Oppo BDT-101CI - discussions, improvements, mods

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Member
Joined 2009
Paid Member
Well, a new (also quite old right now) device from Oppo.
I got at least one such player in my hands (and on my bench)... Quite much enough to be improved on it, as I can see so far...

This player is more meant as a digital transport. It have a (customised) modular concept.
It is equipped with the same excellent driver transport, it looks enough simple (quite cheap, I can say, and it is indeed cheaper than the rest of the Oppo devices).
The same power approach, using a SMPS, even thought the power needs of this device are quite low (overall under 2A). This SMPS concept it have to be for sure modified/improved...
The main board is quite simple. It is powered at 12v, and then regulated accordingly to lower voltages needed by the processor, USB ports, interfaces, and so on. The processor it is clocked by a resonator of 27Mhz (placed on the back side of the board). I already see another improvement here: an quality one oscillator. I will use for sure my own approach: a battery powered clock. I think it will work just fine a 54Mhz SAW oscillator, divided by 2 to provide a better quality clock for the main processor...

When about the audio board, this model is based on a AK4490 DAC. This device is very much integrated, it output fully differential directly, no I/V converter involved, and it sounds well enough (so as it is implemented by Oppo). BTW, the DAC chip AK4490 is priced at Digikey for 5$ (one piece)...
The sound board/DAC is powered by two differential raw rails (+/-15v from SMPS), which are further linear regulated (LM317/337) to 12v. Also a 5v regulator it is used for the analogue stage of the DAC chip. This 5v is down regulated from the raw +15v rail, while the analogue 3,3v for the DAC chip is regulated down from the +12v line. The 3,3v for DAC digital stage it is regulated from the digital 5v rail, created on main board, and also used for USB ports. The linear regulators on this board are switched on power up, by some FET transistors, while the main +/- 15v rails are always on.
The DAC chip is AC coupled to the final buffer, which is further AC coupled to the outputs. I`m not very impressed by this chosen approach...
The final buffer it include a double opamp chip (N5532), which is not a very special quality device... Common, standard, low cost chip...

The DAC it looks like it works in a synchronised modus, and it is clocked by the main digital audio signals (PCM/DSD).
At least not so many clocks/oscillators in this device, than the one for the main processor. And this one is a resonator actually...

Examining closer the audio board, I was intrigued by the last components in this chain. After so many AC coupling stages used into the signal path, I could see more (see picture here), a cap to the ground right on the RCA output, as a component market TSS also tied to the ground. This component it looks like a big ceramic cap, and I can not figure out what its role it may be. I seems like a protection device for some static protection on audio outputs... Well, I do not know what this it may be, but I really do not like it in the signal path at the very end of the chain... Also the MUTE mechanism is based on transistors switching. I would prefer more the relais... Maybe another improvement idea in this respect...

So, we shall only see forward to the improvement steps to come. First and sure big improvement: the power supply (replacing the original SMPS, with my LPM).
 

Attachments

  • AnalogueOUT.jpg
    AnalogueOUT.jpg
    789.1 KB · Views: 257
Last edited:
Member
Joined 2009
Paid Member
Main Board mods...

I think to make a correction to one of my previous posts. This Oppo model it have more clocks inside, than only one...
There is the HDMI chip which it have its own (27Mhz) resonator, and it internally generate its clock signal, as the main processor in Oppo`s design. Why they used separate 27Mhz clocks for these two main chips, is of course not known (except them self).
It seems that these resonators are so cheap as the designers can use a lot of... Every resonator for every single chip which it may need to be clocked...

Personally, I appreciate that the quality parameters of the device it gain more if these two chips, which hopefully it work on the same clock frequency, it may have one master oscillator to run it both with the same syncro clock signal.
As you see here, I used on this mod an Abracon 27Mhz oscillator, powered by battery, and distributed the same clock signal to these chips. Another option is to use my custom made SAW 54 Mhz oscillator, and then dividing the frequency, to get 27Mhz. This option is already in production...
The improvement of the main board is not yet finalised, so please ignore some remaining improvisations.
The next step is to install my new version LPM into this device...

