Thx. 3A would have your Rcore transformer at its max rating so I see that's the limit of your PSU (for all voltage needs combined).
I have ordered a 203 but it hasn't shipped yet so I'm not yet in a position to look inside the unit itself. (I've also put myself down as interested in a final run of the 205.) What voltages are required to be supplied by the PSU? From heyj's pics I see the 203's analogue board requires +/- 15V but, like him, I wouldn't be worrying about that. Does the digital board only expect 12V? What about the 205?
I have ordered a 203 but it hasn't shipped yet so I'm not yet in a position to look inside the unit itself. (I've also put myself down as interested in a final run of the 205.) What voltages are required to be supplied by the PSU? From heyj's pics I see the 203's analogue board requires +/- 15V but, like him, I wouldn't be worrying about that. Does the digital board only expect 12V? What about the 205?
Well, based on my previous experience with the former Oppo models, I finished quite fast all the improvements for 205 model. Except the post DAC analogue stage. The new chip performances it was a challenge for finding the best or at least a good way to take advantage of this new chip capabilities.
There was/are few alternative: one is improving the Oppo design of this stage. The opamps used by Oppo designers are of a very good quality, but it do perform this design as it should in conjunction with the new Sabre Pro chip? Well, I was not satisfied at all, about what I found on this section. A little bit improved old design…
Another alternative, very convenient and enough simple is using transformers. I used some Lundahl, maybe not just the right ones, which it fitted well enough for the new DAC chip. It sounded good, but the outputted signal it was low enough, and not so much to be done to fix this. In addition, the signal dynamics it was not as I expected from the advertised figures for SabrePro…
One alternative more, it was using the quite professional design and concept of the TPA guys, with their I/V stage for SabrePro, called Mercury. Well, this alternative is only for personal use, it is quite expensive and not convenient at all, as it is seldom available for sale (because too big demand). However, I had to give it a try. It fitted very well into the Oppo. As my approach for this stage is modular, I used the installed motherboard for my module, adapting very easy the Mercury to it. And it worked like a charm. Increased dynamics, and all the rest of high quality audio out of the SabrePro DAC.
So, my last alternative (so far) to solve this problem: my old analogue module, I have used with very good success for ESS9018 in previous Oppo models.
I used initially exactly the same circuits based on OPA1632 (I/V followed by a buffer), as for ESS9018 chip. The result it was (expected also) quite disappointing… The outputted signal too low and not very special quality whatsoever. Knowing the new DAC chip it deliver a large current on its outputs, I appreciated that the I/V stage in my module it have to be adapted accordingly to ES9038Pro. Lot of experimenting with this I/V stage, not so successfully… I have also noticed that if the feedback resistor of the I/V circuit, it is getting lower and lower, the results become better and better… Until I came to a few ohm resistor… Well, it was enough obvious for me, that at a few ohm FB resistor for OPA1632 I/V stage, that it is no more a I/V stage, but something else, or nothing…I just bypassed it, disconnecting it from power, and connecting directly the DAC output channels (2+2 unified), to the OPA1632 buffer. Bingo! It worked! And how nice it worked! It was indeed a revelation…. I could finally experience and enjoy the amazing dynamics of the new ESS9038Pro. Huge! And even better than I could get out of Mercury (in the very low end of spectre)… The base instruments it send out indeed the vibrations through the speakers, exactly as in live music. Even the lowest and subtle note or sound in lowest end, it is felt it, not only heard it. But how it works so? Well, I cannot answer yet to this question…
A just perfect match and so unbelievable simple. And this buffer attached directly to the DAC outputs it can be customised for the right output level…
Therefore, I thought making some measurements, comparing the Mercury with this my module, where only the OPA1632 buffer section is used. The results of the measurements are to be seen hereby.
Some words about Mercury environment. The feedback and gain of this device it can be customised. I used the recommended gain setup for ESS9038Pro, but I got a 1,4Vpp output level on balanced. Too low! Then, I have set it the gain as for ESS9028 (243 Ohm FB resistors), and getting so a more acceptable output level, as seen in the screenshots. The power for Mercury was set to +/- 12v.
