporate the best in today components... In my opinion a well designed product, when is using the last/top electronics components available, can sound exceptional. BDP95 can this! Only some details (like for example a well designed serial power supply, and a more accurate power distribution to the boards/components) can bring much more to the same product. And unfortunately is not only the power supply problem in this player.... For example, long traces (50-100 mm) from the main oscillator/clock to the target chip is quite unknown in the electronic design. It is elementary that these traces have to be as short as possible... Well, not for Oppo... The main clock on the audio board (54 Mhz) run his output on very long traces. A well designed antenna for this oscillator! I just had to use a mini coax from the clock to the ESS9018. One can also hear the difference...
For not say more about the heat sink for the main processor... I can see that the designers of the main digital board, had an idea about a right heat sink for the processor. The electronics components were placed right on the board for the use of a such good radiator. But was never installed in that place. It was used a ordinary steel plate with a thermal guider compound... I mounted on that processor one heat sink from an old computer processor. I had so a better picture with exceptional details... Just simple, just immediate results...
But I can understand that the main line was for Oppo to come out with an under 1000$ best as possible player. Well, they quite succeed... Just for 100 bucks more could be the best player ever...

Coris 14th August 2011 08:32 PM

I see now that I`ve wrote it wrong figures about that regulators (3,3v/1,8v). Is of course about 3,3v/1,2v... It was a mistake in my writing, but the rest still be as is state it...


All times are GMT. The time now is 01:50 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2