MDAC First Listen (part XXIX)
- Thread starter Asphaltradler
- Start date
- Status
JohnW
pfm member
Ale,
OK, I'll work on your unit after I get a few hours sleep - now seems like a good time to "take a break" from development and give my mind sometime to gather its thoughts for the next stage of MDAC2 design.
iansr
pfm Member
Knock it off John, I've got this thick foamy liquid stuff running out of the corners of my mouth and dripping on to my chest. It's not a good look.
JohnW
pfm member
Yep - you unit had the Multicomp resistors fitted... it should be ready tomorrow AM, I'll Email you with shipping cost (and the 4x Vishey "Toy" resistors).
JemHayward
pfm Member
How about a version of the MDAC2 board without the analogue stage populated, and then build the TDAC output stage in a separate box with its own high voltage PSU, with some form of cable connecting the two, and then the TDAC box could have an analogue input fed from our Khozmo attenuator...
(I've probably just demonstrated my ignorance...)
Reverted back from nCore monoblocks this evening, after a year or so, to my retired integrated ShengYa A17CS (see Vincent A10).
Integrated has tube presection with transistor powersection. Will later try to switch between Sovtek (AX12LPS now) and Brimare tubes in the pre section to evaluate if there's any difference.
MDAC @ 0db, never expected this huge difference when switching, took me & spouse (better hearing) by surprise.
Switching amp from class D (nCores) to class A (first 10 W's then AB) took me by surprise, turned off my old AudioPro subwoofer. Not needed when using the cheapish chinese Shengya A17CS. If being honest to nCores, guess they reveal details,
/Cheers (and I hope that all of you cheers our Swedish olympic competitors)
Lars
Integrated has tube presection with transistor powersection. Will later try to switch between Sovtek (AX12LPS now) and Brimare tubes in the pre section to evaluate if there's any difference.
MDAC @ 0db, never expected this huge difference when switching, took me & spouse (better hearing) by surprise.
Switching amp from class D (nCores) to class A (first 10 W's then AB) took me by surprise, turned off my old AudioPro subwoofer. Not needed when using the cheapish chinese Shengya A17CS. If being honest to nCores, guess they reveal details,
/Cheers (and I hope that all of you cheers our Swedish olympic competitors)
Lars
laverda
pfm Member
JohnW, out of interest, do you have any 'Toy' resistors kicking about? You should shortly receive my Sovereign+ board, and I did wonder how much of a job the Toy update would be? I'm still totally committed to getting MDAC2, but I figured it was worth asking.....
Ah Ha......
John W has a set of TOY resistors shipped by HFC for me many months ago for this upgrade......but things have moved on...apace !!
I was going to ask JohnW if there was any chance of a deal of some sort for a new board ...... What do you think John.
Regards
Graham
jamjar784
Heavy Up
Hi John
I've been looking for information about the ESS left channel locking bug. It seems there are two web sites now http://mdac2.lakewestaudio.com/ and http://mdac2.referata.com/wiki/Main_Page but the Wiki isn't working any more which had the firmware upgrades and other useful info.
I previously downloaded several of the firmware versions Dominik created, but the verdict seemed to be that these affected SQ a little so you only recommend 0.90 and A.10? Is mdac-update_A.10.exe (Feb 2013) this version? What is included in this version - does it help with the ESS left channel bug?
In the new MDAC 2 are you only using the right channels of the 2 ESS DAC chips or how do you plan to eliminate this problem?
Many thanks
PS Anyone interested in the answers to my questions may find this "M-DAC Firmware Updates/Other Useful Information" by finesensations, Hubert481, sam_cat, and others helpful.
I've been looking for information about the ESS left channel locking bug. It seems there are two web sites now http://mdac2.lakewestaudio.com/ and http://mdac2.referata.com/wiki/Main_Page but the Wiki isn't working any more which had the firmware upgrades and other useful info.
I previously downloaded several of the firmware versions Dominik created, but the verdict seemed to be that these affected SQ a little so you only recommend 0.90 and A.10? Is mdac-update_A.10.exe (Feb 2013) this version? What is included in this version - does it help with the ESS left channel bug?
In the new MDAC 2 are you only using the right channels of the 2 ESS DAC chips or how do you plan to eliminate this problem?
Many thanks
PS Anyone interested in the answers to my questions may find this "M-DAC Firmware Updates/Other Useful Information" by finesensations, Hubert481, sam_cat, and others helpful.
http://mdac.referata.com/wiki/Main_Page is the page to use, the wikkii one was discontinued some time ago. AFAIK JohnW prefers 0.90, but many others lean towards A.10 and A.08, so listen for yourself and take a pick.
Regarding the left channel bug - I remember John mentioning that in one of these first listen threads - yes, using only the right channels is the way to go, at least for now.
Also, that referenced firmware updates thread is rather old.
jamjar784
Heavy Up
http://mdac.referata.com/wiki/Main_Page is the page to use, the wikkii one was discontinued some time ago. AFAIK JohnW prefers 0.90, but many others lean towards A.10 and A.08, so listen for yourself and take a pick.
Regarding the left channel bug - I remember John mentioning that in one of these first listen threads - yes, using only the right channels is the way to go, at least for now.
Also, that referenced firmware updates thread is rather old.
Hi Jirij
Thanks for the link, that has exactly what I was looking for. V.0.99 has some workaround code for the LCL bug.
I also found some posts from John and Dominik about the ESS 9018 problem in this thread JohnW 9th Aug 2013
The left channel locking bug is related to the ESS's internal SPDIF interface - MDAC will use an external SPDIF decoder... unfortunately(ignoring the Locking bug) the ESS offers greater jitter attenuation via its SPDIF block then via I2S ports... but we are left with little option.
John/Dominik
Did the 0.99 code for LCL bug make it into A.10?
Could you explain a wee bit more about how the SPDIF interface is causing this problem as I get this using USB interface?
Thanks
JohnW
pfm member
To reduce the number of data lines toggling within the MDAC for its internal data transportation we use the benefit of single line SPDIF. The SPDIF is used only for DATA recovery - the clock is generated by the master clock so bypassing the deficiencys of SPDIF.
The ESS has a design bug with its internal SPDIF block which with the benefit of hindsight we will not be using on the MDAC2 design.
Im surprised you have the Left Channel Bug with USB keep D3E set to Full suppression.
Avoid 0.99 software its the worst sounding try A08 . As far as I can recall, A10 does not implement the Left Channel Locking bug work around.
JohnW
pfm member
Ale79,
Your units on test, I'll ship it on Tuesday - I'm in the process of upgrading JTC unit to full Toy and would like to compare to your unit as there are slight differences in the upgrade build (I'll then upgrade both to the best sounding).
JTC,
As I love inflecting pain upon myself - I've upgraded your PCB to full toy + AV bypass, I'll charge you just for shipping + the Toy resistors... BTW, it appears it was the Crystal that failed (there was also a couple of capacitors in distressed which I don't understand the reason for - I can only presume QC issues).
Tonyb, your attenuators will be shipped out on Wednesday.
Tomorrows wiped out for both Renata and I will be making same day return flights to retain our Airmiles which would otherwise expire by the end of March... Not good for our CO2 footprint I know, basically we board a plane to the cheapest destination, and return back to Prague on the return flight (unfortunately Renata and I have to take different flights as someone has to mind the pups).… not sure what the flight crew will think of our joy rides… waste of time and money. To make best use of the trip to Prague I'll try to visit the Surface mount production line that I have in mind for the MDAC2 PCB(s).
Your units on test, I'll ship it on Tuesday - I'm in the process of upgrading JTC unit to full Toy and would like to compare to your unit as there are slight differences in the upgrade build (I'll then upgrade both to the best sounding).
JTC,
As I love inflecting pain upon myself - I've upgraded your PCB to full toy + AV bypass, I'll charge you just for shipping + the Toy resistors... BTW, it appears it was the Crystal that failed (there was also a couple of capacitors in distressed which I don't understand the reason for - I can only presume QC issues).
Tonyb, your attenuators will be shipped out on Wednesday.
Tomorrows wiped out for both Renata and I will be making same day return flights to retain our Airmiles which would otherwise expire by the end of March... Not good for our CO2 footprint I know, basically we board a plane to the cheapest destination, and return back to Prague on the return flight (unfortunately Renata and I have to take different flights as someone has to mind the pups).… not sure what the flight crew will think of our joy rides… waste of time and money. To make best use of the trip to Prague I'll try to visit the Surface mount production line that I have in mind for the MDAC2 PCB(s).
JohnW
pfm member
Interesting, which has the better Sound stage?
I'm so anti Putin (and the level of corruption) that I refuse to watch the Olympics! but good luck
(Guess you guys are one of the favourites for the Ice Hockey)!
JohnW
pfm member
All this TDAC talk makes me hungry
Keep up the good work John!
Ok, latest development is to added position for "headers" on the MDAC2 PCB which will allow us to add the Tube output stage at a later date. So owners of the MDAC2 can "Upgrade" to the Tube platform one its released (the MDAC2 will have to be returned to Czech Rep for Tube option).
Unfortunately, the difficulties with the "TDAC" is not so much the tube circuit itself, but we will need to rehouse the MDAC2 PCB + Tube section with its extra PSU into a new chassis - its this new chassis that adds the extra complexity. At this time, I'm not sure how we will handle this... Maybe Arek can help out the Mechanical work, I dread European pricing on Mechanical parts…
As I’ve now come to terms with the fact that the MDAC2 design will be based on 2 PCB’s – the logical step is to totally isolate the Digital and analogue stages – so one PCB contains the Dual ESS DAC’s, Clock, Analogue stage and PSU. While the second PCB (mounted above) – has the FPGA, USB and digital inputs etc.
We can add a galvanic isolation barrier between these 2 PCB’s so the critical “Analogue stage” is isolated from the “Noisy” digital sections, insuring the lowest analogue ground noise. While added complexity to the project in design time its has many advantages – not least that it’s really the correct method to do it:-
1. Should practically eliminate “Second order” digital processing effects on sound quality (Sound quality should not be effected by software routines etc).
2. Provides galvanic isolation from the digital section if the additional PSU is installed (an option).
3. Provides a future upgrade path (either the Digital or Analogue PCB can be upgraded in the future at reduced cost without scrapping the whole unit).
4. Allows as to pre-release a simpler interim Digital board very quickly (no FPGA and only supporting USB upto 24/96) – so basically the same functionality of MDAC but with the improved sound quality of the MDAC2 (Dual ESS + new output stage etc) with galavincally isolated USB etc. This allows those waiting for the MDAC2 to gain the improved sound quality ahead of the “official” release of the fully featured MDAC2, the extra MDAC2 “Digital” features can be added later with a simple board upgrade by the owner with the “fully featured” MDAC2 digital board.
This de-risks the project, allowing us to swiftly deliver the “basic” improved sound quality of MDAC2 while we continue the development work on the more advanced digital board with the promised features (DSD64, 128, 256, PCM 384fs + FPGA etc), so now a 2 stage process, but allowing us to deliver the MDAC2 "Sound quality" closer to the original target dates
MDAC2 audio performance will be further improved once we start introducing the more advanced FPGA firmware upgrades (upgraded via a software download) these will incorporate a more advanced digital filter etc…
The "Interim" Digital board will be sold at a nominal cost to cover its production and development (lets say around GBP150).
Comments? (Good idea or bad idea?)
- Status