I have set it the gain for my buffer module so to get a convenient level for my system (6Vpp).
As you can see out of these screenshots, the Mercury bandwidth in this installation it get quite worse with frequency (at 80khz), while the only OPA1632 buffer connected to SabrePro chip output it perform a little bit better at higher frequencies. Also, look at the FFT graphic calculations in the screenshots. FFT it is on Volt scale.
The figures in itself are not very important, or relevant, but more important is the comparison in between these two solutions (a well and original designed Mercury, that cleverly take advantage of DAC capabilities, and a simple OPA1632 buffer directly coupled to the DAC outputs).
I think the attached pictures it have to be downloaded first, as here are quite small...
There was/are few alternative: one is improving the Oppo design of this stage. The opamps used by Oppo designers are of a very good quality, but it do perform this design as it should in conjunction with the new Sabre Pro chip? Well, I was not satisfied at all, about what I found on this section. A little bit improved old design…
Another alternative, very convenient and enough simple is using transformers. I used some Lundahl, maybe not just the right ones, which it fitted well enough for the new DAC chip. It sounded good, but the outputted signal it was low enough, and not so much to be done to fix this. In addition, the signal dynamics it was not as I expected from the advertised figures for SabrePro…
One alternative more, it was using the quite professional design and concept of the TPA guys, with their I/V stage for SabrePro, called Mercury. Well, this alternative is only for personal use, it is quite expensive and not convenient at all, as it is seldom available for sale (because too big demand). However, I had to give it a try. It fitted very well into the Oppo. As my approach for this stage is modular, I used the installed motherboard for my module, adapting very easy the Mercury to it. And it worked like a charm. Increased dynamics, and all the rest of high quality audio out of the SabrePro DAC.
So, my last alternative (so far) to solve this problem: my old analogue module, I have used with very good success for ESS9018 in previous Oppo models.
I used initially exactly the same circuits based on OPA1632 (I/V followed by a buffer), as for ESS9018 chip. The result it was (expected also) quite disappointing… The outputted signal too low and not very special quality whatsoever. Knowing the new DAC chip it deliver a large current on its outputs, I appreciated that the I/V stage in my module it have to be adapted accordingly to ES9038Pro. Lot of experimenting with this I/V stage, not so successfully… I have also noticed that if the feedback resistor of the I/V circuit, it is getting lower and lower, the results become better and better… Until I came to a few ohm resistor… Well, it was enough obvious for me, that at a few ohm FB resistor for OPA1632 I/V stage, that it is no more a I/V stage, but something else, or nothing…I just bypassed it, disconnecting it from power, and connecting directly the DAC output channels (2+2 unified), to the OPA1632 buffer. Bingo! It worked! And how nice it worked! It was indeed a revelation…. I could finally experience and enjoy the amazing dynamics of the new ESS9038Pro. Huge! And even better than I could get out of Mercury (in the very low end of spectre)… The base instruments it send out indeed the vibrations through the speakers, exactly as in live music. Even the lowest and subtle note or sound in lowest end, it is felt it, not only heard it. But how it works so? Well, I cannot answer yet to this question…
A just perfect match and so unbelievable simple. And this buffer attached directly to the DAC outputs it can be customised for the right output level…
Therefore, I thought making some measurements, comparing the Mercury with this my module, where only the OPA1632 buffer section is used. The results of the measurements are to be seen hereby.
Some words about Mercury environment. The feedback and gain of this device it can be customised. I used the recommended gain setup for ESS9038Pro, but I got a 1,4Vpp output level on balanced. Too low! Then, I have set it the gain as for ESS9028 (243 Ohm FB resistors), and getting so a more acceptable output level, as seen in the screenshots. The power for Mercury was set to +/- 12v.