P.S. I finally decided to remove the original mute circuits on audio board. The improvement of the audio signal quality it was immediate and obvious. Well, there are some pops on start up, on outputs. So, I designed a simple mute circuit, using relays, which it will be added accordingly to this board... Transistors control in a mute circuit is definitely not something good at all, for the outputted signal... I got fully confirmed this, once more, in this 101 player.
 

Attachments

  • MainBoard Mod.jpg
    MainBoard Mod.jpg
    598.6 KB · Views: 244
Last edited:
Its bizarre they would use 2 27Mhz clocks.

One would be for the CPU, but that has HDMI integrated correct?.


There is a secondary HDMI board that can be removed. Are you sure the 2nd 27 clock is not feeding this external board? But I would presume the clock for it would be on that external pcb, not the main board.
 
Member
Joined 2009
Paid Member
Its bizarre they would use 2 27Mhz clocks.

One would be for the CPU, but that has HDMI integrated correct?.


There is a secondary HDMI board that can be removed. Are you sure the 2nd 27 clock is not feeding this external board? But I would presume the clock for it would be on that external pcb, not the main board.

Yes, it is bizarre indeed...
The main processor it have integrated a HDMI processor too. However, Oppo still use the same approach with two HDMI channels (as these are also available). In this 101 model, there is a HDMI chip (Silicone Image xxxx), which it process on its own the video signal, and it send it to the standard HDMI output. The another HDMI channel, integrated into Mediatek ARM, it goes to another chip (processor), then it goes to the HDMI module (optional). For this second HDMI channel, it may be used the clock provided by the main processor (based on its 27Mhz resonator), or, indeed, it may be another clock on that module too... I did not investigated yet, closer this aspect....
The Silicone Image chip it have its own similar resonator, 27Mhz. Even the same marking, as for the Mediatek one... Mediatek resonator is on the down side of the PCB.
However, is better to unify the clock for these two chips...
 

Attachments

  • HDMIchip clock.jpg
    HDMIchip clock.jpg
    744 KB · Views: 188
Member
Joined 2009
Paid Member
There is the same approach in another Oppo models too. These payers it has always two HDMI channels available. One channel it is driven by a dedicated processor/enhancer chip (which in 105D is a Darbee processor), another channel is available from the main processor itself. This last one it may be not so much processed, I suppose...
I assume in 101 it is also a HDMI processor integrated into the Mediatek chip (see Mediatek infos for this particular chip).
You can see the four differential traces for HDMI signal going out of the main processor, going into another chip beside, and then going to the optional module connector. The second channel (standard HDMI output of the device) is driven by the Silicone Image processor.
However, I appreciate much more easy to implement a main/master clock for these two chips, as it have the same frequency and similar voltage levels. For us it is fortunate as they did it so... It looks like these two 27Mhz clocks are the only ones in this model (except the oscillator/resonator for USB chip/stage). The DAC is used in a synchronous mode, and so the clock signal is coming from main processor, included into the particular audio digital signals.
As you can see in my modded device (pictured here), I used only one 27Mhz oscillator (my clock board) to drive both chips, and it works just fine. Another option it may be the use of a 54Mhz oscillator divided by 2, as master clock for this player. I already designed and I have in production this divider, which it will be a optional module for my clock board.
Well, I haven`t in this device the HDMI optional module, to appreciate better this part of the case too...
In my opinion, there is much better to have the HDMI processor clocked by 27Mhz and not by a 20Mhz as it was in previous Oppo models... The only design problem I can see so far here is the use of two resonators instead of only one 27Mhz oscillator to clock both chips from the same signal.
 
