diyAudio (
-   Digital Source (
-   -   Oppo`s BDP105 - discussions, upgrading, mods... (

Coris 31st October 2012 04:52 PM

Oppo`s BDP105 - discussions, upgrading, mods...
A new Oppo player is on the way to marked.
This thread is meant to facilitate discussions and ideas changing about BDP105 player, about eventual upgrades, and modifications.

Oppo has come out with only few pictures of the boards of BDP105.
A preliminary examination of the inside pictures it shows that Oppo used a high performing chip op amp for headphone stage. This chip can run up to 2-3W, so is not any problem to drive the most of the headphones. The lowest driving impedance of this op amp is 8 ohm. One can see the placement of this stage on the analogue board here by. It looks to me that the headphone op amp is DC coupled to the output (through resistors)... I`m almost sure now that all the other outputs of the BDP105 are AC coupled. It seems that is planted a large capacitor on the signal output path on each channel! This surprising me very much, and I think is very unfortunate...

It seems to me that the DAC chip itself is used in a different way than in the earlier model (95). A short clarification about the ESS9018 DAC chip: this chip have 8 independent DAC channels inside. Those channels can be connected (hardware) together (2, 4, 6, 8) When all the 8 channels are connected in parallel (for only one stereo channel), one have the max accuracy of the resulting signal. 8 DACs working together and output the same signal (one may use two ESS9018 to have a LR stereo in such case). In stereo configuration of one DAC, are used 4 + 4 channels connected together for max accuracy. Multichannel configuration use all the 8 channels independently.
In the BDP 95 model, the stereo output were obtained from 4 + 4 channels (4 channels connected together). This assured max accuracy for the stereo signal in BDP95 model. In the upcoming BDP105 it seems that Oppo used a configuration for ESS9018 which assure 2+2 channels for balanced/unbalanced output, and 2+2 channels for headphone output. I`m not very sure about the allocation is strictly in this way, but is very sure to me that the Sabre chip is hardware divided between RCA/XLR output and headphone output.

Personally I`m quite critic to this design. I should want that one could have the possibility to use 4+4 channels of the chip for both (switching) normal out and headphone out. This way could assure the max accuracy in both cases. When one use the headphone is no any reason to have half part of the DAC chip hardware dedicated to another output which is not in use, and opposite. This judgement is made as a consequence of what is to be seen in the high resolution inside picture of 105 model. Oppo are welcome to correct me if I have seen wrong in the picture, or to precise more on this aspect...

It seems that this player model have a better distribution of the analogue power supplies and regulators on the boards. Having two dedicated outputs from the toroid transformer it make the PSUs more efficient and it assure less heat dissipation inside the enclosure. This design it were already a must for the 95 model, and it looks like it is implemented now in 105 model.

qusp 31st October 2012 05:28 PM

I guess you are talking about tpa6120A? the datasheet is a lie, their reference schematic calls for 10-100ohms output damping resistors in series with the output, not in the feedback loop, i'd like to see the frequency response with 8 ohms headphones....

Coris 31st October 2012 06:14 PM

Yes, is about TPA6120A2. At last one may take the data sheet as reference... What is there is quite impressive... Personally I did not analysed in details this chip or measured, but I have listened though it. And it was a very good sound out of it...
Actually one may do the same in this case, but for now (without the player on place) are only observations based on pictures/documentations, and maybe speculations...
If you have better experience with this chip, then are welcome your knowledge.
We have to note that this player is not an very high end product... But an player good enough, using quite high end components for a quite reasonable price...As I see in the pictures, the design faults are not eliminated much by Oppo even we have a new and "designed from the ground up..." player. But I can not understand why they chosen AC coupling on all output of the player...

qusp 31st October 2012 06:33 PM

its not a bad chip as long as you dont follow the application note, or use it with low impedance headphones if you do. but if they have followed the datasheet spec, it is not suitable for 8ohms, not even close. you have to enclose the output resistance in the feedback loop to make it suitable, but they never do, seems designers these days just cut and paste without thinking.

you cannot just remove the series resistor, its a current feedback amp and cable capacitance really ***** with it. the datasheet calls for minimum 10ohms and up to 100ohms, thats completely out of step with modern headphones, most headphones you will find these days that are that low impedance are multidriver in ears, with high output impedance the frequency response and particularly bass response is all over the map because the impedance usually dips quite a bit lower due to the crossover.

its a well known issue for a couple of years and I cant believe yet another manufacturer has just blindly made the app note circuit and then quoted the manufacturer spec.

haha output caps you say? the output resistor is probably even bigger then, sounds like a lazy approach all round. surely with their staff they could have worked out some way to switch the outputs for headphone or line out

Coris 31st October 2012 07:48 PM

At this moment one do not know details about how Oppo has chosen to use this headphone chip. If is an 10 ohm or 100 on that output. I hope they used an 10 ohm resistor. Else is quite stupid to use 100... I think is not big problem to lower that resistor to 5 ohm if the headphone cable is not very long...
I have been experimenting with this chip one time, using it as a buffer before an amplifier (with an 300 ohm load resistor). I ware impressed by the result... Afterwords I found better ways to do it...
About this stage of the BDP|105 we will see more later ...

