Oppo's BDP105 - discussions, upgrading, mods...

The new Crystek oscillators are the new industry standard low jitter clocks....many, many companies are using these on their digital input boards. I use them with great results. I have not played with any see SAWs. Of course, I want the SAWs to win....they are much cheaper!

The measured performance of Crystek XO's is a lot better than these SAW's,
especially at low frequencies.

If the SAW oscillator actually does sounds better, all else being equal, then
a few theories need to be re evaluated. Or maybe better = 'nicer' but
not necessarily more accurate.
 
If the SAW oscillator actually does sounds better, all else being equal, then
a few theories need to be re evaluated. Or maybe better = 'nicer' but
not necessarily more accurate.

Absolutely agree on the first point, but I will say categorically that the SAW oscillator we have heard, pretty sure this includes what Coris is hearing, is actually less euphonic and I wouldn't describe it as nicer, but more real and honest. Maybe ruthless at times as well - and hence fascinating too. I can hear certain mixing effects in the music exposed quite dramatically, things like tape noise on classic rexorcings are oh so more analog and realistic sounding. That word "realism' continuously comes to mind. Also, euphoric comes with a level of bloat at cert kind of frequencies, lower, midrange and mid-bass - yet I hear significantly tightened up. I also hear bass textures much more clearly. Deep bass is very authoritative but no exaggeration. But air around instruments, the separation of that air against the actual size of the instruments or voice, where they merged into one, they now can easily separated by the ear. And PRAT, oh boy.

I don't anybody who tries them will call the euphonic. Pretty confident about that.

But as more will try them, the more feedback we get, thisn I believe will al be settled by consensus.

Cheers, Joe R.
 
Member
Joined 2009
Paid Member
I really do not know why people are sceptic to try something else. In this field of DIY one may not be at all conservator... or hanging on an "good old principle". Is the base of experimenting to try new things.
Maybe one or another are or may be euphoric about self experiments and its results. That because is a very big point that another ones (many) try the same and come out with them observations. This only push things further.
In this field of oscillators there is quite easy to replace an good old Crystek with an SAW one. SAW is cheap and quite easy to be found (in quite limited ranges of frequencies, but anyway...).
Replace the oscillator in your DAC/Buffalo and hear the results. Satisfied or not, share your observations, and in the last will be a conclusion. A negative conclusion about SAW, then we will go back to Crystek or try another things... It may not be easier than this...
Everybody will have something to gain out of such "procedure"...
 
I really do not know why people are sceptic to try something else. In this field of DIY one may not be at all conservator... or hanging on an "good old principle". Is the base of experimenting to try new things.
Maybe one or another are or may be euphoric about self experiments and its results. That because is a very big point that another ones (many) try the same and come out with them observations. This only push things further.
In this field of oscillators there is quite easy to replace an good old Crystek with an SAW one. SAW is cheap and quite easy to be found (in quite limited ranges of frequencies, but anyway...).
Replace the oscillator in your DAC/Buffalo and hear the results. Satisfied or not, share your observations, and in the last will be a conclusion. A negative conclusion about SAW, then we will go back to Crystek or try another things... It may not be easier than this...
Everybody will have something to gain out of such "procedure"...

Coris,

Don't worry I'll be trying them at some stage, I have a Sabre running on
100MHz kicking around here.

Generally from a diy'ers perspective it's all good and a free lunch :) You get
to try new parts and get great results. That's what is good about this place.

From technical perspective it's also good because we have something new
to investigate. :)

So these SAW oscillators apparently sound better than a conventional
oscillator but as far as I can see measure worse.
Have we missed something? I'm interested to find out.

Joe, maybe you could get a phase noise plot done from 1Hz -> 10kHz
on one of your SAW / Terrafirma oscillators?