I have set it the gain for my buffer module so to get a convenient level for my system (6Vpp).
As you can see out of these screenshots, the Mercury bandwidth in this installation it get quite worse with frequency (at 80khz), while the only OPA1632 buffer connected to SabrePro chip output it perform a little bit better at higher frequencies. Also, look at the FFT graphic calculations in the screenshots. FFT it is on Volt scale.
The figures in itself are not very important, or relevant, but more important is the comparison in between these two solutions (a well and original designed Mercury, that cleverly take advantage of DAC capabilities, and a simple OPA1632 buffer directly coupled to the DAC outputs).
I think the attached pictures it have to be downloaded first, as here are quite small...
Attachments
Last edited:
I am just looking for some clarity regarding swapping out the AC coupling caps on the 205.
I am using the unbalanced stereo pair for LF and RF. I believe that there is only one coupling cap for each channel.
I don't want to wire-out these caps because I am simply not sure I can.
Working through this thread I can see 3 options suggested
1. Os-con Polymer (100uf) with a polyester bypass (1uf).
2. Non-polar electrolytics (100uf ??)
3. Silmic II at 470uf (although I think that was for a 203, which may be different).
I'm not familiar with the os-con type cap so I am unsure which is the better option.
I can get either panasonic SEPC or Nichicon NS series 100uf 16 v
Audio Note Kaisei has a non-polar electrolytic (100uF 25V) with 5mm pitch which should fit and I think the Mundorf ECap AC (100uF 50V) audio coupling bipolar electrolytics could be made to fit.
If the electrolytics were preferred would the 1uF bypass still be useful?
I am using the unbalanced stereo pair for LF and RF. I believe that there is only one coupling cap for each channel.
I don't want to wire-out these caps because I am simply not sure I can.
Working through this thread I can see 3 options suggested
1. Os-con Polymer (100uf) with a polyester bypass (1uf).
2. Non-polar electrolytics (100uf ??)
3. Silmic II at 470uf (although I think that was for a 203, which may be different).
I'm not familiar with the os-con type cap so I am unsure which is the better option.
I can get either panasonic SEPC or Nichicon NS series 100uf 16 v
Audio Note Kaisei has a non-polar electrolytic (100uF 25V) with 5mm pitch which should fit and I think the Mundorf ECap AC (100uF 50V) audio coupling bipolar electrolytics could be made to fit.
If the electrolytics were preferred would the 1uF bypass still be useful?
The best solution about this AC coupling is to bypassing it completely (shorting the existing caps). In such case you should ensure there is a low DC offset on that outputs (measure the DC offset before the caps). It should be of few mV only.
If you should keep the AC coupling approach, then is better using non-polar caps bypassed by some good smaller film caps.
In the picture hereby are marked the RCA AC coupling caps.
If you may want removing the existing caps, then I suggest braking it by moving it back and forth, rather than unsolder it so as it are. After brecking their legs, it is easier removing (unsolder) only the rest of their terminals.
If you should keep the AC coupling approach, then is better using non-polar caps bypassed by some good smaller film caps.
In the picture hereby are marked the RCA AC coupling caps.
If you may want removing the existing caps, then I suggest braking it by moving it back and forth, rather than unsolder it so as it are. After brecking their legs, it is easier removing (unsolder) only the rest of their terminals.
Attachments
Last edited:
Coris
The reason for the low output on the Mercury is that this is designed assuming that 4 channels of the DAC are in parallel. The 205 actually only uses one pair of the channels of the DAC chip. It uses another pair for the headphones and yet another for the single ended output with one pair spare. To get the designed gain, you need to make the IV resistor bigger x4. TBH the 205 is only giving you a quarter of what the ES9038Pro is capable of.
When you just connect the buffer directly to the DAC, you are making the DAC operate in voltage mode (rather than current) and sacrificing may be 3-6dB of performance. I would try the Mercury with a higher gain to see how it sounds for a like for like comparision.
