As rfbrw mentioned, you only need one clock. Everything is converted to the destination clock domain rate.
A question I would have is if the AK4137 is the best choice anyway, considering there is zero data published on its DPLL filter compared to SRC4192/4392. The datasheet is severely lacking compared to what you get from AD, Cirrus, or TI. I’m sure it’s audibly transparent, though.
You don’t even need an ASRC if you are only going to use USB input.
Yes, PLLs are known for sounding bad. They muddle the bass, add sibilance to the female voices and affect the transients, like cymbals. PLLs that have 200 femto seconds RMS jitter are in particular to be avoided.
Last edited:
Apparently not stale enough for some here, at least when talking about the 7805 regulators superior sound.
Quiet the chip to use as a DIY hobby project, a fair bit of coding to make it operate, but making it open source helps with collaboration, if you can get people to agree on implementation. I do wonder when folks can not even agree on a power supply
I see it is a 0.4mm pitch part, isn't low cost either, a deep pocket design, 4 layer pcb, fun wow.
Got of love them spec'ing an obsolete mute bjt, 2SC3327, makes you wonder what other nasties they have in store for you.
$908.770 $CDN for an evaluation pcb = outch, got to buy one to get a schematic?
Syn08 are you going to start your own thread or work silently?
I wold consider collaborating in a DAC design.
I see it is a 0.4mm pitch part, isn't low cost either, a deep pocket design, 4 layer pcb, fun wow.
Got of love them spec'ing an obsolete mute bjt, 2SC3327, makes you wonder what other nasties they have in store for you.
$908.770 $CDN for an evaluation pcb = outch, got to buy one to get a schematic?
Syn08 are you going to start your own thread or work silently?
I wold consider collaborating in a DAC design.
XMOS and AK4137 PLLs are good enough, but a 200 femto second jitter PLL, as master clock is not. Got it.
No matter what people try to explain to you, you always find a way to become confused.
Don't know why you have to be such a jerk any time someone says something you don't agree with. One person who does what you do can really damage a forum like this. No wonder all the good people from the old days left.
As a result of your trying to ruin this thread, no AK4499 board is going to be built around here. Not unless its some piece of junk from you that was built without ever listening to it.
Silently, for the time being.
Digital audio is for me a hobby within a hobby, it's a slow learning curve for me. One thing I already figured out, there's nothing magic (or secret sauce) about a chip like AK4499, 99.9% of it's ultimate performance is related to the board layout. Everything else, from reference voltages (the next low noise regulators will do), to the I/V op amps (OPA1612, ADA4627 or ADA4898 will all do great), to the master clock generator (a $1 quartz will do just fine) is marketing crapola. For any audiophile purposes, I would bet an AK4499 reference board that nobody could distinguish the AK4499 from it's much more affordable AK4497 in an ABX test designed to Jakob(x) satisfaction. The only place that AK4499 would truly shine is for instrumentation, an "RTX" build around this chip would be truly amazing, in particular if a matching ADC chip would be available (unfortunately, not).
Yes, a 4 layer board is absolutely required for good performance.
Does spreading audiophile myth help a forum like this?
Like jkeny, one who sells organic sounding DAC?
Listening method which you won't share.
You can try LTC2378-20 with an FPGA for interfacing and some external digital filtering. I haven’t used it, but am using LTC2387-18 in a test and measurement application not for audio and it performs spectacularly.
Your cheerleading has no value here. I’d suggest you stick to posting when you have a clue on what’s being discussed.
Argument over the capabilities of a PLL and VCO notwithstanding, the argument from Mark is probably because there is no need for an external PLL in the design unless I am missing something (like you want to try to lock to external word clock or clean up a recovered SPDIF MCLK further). The USB receiver’s I2S ports can easily be made to be in the same clock domain as the converter.
Last edited:
+1
I agree wholeheartedly with Marks input to this thread with respect to power supplies based on my experience with many different regulator types on other DACs. It would be nice if more respect could be shown to people who are just trying to get the best outcomes.
John
AK4497 seems to me to be more manageable for a first cut at a DAC project.
To bad the evaluation documentation is lacking specifics on implementation. They designed it using ORCAD
The software is grossly out of date, supporting obsolete Win O/S versions, looks like a poorly supported product imo.
AK4497 seems to me to be more manageable for a first cut at a DAC project.
To bad the evaluation documentation is lacking specifics on implementation. They designed it using ORCAD
The software is grossly out of date, supporting obsolete Win O/S versions, looks like a poorly supported product imo.
Orcad is still popular and supports Windows 10 just fine even if the UI looks dated.
AK4497 is indeed a bit easier to use.
Last edited:
I was referring to the AKM evaluation board software O/S support.
I understand that orcad capture still exists, too bad they dropped layout and only offer the Allegro/Cadence variant which imo is overly complicated for simpler designs and new pcb designers.
I used Allegro back in the 90's just after Cadence bought? was it Daisy, to long to remember now, I'd have to look
I understand that orcad capture still exists, too bad they dropped layout and only offer the Allegro/Cadence variant which imo is overly complicated for simpler designs and new pcb designers.
I used Allegro back in the 90's just after Cadence bought? was it Daisy, to long to remember now, I'd have to look
Last edited: