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

Was the preferred Oppo stock or one of your modded versions?

I thought that was obvious. The DirectStream I had here would trounce the standard Oppo in every way. Then again we are talking 5:1 price ratio or close to it? OTH, the stock Oppo is as good as you are likely to get at that price point. But I have clients who sends their Oppo to me and when I return it they exclaim "it's hard to believe this is the same player I sent you" again and again.

I too read with curiosity the virtues of the DSD dac with its Pikes Peak OS and previous version....

Quite familiar with the DirectStream story and much of the inner details. Some of the things said I agree with and some I don't. But the one I was able to borrow was the latest Pikes version as of now and it is very good indeed and I am not going to criticise it here as I have several friends of mine who are dealers. But let me just say we have an Oppo 105 (JLTi Level 4.0) here that competes, and all the added features it has over what is basically just a USB/Networking DAC only.

Re transformers, in the DirectStream DAC I think it is a neat, even a logical, solution. I use transformers myself often, the work well with "voltage" DACs used in Oppo players and sound really good on the Oppo 93, 104 and 103D.

To me the interesting question is whether a DAC's resolution should be determined by it THD+N measurement? In which case the DirectStream measures not that great, but now better than before. But are we taking for granted that THD+N is a good way of determining "goodness" of a DAC's performance, or is it just another one of those numbers that don't tell the whole story, but are at least part of the story? Here it is the +N part that is contentious as I understand the original measured only 17bit according to John Atkinson. But then the DirectStream DAC is software driven and basically designed by a software engineer, Ted Smith, and he can apply a number of treatments/shaping to the +N aspect and that also affects the THD part.

As Terry Demol (many here know him) told me, it is only a matter of time before all DAC will be software DACs. So the DirectStream DAC is a precursor to that. How long will that take? Maybe 5-10 years?

Cheers, Joe

PS: Ted Smith first tried filtering a DSD signal from a Sony VC24 chip back in 2007. The late Allen Wright did it on a VC24 chip fitted to a Sony XB-940 in 2001 - and this simple idea is behind the DirectStream DAC. But we were doing this years before and indeed Allen extracted almost painfully (by being a pugnacious and insistent Allen) the confirmation from Ed Meitner that this would work. Just for the record. :)
 
Last edited:
Thanks for the response Joe ...IMO its the harmonics number in THD+N that might be one of the indicators to the musicality of a DAC. The noise in +N is random and wideband in nature so is benign if its low enough (less than 100dB?) I believe.

If a DAC is fed a simple signal like a sine wave and it produces lots of harmonics (which are unwanted distortions whether they are related to the fundamental or not), what would happen to the distortion profile if fed complex signals like music? Even more unwanted distortions more complex in nature than before. Hence the need to minimize these distortions as much as possible in a DAC so that the original signal comes through as unmolested by distortion(and noise) as possible.

That's my 2 cents anyways:)...
 
I have never heard of an OTA until today when it was mentioned a few posts ago. I looked at its datasheet(OPA860) and after a while wondered why would one why to use it in a DAC such as an Oppo or a preamp. It seems to be an excellent current amplifier that thrives when driving low impedance loads which wouldn't be the case when driving the high input impedance of a preamp or a power amp. Obviously I'm missing something here...
 
I've come to the view that correlated noise in DACs is a major issue in relation to how they sound. THD not so much.

Tend to agree. And the active phrase here is 'correlated noise'.

Both THD and +N have spectral characteristics that have huge impact on the sound. But THD can actually be quite benign in some cases.

Hunt down the effects that most affect sound quality in digital playback and it is noise related. In real estate the mantra is location, location and location. In digital playback the mantra should be noise, noise and noise.

Cheers, Joe
 
Member
Joined 2009
Paid Member
If my previous showed LPM was more like an prototype, here is the first one as product in its last/final version.
The screenshots are take on this final version LPM, with a non-inductive resistive load, and powering the player. On resistive load the ripple (with harmonics) is no more than 3µV loaded for 5A on 5v rail, and 200µV on 15v rail (1A). Measurements points are on LPM side, before the final filtering module/connector.
The production is quite slowly in the beginning, mainly because the transformers, but I do hope it may be better later on. A first batch of 4-5 items it will be available in the next days...
 

