Hello ! Just bought K11 dac/headphones amp, here is board photo with some measurements. There are 18 regulators, with 2 independed amps for 4.4mm phones. First of all I don't like polymer caps in analog circuit, and already swiped one op amp to opa1656. Maybe the pictures will be useful for someone to repair
First of all, I think it will be good to make linear power supply. What do you think about it needed to be regulated or not with such amount of regulators onboard ? Maybe powerfull transformer 9V, diode bridge and 4700uF will be enough for it ?
How much money are you willing to spend modding it?
Also, how much better would you expect it to sound after the mods?
Also, how much better would you expect it to sound after the mods?
Of course this should be an amount not exceeding the cost of the device itself, but I'm don't thinking seriosly about this, it is a hobby so it is not important about money. I expect this can sounds better, amount of better at this time is hard to measure ))
Now I'm thinking about choosing a power supply, have three interesting variants: Base On STUDER 900 Circuit Power Supply, S35-MKII Fully Discrete HIFI Linear Power Supply or board with 4pcs LT3045 in parallel
Now I'm thinking about choosing a power supply, have three interesting variants: Base On STUDER 900 Circuit Power Supply, S35-MKII Fully Discrete HIFI Linear Power Supply or board with 4pcs LT3045 in parallel
Last edited:
Current progress...
I've changed LDOs for VA/VL/VD to LP5907, also many caps. The power source now is low noise Pre-Amp / DAC power supply MJE15034 TL072 Regulator based on STUDER 900
I've changed LDOs for VA/VL/VD to LP5907, also many caps. The power source now is low noise Pre-Amp / DAC power supply MJE15034 TL072 Regulator based on STUDER 900
Attachments
Hello! I have Fiio K11 R2R, the same DAC/AMP combo except, obviously, DAC part, analog part is almost identical. And yes, I`m totally agree about using some normal power supply with this device because there is clear difference in sounding comparing to stock PSU.
Hmm, couple of Nichicon ES caps and four of WIMA caps in audio signal circuit, nice nice) What capacitance you`ve used there?
Hmm, couple of Nichicon ES caps and four of WIMA caps in audio signal circuit, nice nice) What capacitance you`ve used there?
Attachments
WIMA 4.7uF, there is no place for bigger value. At rca output side there is no DC offset at point with ES caps in K11, but I have put this DAC on hold for now. I'll write later about what else I found for improvement in this boxWhat capacitance you`ve used there?
Found that USB controller SA9312 uses passive crystal 12Mhz (in 4pin case). Pin 1 - XTO, pin 3 - XTI
Seems to me it will be better to put here good quality active crystal with output to XTI, this controller makes MCLK and in fact a half or more of work in making audio in whole DAC if using USB.
Seems to me it will be better to put here good quality active crystal with output to XTI, this controller makes MCLK and in fact a half or more of work in making audio in whole DAC if using USB.
IMO better PS is always better if you don't care about it price.Any improvement in using a better PS?
Last edited:
better to put here good quality active crystal with output to XTI, this controller makes MCLK
IMHO, the PLL inside will ruin everything anyway...
Alex.
How do you think, does it make sense to try to cut external MCLK and put active crystal to CS43198 for use it's internal PLL instead of USB controller one ? (if CS chip can work in this mode, have not checked this)PLL inside
Also don't know, I have no experience with CS43198.(if CS chip can work in this mode, have not checked this
If it can work asynchronously, then maybe it makes sense.
Alex.
according to datasheet seems it is impossible in this case because of PLL configuring by external software contol. But I've tried to put active crystal and yes, improving of sound is audible, it is more denser and larger/deeper soundstage (tested with bookshelf speakers). Now tested with Kyocera 12Mhz, soon will put here NDK NSC5083D 24.576Mtry to cut external MCLK and put active crystal to CS43198 for use it's internal PLL
PLL plays big role in such USB transport devices.
So, controller can work with some different clocks according to datasheet, but it is configured in firmware so we can't change 12MHz here. I used Kyocera KC3225
This is small guide to swipe crystal to active crystal (it's out to XI controller pin) only for people who understand what it is and what they do. Be carefull, near glass BGA chip. Big 4R7 restistance (i just have it in such big case) makes RC filter for PLL supply pin, it can be done without it and without cutting too and even without placing 47R at crystal output (and cutting line under it between 3 pin out an XI input)
This is small guide to swipe crystal to active crystal (it's out to XI controller pin) only for people who understand what it is and what they do. Be carefull, near glass BGA chip. Big 4R7 restistance (i just have it in such big case) makes RC filter for PLL supply pin, it can be done without it and without cutting too and even without placing 47R at crystal output (and cutting line under it between 3 pin out an XI input)
Attachments
Last edited:
- Home
- Source & Line
- Digital Line Level
- FiiO K11 disassembly - modding