I have to precise that headphone out is not AC coupled. But I see in the picture (I will publish here later on) some large capacitors on the rest of all output channels of the player. It seems to me that those capacitors are on the signal path in an AC coupling design... What could be their reason(s) to do this is a very big question...

qusp 31st October 2012 08:54 PM

even with 5ohms with an 8ohm headphone its... its old school and covering their *** in case someone uses it with an unusually capacitive headphone cable or driver. you cant lower it really, not without a scope. with a crossover in the signal path you have more capacitance to worry about than the cable and it doesnt just effect multidrivers, with 10ohms it will effect single dynamic driver headphones up to 60-70ohms.

for higher impedance headphones (over 100ohms) its quite a good chip, reasonably powerful and clean sounding, but with low impedance and the app note design it has bass rolloff and a hump higher up. if you include the resistor in the feedback network you dont have this problem, but any time I see it by a non diyer and even some diyers and headphone lovers, theyve just copypasted the app note.

I havent looked at any pics so cant comment on the caps you speak of

Coris 1st November 2012 06:59 PM

3 Attachment(s)
Some (detail) pictures of the board(s) of BDP105 (published by Oppo)

Coris 1st November 2012 07:07 PM

I think I have now an explanation about why Oppo has chosen the AC coupling on all the outputs of the 105 model (if this coupling will be confirmed at last). They could not control in production process the 0v DC offset at the output of the final op amps. An 0v DC offset is very important to assure the high performance of the channels amplifiers (in an DC coupling). An high DC offset before the output capacitors, do not necessary mean that the final amplifier stages work at high parameters... and the useful audio signal which goes through the AC coupling is at very high quality one...
But everything was fine for the 95 model, where all the channels were DC coupled, and the DC offset were in acceptable low limits. The AC coupling choice for the 105 model as a solution against a high variation of the output offset, or the difficulty to control this parameter in production process, it may be a consequence of choosing another type of op amps in 105 model, than it were used for BDP95, lower quality of the passive components, or the opamps in 105 are of a lower quality/parameters ones, and of course cheaper... Both those alternatives are not something positive for Oppo and for the new/upgraded player model...
To control by design and then in production the offset of output channels for an DC coupling, it may not be a difficult task for some professional designers. The datasheet of the ESS9018 itself recommend some solutions to well control the output opamps offset. At Oppo has chosen an old fashion coupling, which is definitely a low quality one, for outputting the audio signal, it seems to me that they were forced to do so, by a bad (again) electronic design in this analogue area, or because of economic reasons (a lower quality components from an accepted provider). The economic reasons mean in this case that they want to invest less and sell the final product for more... If this hypothesis it will be true, then it may not "sound" quite well for Oppo`s reputation...

Back to the DAC channels allocation problem, and the downgrade for the 105 model (comparing to the 95 one) in stereo stage.
Oppo could proceed to a very real and effective upgrade if they should use an double DAC mono configuration (one ESS9018 for each stereo/headphone channel stage). This mean that all the 8 DAC channels will work for one of the stereo channels. In this way ESS9018 deliver the max of 133/135 dB dynamic. Such design would be very possible on the same board, and sure enough, with the same software (or very small changes). I can not see that such configuration should have a very important impact on the final price of the player. It may be in to the same 200$ over the 1000$ price of the former model... The ESS9018 chip it cost 60$ when one buy one or few chips. If a big producer as Oppo buy thousands of such DAC chip for a big production, then the price it may decrease very much. I can not see that using one DAC chip more in this new player could be a economical problem and big impact for the final price of it...

Those problems which come out in the case of the 105 model, and another ones which were for the 95 model too, an stated upgrade which it looks in real more like an downgrade (for the analogue stereo stage) in the new model, it make me a quite low impression about the electronic (analogue) design department of Oppo...

David Gatti 1st November 2012 11:06 PM

Thanks for your detailed insights Coris.

So in terms of outright audio performance would you say that the BDP-95 is probably superior to the 105?

Which model would you find more amenable to modding/correcting of the design issues?

Coris 2nd November 2012 07:26 AM

Here is an answer from Oppo about the DAC allocation problem in BDP105 as it is published in post 1977 in AVSForum thread (Official OPPO BDP-103/BDP-105 Anticipation Thread):

"I spoke with our engineers, and this is how the DAC works for the stereo output / headphone output:

2 channels allocated to the headphone section
2 channels allocated to XLR outputs
2 channels allocated to RCA outputs
2 channels to ground"

My only comment to the last line of this answer is: It could be anything more stupid then this kind of design...

All times are GMT. The time now is 05:35 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