Attachments

  • Last version LPM.jpg
    Last version LPM.jpg
    628.9 KB · Views: 299
  • LPM in action.jpg
    LPM in action.jpg
    441.9 KB · Views: 284
If my previous showed LPM was more like an prototype, here is the first one as product in its last/final version.
The screenshots are take on this final version LPM, with a non-inductive resistive load, and powering the player. On resistive load the ripple (with harmonics) is no more than 3µV loaded for 5A on 5v rail, and 200µV on 15v rail (1A). Measurements points are on LPM side, before the final filtering module/connector.
The production is quite slowly in the beginning, mainly because the transformers, but I do hope it may be better later on. A first batch of 4-5 items it will be available in the next days...

Awesome Corus. How much will these be cost?
 
Member
Joined 2009
Paid Member
350$, excluding shipping...

BTW, minimized heat dissipation by customized transformer, sense control stage, dedicated regulator stage with adjustment for fan speed, grounding of the LPM is tied to the main system ground, to prevent ground loops (options for different grounding provided by design), are some of the features...
The back light for label is not only as fancy gimmick, but it have the function of a status indicator for the fan control circuit (this it can be also replaced by a led indicator).
In normal use into the player (no ventilation), the centre of the heat sink goes up to 45dg. C, while the main regulator area it measure 53dg.C
 
Member
Joined 2009
Paid Member
Assembling some LPM boards I got the good idea. Last minute improvement...
I double checked few times to ensure is nothing wrong with my measurements. Couldn`t believe it... Almost a like a battery.
Please take a look at these last measurements.
 

Attachments

  • Even better....jpg
    Even better....jpg
    484.8 KB · Views: 249
  • Even better...2.jpg
    Even better...2.jpg
    662.5 KB · Views: 244
noise from spearkers with oppo 105

a little off topic here, but It does involve the 105:
I was running a cheap $130 DVD player as a source until I got my ASi/Modwright/OPPO105 back from mod, and it idled silent with the DVD player..
Once I got the front end back, it now emits a hiss/buzz from the speakers.
at half volume, I can hear the hiss from the listening position.
The player runs into a jadis DPL pre, then on to a first watt F6.
There is the modwright power supply and the Asi power supply for the oscillator.
The first thing I tried was lifting the ground to each of the power supplies and oppo, one at a time, to no avail.
I'm starting to think its an impedance mismatch.
The oppo now has output transformers, so i'm not sure of the output impedance.
Not sure of the input impedance of the Jadis either.
Does thins sound likely?
Is there an easy way to measure the impedance of the oppo and preamp?
thanks
 
Assembling some LPM boards I got the good idea. Last minute improvement...
I double checked few times to ensure is nothing wrong with my measurements. Couldn`t believe it... Almost a like a battery.
Please take a look at these last measurements.

Great design and performance Coris ...now what do you claim these power supply improvements do over the stock 105 player?
 
Member
Joined 2009
Paid Member
a little off topic here, but It does involve the 105:
I was running a cheap $130 DVD player as a source until I got my ASi/Modwright/OPPO105 back from mod, and it idled silent with the DVD player..
Once I got the front end back, it now emits a hiss/buzz from the speakers.
at half volume, I can hear the hiss from the listening position.
The player runs into a jadis DPL pre, then on to a first watt F6.
There is the modwright power supply and the Asi power supply for the oscillator.
The first thing I tried was lifting the ground to each of the power supplies and oppo, one at a time, to no avail.
I'm starting to think its an impedance mismatch.
The oppo now has output transformers, so i'm not sure of the output impedance.
Not sure of the input impedance of the Jadis either.
Does thins sound likely?
Is there an easy way to measure the impedance of the oppo and preamp?
thanks

In my opinion it should be a grounding problem somewhere in your system. There is a little bit difficult to identify the issue from your description only... I think is not very important to measure the impedances. I would like to connect the source you use, directly to an amplifier, to observe if the noise it disappear. I do hope you will find out where the issue may come from...
 
Member
Joined 2009
Paid Member
Great design and performance Coris ...now what do you claim these power supply improvements do over the stock 105 player?