cheers

T
 
Member
Joined 2009
Paid Member
I succeed at least to power up my new BDP105...
Well very few and quick personal impressions about this new Oppo player (as it is out of the box):
I have to say/recognize that this machine is an step forward, comparing to the former model (unmodified) 95. The picture is just exceptional. I used for test only static HD pictures. I had no time yet to put on it an video/movie. The conversion 2D - 3D is quite poor. If the picture has an whatsoever background, then the 3D effect is a little bit there... But this is not so important anyway.
How it sounds (in my opinion)? Well it sounds good. Much better than (untouched) 95 model. I could hear many details, a quite deep and well defined bass, and for sure more fidelity than from 95. But is something wrong with the sound stage of this new player... No any sound stage, I mean...
I have an record on SACD which can give me a quite wide sound stage (on stereo out) when I hear it on my modified 95. But on 105 the instruments are all together just in between the two speakers (in the middle of the stage) on this my DSD recording. I played Erick Clapton unplugged (CD). This (old) recording is very good, the sound stage is wide and very well defined. The sound out of my 105 come (on this CD) well enough from the directions of the respective speakers, but nothing in between... The sound stage on my 105 device is wrong and very poor!
I haven`t tested yet how it sounds on USB in. I will do it in the next days...

It could be nice if somebody else (Joe, or somebody else who have had this player already) can come here with some impressions about this new model of Oppo as it is, out of the box (unmod). Is good to have something to compare: before and after modifications...
I can not wait to jump in to this player to fix what is to be fixed out...
 
Member
Joined 2009
Paid Member
I will not say that 105 is a bad player. I said already that it sounds better than 95 (out of the box). In the same time, I have my personal observations (in the earlier) post . And follows here.
I`ve plaid an UltraHD CD (DXD recorded). It sounds very well indead, detailed and the sound stage is there (not very well defined...). I only miss a little bit more dynamic in the sound...
I`ve tested the USB in with the files I`m used to play on my mod 95. I can say that this stage of the player sounds much better than the transport one. It could be a little bit more dynamic too... For me it works much better on KS than ASIO (i use Foobar/win7).
I was not very impressed by the headphone out (Sennheiser IE 80), but it sounds well. I could register a quite big background white noise on this output. Well audible when no audio signal is present. This noise have the same level at increasing volume, and it hears as an chip generated noise or an induced one from somewhere. I did not measured anything. Is just what I can hear out from this player now... I use my headphones most on my Note II, and I can definitely not hear out of my phone that white noise in silent passages, as I can hear it out from 105...
I may say that I will not pretend those my first impressions as a well qualified review of BDP105. Not at all...

Thanks gwmorgan for information. I`m still waiting for own observations from owners of this player, they who have hear it and have or not critics about one or another aspects in audio stage first. I think almost all users may agree that the picture (video stage) of this machine is just impressive...
 
Last edited:
I have not modded the headphone outs yet....but if the feedback cap in the IV stage is a 100pf like used in all the other output stages then there is practically no filtering. Look at the data sheet for the TPA6120A2 headphone amp. They show a DAC using an IV converter and the feedback resistor across the IV converter is 1K and the cap across the resistor is a 2700pf.....not 100 pf. I have not removed the tiny surface mount ceramic to measure it but looks to be the same as the other stages (where it is 100pf). The other stages use a summer with another pole of filtering. There is only one pole of filtering in the headphone output.....and if it is just a 100pf then its way out in the megahertz range.....not going to filter any noise. I will be tweaking the headphone circuit soon and measure that cap and if it is just 100pf I will raise that value to at least 1000pf (Wima polyprop with its outside winding facing out.....for best sound). This may eliminate that noise you are hearing.

So, try removing that cap (one on each phase of each channel....4 in total) measure it and if it just 100pf then try a bigger value.
 
Member
Joined 2009
Paid Member
Thanks Ric for your input. I will of course take a closer look on the headphone stage in the near future. My intention for the moment was to come out with my observations about what I could find out on how it works this player out of its box...

I may say that I never tested with headphones the outputs on my modded 95. I will do it. In the analogue stage after the DAC in my 95, I use few pF caps on that I/V resistor. The same in the feedback of the following opamp stage (final). I succeed to reduce the high frequency noises on outputs in another way... Anyway those noises (Mhz...) which it come out from the DAC are not possible to be heard... Maybe the eventual harmonics in conjunction with the audio signal...
The point is to reduce the level of those high frequencies radiations and prevent to be send it out of the player (in cables).
 
Member
Joined 2009
Paid Member
In my opinion, caps in the signal path have negative impact over the signal quality. I appreciated that caps over 50pf used in circuits after DAC it affect negative the sonic results. The same about the filter used in between I/V and final opamp. So, I eliminated the filtering and I lowered the caps capacity to the lowest possible level. Without those caps it sound not god either... and the high frequency noise is to high... Then I found a way to reduce drastically the HF noise after the final opamp.
I`ve done in this way, and i deal now with an 32mVpp output HF noise for an 28Vpp audio signal out of the final opamp... I think is pretty well like this...
 