Yes, it is right your jugement. I have connected the Mercury to the two of the DAC chip output channels unified (tied together). Indeed, the gain resistor for Mercury, is in this case 4 times bigger (240R) than for 4 channels (60R).
Now I know why it is working this way, with only an attached buffer amp stage. If I would read (or look at) more carefully the SabrePro chip datasheet, at the right time, I could save much of my experimenting time. In the end I came to the same post DAC analogue circuit as the one recommended by ESS (for current mode operation)... I may not agree with your assertion that I use the DAC output in voltage mode (sacrificing 3-6dB of performance)... Also I found a simple way to control and reduce the residual HF noises on DAC output. A way which it not reduce the bandwidth in higher end, and also it improve even more the sound scene quality.
I got a a linear bandwidth from 10Hz to 80Khz, for a customised signal level on output. So, it is no any performance sacrificing using my circuit, but the opposite. And it is about only two OPA1632 for balanced outputs.
I will publish soon another comparative measurements, for my improved now module. I have to admit that I was surprised getting a slightly better sound and quality than from Mercury, when about volume of the sound scene, dynamics, as a very solid, balanced, powerful and smooth low end (in this Oppo installation).
Last edited:
I can’t actually see any schematics so I am only guessing at what you are doing so please don’t be offended by any of my comments.
What is your view of the 7.1 analogue output sound quality for the 203 compared to the 205? See my separate post.
No, I am not offended by your comments. There is all right. I only answered and commented myself your post.
I agree about schematic... I should published the ESS schematic, but I am not allowed... You know...
I tested/listened only to the stereo part of the analogue stage of these models. The stock 203 is not very special, medium to low in my opinion. About a stock 205 I did not liked at all the XLR/RCA outputs, but headphone out it sounded much better. I assume the multi-channel outputs may not sounds better than stereo ones, as the mute function is designed with transistors...
There is a slightly overall better quality for 205, as it is based on a new generation DAC chip, which it however provide a different sound than the predecessor. Here is more about the exigence of the listener... For someones the stock 205 it sounds nice.
For sure you have to experience live how it sounds these two last models, to make your own appreciation. Else impressions in this field may be quite subjective, from a listener to another. However, it sounds wonderful after improvements (as it should, actually...).
I agree about schematic... I should published the ESS schematic, but I am not allowed... You know...
I tested/listened only to the stereo part of the analogue stage of these models. The stock 203 is not very special, medium to low in my opinion. About a stock 205 I did not liked at all the XLR/RCA outputs, but headphone out it sounded much better. I assume the multi-channel outputs may not sounds better than stereo ones, as the mute function is designed with transistors...
There is a slightly overall better quality for 205, as it is based on a new generation DAC chip, which it however provide a different sound than the predecessor. Here is more about the exigence of the listener... For someones the stock 205 it sounds nice.
For sure you have to experience live how it sounds these two last models, to make your own appreciation. Else impressions in this field may be quite subjective, from a listener to another. However, it sounds wonderful after improvements (as it should, actually...).
Last edited:
Very very simply said, and generalizing a lot.... the 203 in overall sounds like the treble setting is just one notch up in comparison to the 205. That is what I noticed the most with every source.
Than a bit more in depth:
Especially with "low bitrate" sources like cd's it is sounds likes the 203 is just trying to get more sound out of it than there is in the actual source, and so it sounds a bit unnatural.With "high bitrate" sources this effect is less, and it sound a lot more the way it should be, and how it was meant to be by the producers. (I am talking about good records here, a bad recording always stays bad, no matter what equipment you play it on).
A good LPM power supply makes the general "treble problem" a lot less. But still I think that the 203 analog output is not that great with cd's. I got a LPM PSU and replaced the clocks for better ones, and it all is a big improvement. But I still play all my cd's through the DAC's of my HK AVR8500, for everything else I use the analog outputs of my modified 203
The unmodified 205 sound better with "low bitrate" sources. Even with my 203 modifications I feel that the unmodified 205 sound better with cd's(low bitrate) than my modified 203.
With the "high bitrate"sources I like my modified 203 more than the unmodified 205.
But all of this is off course just my personal experience and that of a few of my friends with the same hobby with whom I compared things....
But don't forget that the unmodified 203 and 205 still sound a lot better than most stuff on the market today. So my remarks are to bee seen in a relative way.
Greetings Jack
(Sadly enough my PSU has broken down, and I am waiting for the replacement, so I am currently listening to my music on my Marantz dv8300 player)
Than a bit more in depth:
Especially with "low bitrate" sources like cd's it is sounds likes the 203 is just trying to get more sound out of it than there is in the actual source, and so it sounds a bit unnatural.With "high bitrate" sources this effect is less, and it sound a lot more the way it should be, and how it was meant to be by the producers. (I am talking about good records here, a bad recording always stays bad, no matter what equipment you play it on).
A good LPM power supply makes the general "treble problem" a lot less. But still I think that the 203 analog output is not that great with cd's. I got a LPM PSU and replaced the clocks for better ones, and it all is a big improvement. But I still play all my cd's through the DAC's of my HK AVR8500, for everything else I use the analog outputs of my modified 203
The unmodified 205 sound better with "low bitrate" sources. Even with my 203 modifications I feel that the unmodified 205 sound better with cd's(low bitrate) than my modified 203.
With the "high bitrate"sources I like my modified 203 more than the unmodified 205.
But all of this is off course just my personal experience and that of a few of my friends with the same hobby with whom I compared things....
But don't forget that the unmodified 203 and 205 still sound a lot better than most stuff on the market today. So my remarks are to bee seen in a relative way.
Greetings Jack
(Sadly enough my PSU has broken down, and I am waiting for the replacement, so I am currently listening to my music on my Marantz dv8300 player)
Attachments
Last edited:
The Oppo 203 it was not meant by its designers, as a audiophile device. The stock audio board of 203, based on AKM DAC is meant for audio monitoring purposes only.
Indeed, how the analogue audio section of 203 it was designed and produced, it can provide only average audio quality. 203 model is meant mainly as a digital transport, when the user it will mostly be interested in using the digital outputs and its 4k picture capabilities. At least the lower price of this device it confirm the above assertion.
The 205 model is meant as an high end (so called) device for both picture and sound (sold for a convenient price). Indeed, the 205 hardware potential it is for a high end device, but the design and implementation (low production cost), it not use the whole hardware quality potential this model it contain, and it is capable for.
Both these devices it need improvements to perform accordingly at their highest quality potentials.
An improved 203 audio section (DAC board) it perform actually very good, and it provide a lot better sound quality, than a stock 205 does. AKM DAC chip is very capable, and it perform excellent in a properly upgraded environment. There is not enough only improving the 203 with a linear PSU and another clock system, for getting a good audio quality out of it.
This it apply also for 205 model as well. Its pretended design for high end audio section, it need extended improvements to perform at a real high end level device.
Indeed, how the analogue audio section of 203 it was designed and produced, it can provide only average audio quality. 203 model is meant mainly as a digital transport, when the user it will mostly be interested in using the digital outputs and its 4k picture capabilities. At least the lower price of this device it confirm the above assertion.
The 205 model is meant as an high end (so called) device for both picture and sound (sold for a convenient price). Indeed, the 205 hardware potential it is for a high end device, but the design and implementation (low production cost), it not use the whole hardware quality potential this model it contain, and it is capable for.
Both these devices it need improvements to perform accordingly at their highest quality potentials.
An improved 203 audio section (DAC board) it perform actually very good, and it provide a lot better sound quality, than a stock 205 does. AKM DAC chip is very capable, and it perform excellent in a properly upgraded environment. There is not enough only improving the 203 with a linear PSU and another clock system, for getting a good audio quality out of it.
This it apply also for 205 model as well. Its pretended design for high end audio section, it need extended improvements to perform at a real high end level device.