Last edited:
Member
Joined 2009
Paid Member
Looking once again, and closer at the picture I already posted here, I think I may review my above assumption concerning the HDMI processing in 101 model.
As I can see, it seems that the four HDMI differential lines going out of the main processor (marked on the picture here), are getting into the P13HDX chip. The denomination of this component it suggest to me a kind of transmitter, driver or distributor for HDMI signals...From this chip, the HDMI lines are send further out in two channels. One HDMI channel it goes to the Silicon Image chip, and another channel it goes to the optional HDMI module connector. It looks like this P13HDX it distribute further the HDMI signals, processed into the Mediatek ARM (integrated HDMI co processor, or decoder). If it may be some more processing involved in this distribution, we may not know so far...

I found at least a picture for the optional second HDMI output module for 101. There is indeed a second Silicone Image chip on this module, with of course its own 27Mhz clock resonator (same marking on that as on the one for the main processor).

So, only one main oscillator 27Mhz it can be used to clock the whole device (if a such oscillator it may be able to drive three clock receivers chips...). I think it may works fine also.
 

Attachments

  • HDMI-101.jpg
    HDMI-101.jpg
    679.6 KB · Views: 101
  • IMG_9549.jpg
    IMG_9549.jpg
    213.2 KB · Views: 84
Last edited:
Member
Joined 2009
Paid Member
One thought more about the clocks approach in 101...
I suppose the designers chosen to use a resonator for each of these processors internal clock system, because it is more easy, simple and cheap so, than implementing a master clock distribution circuit, with quite long clock (27Mhz) signal traces on the large enough main board. Such clock distribution circuit, it could be more complicated, when about to clock so a remote (optional) module, as this second HDMI output it is. In this case, The clock signal it should goes through a cable, through connectors, it had to be particular shielded, and so on... Much ore complicated, and more expensive... So, it was chosen each (internal generated) clock for each processor chip, even there is about the same frequency...
I used coaxes for my mod as unified clock system, but such approach it not fit for a production line (mass production)... However the main processor and the HDMI chip om main board are quite near to each other, so these two chips it could be clocked by the same signal... Anyway...
 
Last edited:
Member
Joined 2009
Paid Member
Here is my new LPM - 101 version, with added power module for audio board, and a improved device with installed this LPM installed, as some other mods in place (but not finished yet...)
 

Attachments

  • LPM-101.jpg
    LPM-101.jpg
    705 KB · Views: 137
  • LPM-101b.jpg
    LPM-101b.jpg
    607.6 KB · Views: 136
  • Oppo101 modded.jpg
    Oppo101 modded.jpg
    627 KB · Views: 127
Member
Joined 2009
Paid Member
Here is the finalised, fully improved BDT-101...
Also are some plots, showing the LPM`s rails, measured while device is functioning. Please compare it with the previous published here measurements on original device, powered by its SMPS.
 

Attachments

  • Fully modded BDT-101.jpg
    Fully modded BDT-101.jpg
    921.2 KB · Views: 112
  • LPM 12v rail - BDT 101.jpg
    LPM 12v rail - BDT 101.jpg
    420.5 KB · Views: 109
  • LPM analogue power rails - audio board.jpg
    LPM analogue power rails - audio board.jpg
    614.8 KB · Views: 68
Hi Coris

Somebody just sent me a BDT-101 to take a look and I am just right now looking at the mainboard and can see the xtal under/near the MT8563 main-chip. It is marked 0153T and E5614. But is clearly 27MHz since the Sil 9612 definitely uses 27MHz as per PDF datasheet and has the same numbers on it - strange there would be no "27" there. But the 9612 chips IS 27MHz. So that clears that up.
 
Last edited:
Hi Coris

Maybe you can save me a little time, if you have injected 27MHz clock signal into the MT chip, did you inject on pin 3 of the pad where the Xtal was, or pin 1 or 3? Would think pin 3?

Pins 2 & 4 are grounds, as per usual.

Also, did you remove any small SMD ceramic pF caps - like there is usually two, like on the Oppo 105 etc. I can see the usual two caps in close proximity to the Xtal. Hope you have a quick answer.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.