Hi Guys

Looking at that 27MHz on the mainboard - you have to remove the entire board to remove the OP8581 Mediatek chip. Remove 7.1 board, the stereo board then remove entire back panel, then finally remove digital mainboard. Only then can the heatsink be removed by squeezing the ends (four plastic 'harpoon' ends) one by one and off comes the heatsink with it sticky (clingy) thermal pad.

I reassembled everything - can simply push back heatsink when finished - and I can operate the player carefully without yet pushing the pins in. Checked running an audio disk for an hour (no heavy video processing, which may not make a difference anyway) and estimate the heat to be 40-45C. So I would not recommend running without heatsinking.

Take a look at photos attached. The caps C19 and C140 I don't know the values - but they will be a few pF. R33 is 10R and R61 is a typical 100K.

To insert a clock signal, the two caps will need removing and also removing 10R may in fact isolate the Xtal so that it need not be physically removed. We shall see.

Cheers, Joe R.
 

Attachments

  • MT-OP8581.jpg
    MT-OP8581.jpg
    511.6 KB · Views: 598
  • 27MHz.jpg
    27MHz.jpg
    603.8 KB · Views: 569
Last edited:
Member
Joined 2009
Paid Member
Nice job Joe! It looks like Oppo have designed specially this way, to make it difficult for they who want to mod their machines...;):D

I see that this oscillator is marked X2. The question is where is X1, and what is used for... The same notation it were used for 95: X2-27Mhz, X1-25Mhz...
It may be X1 the clock for QDEO video processor?
I will take soon a close look on my device.
 
Nice job Joe!

I see that this oscillator is marked X2. The question is where is X1

May in this instance Oppo is following this logic: X1 is 54MHz for Sabre DAC and X2 for the single main system clock. I don't see another X1 anywhere else. There are no clocks under the board - wish I had taken a picture of it, but sure I didn't because there was nothing there. Interesting, I removed three boards and all components were on the upper side.

Oh, on the photos I took while I had the player open, I do see an X3 - nearer where transport ribbon connector is - not sure what the frequency is, but it is not clocking anything in the OP8581, not at that location - near a Weltrend WT61P8. Some kind of display controller chip?

Cheers, Joe R.
 
Last edited:
Thanks, Joe R and Zenelectro and Coris, for your kind words and very helpful pointers. I've been traveling all through the holidays, so apologies for the delayed response.

I'm not equipped in any sense to participate in the exploratory surgery, but I will definitely follow your investigations with interest, and look forward at some point to learning what may be done to get the most out of this player (and inform purchase decisions about future designs).

I will say that my 105's sound quality has evolved pretty dramatically with use. It had about 72 hours of continuous signal (some SACD, some red book CD) before I left for the holidays with the unit on repeat play. When I returned, the repeat function had stopped (I suspect a power blink), so I don't know how long it burned in before stopping (anywhere from another 2-192 hours, but I suspect less than half the higher number). Repeat play works fine now, as does everything else, and the unit has processed another 96 hours or so of signal, for a total of 160+. I'm still hearing (smaller but noticeable) evolution. To minimize adaptation effects in evaluating, I'm doing ABA comparisons with my current player, a BAT VK-D5SE (very fine analogue stage, badly outdated digital section) and noting how the relative sound changes (same balanced interconnects for both, same length, feeding same preamp (BAT VK-52SE), amp (BAT VK150SE monoblocks) and full-range stereo speakers (Von Schweikert VR-4 Mk III).

As often with new components, the out-of-the-box sound was tight and "plastic-y" with poor differentiation between attack, sustain, decay tail (short), and reverberant components (attenuated) of single notes at all freqs. Plucked bass (acoustic or electric) was not boomy but "rubbery" in perceived quality. I'll spare you the blow-by-blow on changes but note that the perceived frequency balance has shifted (and continues to shift) in various ways, the clarity and articulation have developed dramatically, and the soundstage has become more and more distinctive to each recording--an excellent thing. Good recordings are beginning to have remarkable depth (if present in the venue) and specificity of instrumental and section placement. I can hear interference "beats" between two instruments playing the same note or between the reverberant sound of two piano notes with a clarity I've never heard from any digital source. Instrumental timbres are already quite lifelike and with remarkable differentiation. One of my references for this is the landmark David Jones recording of Ivan Moravec's Bösendorfer on his first Chopin disc (Connoisseur Society LP, VAI CD). The subtleties of the piano's voicing through the different registers, attacks, and pedal techniques are very revealing, and the 105 is beginning (in the last dozen hours of signal or so) to have a very natural balance (including the dramatically extended and resonant bass octaves of the Bösendorfer Imperial) as well as a more fine-grained harmonic texture than I've yet heard.