Well, I may tell you that I was previously quite sceptic when about this approach to use a serial power supply for a digital system. The digital system is noisy by its nature, and it will spoil the very good noise performances of a analogue PSU. As one can see in my previous screenshots, the noise levels it increase quite much when is to connect the serial PSU to the digital load.
However I had the opportunity to test myself a such serial PSU approach on my own 105. The differences are important, is only what I can say. I can not fully explain why it is like this, but I suppose a serial PSU it however lower much the overall noise in the system, comparing with a SMPS. A SMPS it actually induce lot of HF noises by its own, on large harmonic spectre, that it worse even more the noise performances of a digital system.
Powering the digital stage of these players by a analogue PSU it have an strong improvement impact on both sound and picture. I did used a such PSU on an already improved/modded 105 model. I was able to notice at once the differences after replacing an SMPS with the analogue PSU.
Lowering the noise levels in the video processing stage of the player it mean an increasing in tonal fidelity and richness of the picture. The tonal improvement in picture it mean actually a much better perception of the depth in a 2D picture. This fact correlated with a very good fine details resolution, as with the fidelity increasing in colour area, it make the picture just astonishing.
On a device with already improved video stage (better quality clock oscillators, battery powered, and better cooling of the active devices), it is very easy to notice the improvements when to power that system from a serial PSU. And a lower noise serial PSU it increase even more the obviousness of the improvements.
The similar results is to be observed on sound stage of the device, even that device it have almost everything improved on the signal path. There is actually the soundstage quality which get improved here.
Of course the description of these obvious improvements is more or less particular from person to person, but the overall conclusion is, it occur clear improvements for both picture and sound when to power the digital processing stage from a serial PSU/LPM. As better the noise quality of that PSU the more obvious the improvements become. This one may only experience by themselves to have the right conclusions...
 
Assembling some LPM boards I got the good idea. Last minute improvement...
I double checked few times to ensure is nothing wrong with my measurements. Couldn`t believe it... Almost a like a battery.
Please take a look at these last measurements.

Perhaps a scope with a 10:1 probe and a 1 mV division is not the best instrument to verify the noise performance of a low noise power supply.

Apart from this general comment, I don't understand the conclusion you draw from these pictures. The second set was taken at lower sensitivity (50 mV versus 10mV including probe conversion), and the peak values measured are considerably higher than in the first measurement. So to me it looks like matters are worse here.
 
Member
Joined 2009
Paid Member
Well, I do not intend to buy another scope to do these measurements... This former Agilent upgraded to 200Mhz bandwidth, and max memory aquisition is good enough to measure/appreciate a PSU noises, so with its 1mV scale division and the 350Mhz/10:1 probe.
Previous measurements was taken on High-resolution mode, 20mV scale with 1µV FFT scale. The second measurements was taken on same High-resolution mode, 10m scale with the same 1µV FFT. The FFT picks should be higher on second case. They are lower...
The normal mode measurements are worse in first/previous measurements, and the second ones shows a much "quiet" FFTs on an enough large spectre. These last measurements correlated with the improvements I have done it make very much sense, and it is exactly how it have to be, as consequence of the last modifications.
This LPM it perform over expectations, and as the test showed to me, the improvements of the powered device`s performances er very obvious.
At least one can test it as one may want and then reviewed it accordingly. I`m not afraid at all about.
And one more thing: one should remember that this LPM is not defined as an (ultra) low noise PSU. It is just a serial/analogue power supply, which perform very well, designated to power a digital board/system. As one can see, this LPM connected to the digital load it worse its noise performances, due to the noises generated and induced by the digital system. This is very normal. Even thought all these, the differences between a digital system powered by the original Oppo SMPS and the same system powered by this LPM are very important.
As the measurements shows very clear, my LPM it perform much better than Oppomod. This is a fact and shows that it successfully hold its goals.
 
Last edited:
Member
Joined 2009
Paid Member
I have registered that some 103 model owners, intend to use this LPM to power the digital only stage in their device.
I would like to observe about some things in such case.

My LPM total height is a little bit under 60mm. With the provided spacers, it have 60mm as standard height. This height it fit quite well the available space in a 103 enclosure (internal 65-67mm height).
There is a question about exhausting the heat dissipated by the LPM. The average power needs of an 103 model should be lower than the average power needed by an 105/105D model. So the heat dissipation it is quite low. However, I would like to suggest here a solution to lower the temperature, when using a serial PSU inside a device designed for a SMPS.
The 103 model do not have the available space and ventilation perforations existent in the later expensive models, so using a fan is not efficient at all.
As the hight of this LPM is very near to the one of the 103 enclosure, it may be a good solution to increase even more the hight of the LPM (using accordingly spacers, easy to buy on Ebay or so), so to have a few mm distance from the upper cover of the player. Then this small space it can be filled with a appropriate thickness thermal pad. I would like to use a 5mm thick thermal pad 100x100mm aplied over the LPM`s heatsink, and having so a good thermal contact with the whole upper cover, when this is mounted in its place.
This approach it ensure a very good cooling and it hold a normal low temperature inside the 103 enclosure. The thermal pads are also very easy to find and are cheap enough.
When I will have the precise spacer dimensions for an 103 model, I may provide these together with the LPM.
 
Member
Joined 2009
Paid Member
The measurements in this connection are made and shown more for comparative purposes, but not for a very detailed, precisely an strictly analytical task, to appreciate the absolute noise floor of a PSU (this LPM).

The approach in itself to design, and use the low as possible PSU, to power a whole digital stage (as the main board in Oppo players, or even a computer) is just a nonsense, and very far from a reasonable and efficient way to do it.
However, and for reasons I could not find yet a good explanation, when powering a digital system using a serial PSU, there are many improvements for the outputted digital signals/streams from that system. In case of Oppo players, there are huge improvements in both picture and sound.

Here are the facts so far in this field:
We have the original Oppo SMPS used in all their models, which exhibit (measured under the same approach(conditions) a very large HF spectre of noise, which it goes up to 200mVpp levels. This SMPS do not dissipate heat.
Alternative to this SMPS (as a serial approach) was/is the Oppomod PSU. This one exhibit a much reduced frequency spectre of noise, and the levels it goes up to few teens mVpp. Huge improvement over the original SMPS.
My LPM it exhibit noise only on the very low frequency spectre (25, 50, 100, 200, and 430Hz) of a level in µV range. This is just exceptional! One have no other choice than only admit this fact...

Well, the FFT pick levels are calculated as RMS, so as Vpp it will be higher than is to be seen in pictures.
The noise floor in FFT it may not be the same as the device noise floor. No matters in these appreciations. Who care about such details? Here is about differences in scales/ranges, from mV to µV... This it matter in fact, not exactly the figures.

Personally I`m very happy to have the opportunity to provide such results.
 
Well, I may tell you that I was previously quite sceptic when about this approach to use a serial power supply for a digital system. The digital system is noisy by its nature, and it will spoil the very good noise performances of a analogue PSU. As one can see in my previous screenshots, the noise levels it increase quite much when is to connect the serial PSU to the digital load.
However I had the opportunity to test myself a such serial PSU approach on my own 105. The differences are important, is only what I can say. I can not fully explain why it is like this, but I suppose a serial PSU it however lower much the overall noise in the system, comparing with a SMPS. A SMPS it actually induce lot of HF noises by its own, on large harmonic spectre, that it worse even more the noise performances of a digital system.
Powering the digital stage of these players by a analogue PSU it have an strong improvement impact on both sound and picture. I did used a such PSU on an already improved/modded 105 model. I was able to notice at once the differences after replacing an SMPS with the analogue PSU.
Lowering the noise levels in the video processing stage of the player it mean an increasing in tonal fidelity and richness of the picture. The tonal improvement in picture it mean actually a much better perception of the depth in a 2D picture. This fact correlated with a very good fine details resolution, as with the fidelity increasing in colour area, it make the picture just astonishing.
On a device with already improved video stage (better quality clock oscillators, battery powered, and better cooling of the active devices), it is very easy to notice the improvements when to power that system from a serial PSU. And a lower noise serial PSU it increase even more the obviousness of the improvements.
The similar results is to be observed on sound stage of the device, even that device it have almost everything improved on the signal path. There is actually the soundstage quality which get improved here.
Of course the description of these obvious improvements is more or less particular from person to person, but the overall conclusion is, it occur clear improvements for both picture and sound when to power the digital processing stage from a serial PSU/LPM. As better the noise quality of that PSU the more obvious the improvements become. This one may only experience by themselves to have the right conclusions...

Interesting observations Coris. Thanks!