jp, sorry it is not my intention to criticise the design.
My intention was that, in general, people on these forums should spend more time consider the way they are applying the designs that they chose, rather than the flavour of the month approach that happens in consumer audiophile circles. As DIYers, we should know better and apply critical thought in our decisions. I only mentioned the acko for comparison of its power supply and so that people could learn to observe the differences and make a comparison of markedly different approach to design for the same DAC chip.
It has been discussed a few times on different threads (I think on Ian's FIFO thread and Lucian's WaveIO thread, probably others that I have missed also) but a poorly made i2s connection will often do more harm than good for the signal transport, people should not automatically assume that i2s will be the 'magic potion' that unlocks digital audio.
My point basically is, whatever transport and DAC you chose, do it well and use the complementary strengths of each device. Simply picking the current GB's that will finish soon is not an audio system design or a (IMO) productive learning exercise. As a rule, DIYing in a hurry without proper research will end with wasted money and a less than optimal result.
In the case of the Subbu/JP ES9023 DAC, my thoughts are that the best performance will be had with SPDIF for your signal input.
Chris
My intention was that, in general, people on these forums should spend more time consider the way they are applying the designs that they chose, rather than the flavour of the month approach that happens in consumer audiophile circles. As DIYers, we should know better and apply critical thought in our decisions. I only mentioned the acko for comparison of its power supply and so that people could learn to observe the differences and make a comparison of markedly different approach to design for the same DAC chip.
It has been discussed a few times on different threads (I think on Ian's FIFO thread and Lucian's WaveIO thread, probably others that I have missed also) but a poorly made i2s connection will often do more harm than good for the signal transport, people should not automatically assume that i2s will be the 'magic potion' that unlocks digital audio.
My point basically is, whatever transport and DAC you chose, do it well and use the complementary strengths of each device. Simply picking the current GB's that will finish soon is not an audio system design or a (IMO) productive learning exercise. As a rule, DIYing in a hurry without proper research will end with wasted money and a less than optimal result.
In the case of the Subbu/JP ES9023 DAC, my thoughts are that the best performance will be had with SPDIF for your signal input.
Chris
Hi Chris, you are right on this. It is hard to make a low budget design that serves both the best way (so both SPDIF and I2S each the most optimal way). IMO it is either this or that. I don't agree with the flavor of the month remark as the 2.6 is a well proven design that took a lot of time. Makes me wonder how it compares to the AckoDAC despite all its technical improvements. Sometimes better technical specs are not the final word. Uh oh, that will give some pages of comment
Let's just say I could very well be using the 2.6 for years from now on. I just wish you had built the 2.6 if only for discussions sake.
Last edited:
Yes I bought an amanero too, thinking it would intgerate with the subbu, but it seems it is not ideal match?
Damn spdif, I wish it would die and go away!
I have read some posts recently
I think you can skip wm8804 and singal in directly from R7 R8 R9
if I'm wrong plz correct me! thanks..
I think you can skip wm8804 and singal in directly from R7 R8 R9
if I'm wrong plz correct me! thanks..
Perhaps I have not been clear regarding the amenero interfacing. I did not want to post again, but people still seem confused about what my point has been. Basically it is that there is a tendency to use i2s without consideration or understanding of the interfacing of high frequency signals.
On the v2.6 the 3 individual pins that are, in high frequency signal terms, a long way from the ground termination and without individual ground return paths. This makes a less than optimal interface for transfer of the i2s signals from any i2s source that is connected. If you are considering using i2s, please consider that it can very easily be done poorly and it's performance may actually inhibit the performance of the DAC. A wire terminated poorly in this area of a DAC can act as an antenna for noise. If this is the case you will not be getting the maximum from the DAC design as it is presented. This may have been addressed in the v3 design and become a non-issue.
Another consideration is that you will not easily be able to modify the v2.6 design for a well terminated synchronous master clock. Synch vs Asynch has been discussed in the main ES9022 thread. It would be totally off topic to discuss that any further here, though it is another possibility with the amenero or other USB->i2s transports.
The most novel features of the amernaro board are not compatible with the ES9023 DAC chip used. The ES9023 does not support input at greater than 192k and does not support DSD. With those two features excluded the amenero usb->i2s device will work with this DAC, at all of the data rates supported by the ES9023 chip.
I have put my name down for a v3 pcb. I have an overflowing queue of other projects that are begging for a slice of my spare time, so I am not sure when I will have time to build this up. I will build it and power from an LT1963 pre-reg supply and input isolated spdif from my WaveIO.
On the v2.6 the 3 individual pins that are, in high frequency signal terms, a long way from the ground termination and without individual ground return paths. This makes a less than optimal interface for transfer of the i2s signals from any i2s source that is connected. If you are considering using i2s, please consider that it can very easily be done poorly and it's performance may actually inhibit the performance of the DAC. A wire terminated poorly in this area of a DAC can act as an antenna for noise. If this is the case you will not be getting the maximum from the DAC design as it is presented. This may have been addressed in the v3 design and become a non-issue.
Another consideration is that you will not easily be able to modify the v2.6 design for a well terminated synchronous master clock. Synch vs Asynch has been discussed in the main ES9022 thread. It would be totally off topic to discuss that any further here, though it is another possibility with the amenero or other USB->i2s transports.
The most novel features of the amernaro board are not compatible with the ES9023 DAC chip used. The ES9023 does not support input at greater than 192k and does not support DSD. With those two features excluded the amenero usb->i2s device will work with this DAC, at all of the data rates supported by the ES9023 chip.
I have put my name down for a v3 pcb. I have an overflowing queue of other projects that are begging for a slice of my spare time, so I am not sure when I will have time to build this up. I will build it and power from an LT1963 pre-reg supply and input isolated spdif from my WaveIO.
Last edited:
Actually I shouldn't, I really have too much on my plate for the next 12months and don't think I could give an unbiased opinion anyway.
I will stay out of the thread for now, hopefully I have given some information and useful pointers for builders of this DAC to maximise the performance of their builds.
Hi,
I knew ES9023 couldn't do DSD and I knew it won't go above 192K when I build the dac and purchased the Amanero, both are highly affordable. The dac is easy for any modification and fun to play with. Interfacing it with any external USB adepter using I2S was some extra fun, might not be an optimal implementation but does work faily well and need not to be too negative on this and just keep the cable short.
You will not lose much with this DAC at this price.
I have a BD player and DAC using ES9018, another DAC using AD1955 and end up spending all my listening on the Subbu 9023.
best regards
I knew ES9023 couldn't do DSD and I knew it won't go above 192K when I build the dac and purchased the Amanero, both are highly affordable. The dac is easy for any modification and fun to play with. Interfacing it with any external USB adepter using I2S was some extra fun, might not be an optimal implementation but does work faily well and need not to be too negative on this and just keep the cable short.
You will not lose much with this DAC at this price.
I have a BD player and DAC using ES9018, another DAC using AD1955 and end up spending all my listening on the Subbu 9023.
best regards
I have both, I built 2 versions so the second one is without the WM8804 and using it with the Amanero.
Purely from S/PDIF, it better all my other DAC and player. With the Amanero/I2S it gives a bit better base, mid and highs are smooth and again these are from my own listening experience. PC is a 4 years old Asus notebook with 256 SSD runing Foobar and remote server for iphone control, amp are LM3875 (From PD) and Rowland model 2. Speakers are Aristoteles from Hørning Hybrid System.
Last edited:
hochopeper,
There is no reason for you to remove yourself from the discussion. You have actually come the closest to drilling down and explaining several elements of component integration that I and I'm sure others have been pondering. I'm sure we all want to end up with just one "Top Choice" DAC in our systems, but I suspect it will take some shuffling to get there. The more information the better.
I'll be presumptuous and start the new thread I mentioned as an attempt to establish a collection/comparison platform that is comprehensive and leaves the individual development threads focused on more narrowly defined subject matter.
There is no reason for you to remove yourself from the discussion. You have actually come the closest to drilling down and explaining several elements of component integration that I and I'm sure others have been pondering. I'm sure we all want to end up with just one "Top Choice" DAC in our systems, but I suspect it will take some shuffling to get there. The more information the better.
I'll be presumptuous and start the new thread I mentioned as an attempt to establish a collection/comparison platform that is comprehensive and leaves the individual development threads focused on more narrowly defined subject matter.
syklab said:
making the connection short doesnt mean all that much i'm afraid, just means the reflections return quicker, without a return path per signal with i2s you turn the connection into an RF transmitter.
its not a matter of quality we are talking about here, we are just trying to inject a bit on info into a thread that seems to have quite a few people who bought these objects without properly thinking it through, or having the experience/knowledge to know what works properly and what is a hack, connecting i2s to this dac is a hack.
Using external i2s over a cable is a delicate operation, it was never meant for cables, only short impedance controlled PCB traces. to connect it with a cable and not do more harm than good requires method that has not been designed into this 2.6 version dac. thats not a slight on the design, that was not part of the design considerations, it was designed as an spdif dac.
I posted this line elsewhere just before, but I feel its worth repeating as its topical here.
i2s, particularly if it includes MCLK is not really a normal digital signal, it is much more similar to a high frequency analogue RF signal that needs to be low noise. thats not an easy combination, clock is a pure analogue signal that is used as the very basis of the analogue audio signal
Last edited:
exactly...
well you could scrap off some soldermask on nearby parts of the PCB to expose ground and use those for the other ground wires
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.