I'm still waiting for the frequency balance of the player to reach its steady state, and I still hear on red book CDs a little glare or emphasis in the soprano and presence registers; that may be inherent to the medium or may resolve, or may reflect the limitations of the stock 105. Articulation of the sense of individual violins or violas in sections during massed passages has developed very well, but its limits test this area. As do the variety of stick strikes on jazz cymbals (wooden tip, nylon tip, wooden shaft, at different distances from the center/edge). Will keep observing.

I've had little chance to compare SACD material to other players, so my impressions there are more subjective, but the soundstage depth (of the Fiedler Symphony Hall Living Stereos) and width (Reiner Orchestra Hall), and placement are pretty impressive, as are the dynamics. One can practically see the mike placements on these recordings. I'll look forward to trying some high-resolution non-DSD files played from memory as well, once I'm set up to do so.

I'll wait for another 100 hours of signal before drawing real conclusions (haven't yet listened to discs recorded in my familiar local venues, for instance) but this is a noteworthy unit even in stock form. Based on listening so far (articulation and low-level detail), the DAC implementation itself must be pretty close to state of the art, and I'm keen to hear what effect (if any) alternative clocks and various analogue stage changes have on its ability to reach its full potential.

Thanks, guys. Will be keen to follow your experiments.

PS: Joe, you're welcome to use anything from my first post that you think stands up to scrutiny. It's courteous of you to ask. Just consider it thought experiments--fair game for scientific challenge and correction.
 
I would like somebody to take a look at this - see below attached Gif files.

The first one shows how it is traced out on the Oppo 105 here.

Note that I am guessing X-IN and XT-OUT.

If that is correct, then I propose removing all four SMD components and insert 100K resistor as shown. That way the Xtal does not have to be physically removed. Personally I find that a good idea - also makes reverting it back a lot easier, if need be.

Other than that, simply remove Xtal and SMD caps.

Note: If anybody knows a manual of a Bluray player that uses MT8580, we can confirm which is X-IN and X-OUT. Based on memory and if the orientation is similar to the past, then my guess is correct. If not, then revert to the other side to inject clock signal. Simple trial and error and 50/50 chance.

Cheers, Joe R.
 

Attachments

  • Old_Xtal.gif
    Old_Xtal.gif
    4.8 KB · Views: 511
  • New_Osc.gif
    New_Osc.gif
    4.4 KB · Views: 493
Last edited:
Member
Joined 2009
Paid Member
About wiring of the oscillator here I will think to an analogy with the 95 model. It may be the similar design, and the chip for sure do not have any changes for the clock pins, comparing with the old model.
So, in 95 model the clock were injected on X1 crystal tab 3 (pin 3). I think the designers used the processor corresponding pin as clock IN and another corresponding pin as clock OUT. This design it may also be similar for the 27 Mhz clock on backside of the motherboard of 95 model. There may not be important reasons that this "logic" and design it will be modified for this new processor model (105).
So, my supposition is that the clock IN pin of the processor it may be connected to the pin 3 of the 27Mhz crystal (105 model).
I fully agree that the most reasonable way is to remove the passive components around this crystal to isolate it, and do not desolder it out from the board.
I will want to identify the trace which goes from the pin 3 of the crystal to the processor, and I will want to try first to connect there the new clock signal (after isolate the crystal).
There are of course 50/50 chances that this logic may be true. So if it does not work this way, then one may connect the clock signal to another processor pin...
In general the design of such clock inputs are made to function well for both way of use: crystals and oscillators. So, may not be catastrophic if one connect the oscillator output on the so called output clock pin of the processor. I will want to use first an 27Mhz oscillator as clock, to see that all may function as it should, and then try something else...
 
Last edited: