Hi Mark, I try an explanation..
Evenharmonics appearance by itself and what he had posted. Somehow I was hearing a whispered phrase with the word "doubleblind" in it and I thought, aha, here he is again. That made me smile and I felt the impulse to post it. Not sure why. I need to add, I haven't missed the phrase at all.
Cheers,
Ernst
@adam1016
Please don't let this thread get derailed. I think its really interesting and informative.
Here is a recent example
RPi4 + IANCANADA ES9038Q2M double-mono DAC + IANCANADA FiFoPi Ultimate questions
Please don't let this thread get derailed. I think its really interesting and informative.
Here is a recent example
RPi4 + IANCANADA ES9038Q2M double-mono DAC + IANCANADA FiFoPi Ultimate questions
Randy, thanks for that link -- This thread is about designing a DAC, want to keep it OT. I'm beginning to flesh out the schematics, and have decided to go with one of the TI digital power sequencing chips to handle all the different regulators and power-on stages. Planning on taking +-15V DC, with most of the load on the positive rail. Thinking I'd like to avoid using a DC-DC converter to get 5V, but this seems like a pretty big voltage drop to an LDO. Also, if I wanted to get 3.3, would you guys recommend taking the regulator input from the 15V rail (more dissipation in the regulator) or from the 5V line (extra strain on the 5V regulator) from a noise perspective? Also, I've decided to go with the AK4118 for handling my input, if I also want a USB input where and how would I put the switch? Do either the AK4499 or the AK4118 have functionality build in that would let me do this? Obviously need a I2SoverUSB or something similar, but then how do I switch the input streams?
My preference for pre-regulation before the low voltage supplies is to set up an intermediate level power bus around 7v-8v. Reason for not going down to 5v is that some regulators have better PSRR and (I think) better performance in other respects if running with a little more voltage drop across them.
The old fetish of required super regulators for the DAC boards, what a pile of.
I've shown in another thread 24bit -130dB distortion and -140dB noise floor using AK4499 and stock low noise regulator ICs for the 5V reference voltage and here we go. I'm sure there's something immeasurable that affects the sound, and a golden ear can easily hear another -10dB noise floor, ridiculous. BTW, the reference board uses AD817 opamp discrete regulators, hardly a high end solution.
This BS, as much as the ultra low phase noise oscillators BS, will never stop.
P.S. AK4499 chips are back at DigiKey. I got a couple myself.
Last edited:
I like the 7805 regulators they use, wouldn't change them. They are for the analog output section where a high-performance LDO would be a bad choice.
There is a lot of BS in audio, but not everybody who says things you doubt is wrong. There are a few people who are described as having good ears, but actually have very normal ears. Those few people have learned what to listen for, that's all. You can scoff, won't change anything.
No doubt we are going to disagree on some things. Jitter may be one. If there is audible jitter and your measurements look great, it just means you are using the wrong tests that don't show the problem, nothing more. I'm sure there are other tests available or someone could invent new ones. If its audible its measurable. That assumes experienced listeners are being used, not neurotic audiophiles (about 1/3 of them).
There is a lot of BS in audio, but not everybody who says things you doubt is wrong. There are a few people who are described as having good ears, but actually have very normal ears. Those few people have learned what to listen for, that's all. You can scoff, won't change anything.
No doubt we are going to disagree on some things. Jitter may be one. If there is audible jitter and your measurements look great, it just means you are using the wrong tests that don't show the problem, nothing more. I'm sure there are other tests available or someone could invent new ones. If its audible its measurable. That assumes experienced listeners are being used, not neurotic audiophiles (about 1/3 of them).
Money is a big motivator. As long as there's money to be made, it will continue on.
None of them are immune to personal bias just like the rest of us and none of them have aural memory span more than a few seconds just like the rest of us.
In this day and age, there are DACs sold with audible jitter when playing music? Hard to believe.
Adam,
You can probably see what Randy was talking about based on the last few posts. It could be that the best thing to do would be to take this topic to another venue where hecklers and trolls are under more constraints to behave.
By the way, evenharmonics is on my ignore list. He will have to stop trolling, stop trying to bait people into his debates, and stop the 'Sea-Lioning' tactics if he wants to get off the ignore list (probably not worth it to him).
If anyone actually has serious questions I would probably be willing to respond.
Regarding input switching, probably best to use gold plated pin headers and jumpers, but if you want automated switching then some high speed CMOS multiplexers in very small SMD packages are probably worth looking at. Some people use CPLDs, but I'm not convinced I like them.
Which multiplexers might be best would depend in part on how many inputs you want to be able to select from.
Also, an important question would be if you want to switch the master clock source. Probably better to use one set of master clocks if possible. Therefore, I would use AK4137 with AK4118 for any SPDIF family inputs. AK4137 can perform ASRC on incoming audio streams to sync them to the local master clock. Since I have used that part before, I can probably help with any questions about how to use it.
On the other hand, if you want to use jumpers and are careful about design then I would say changing master clock sources might be less worrisome.
The thing is, probably best to avoid doing things that will degrade master clock signal quality. Running through unnecessary logic devices would be something I would prefer to avoid if possible.
Hopefully, you understand that layout and sufficient PCB layers are critical for best results in a project of this type. Suggest you read Henry Ott's book, Electromagnetic Compatibility Engineering, if you haven't done so already.
You can probably see what Randy was talking about based on the last few posts. It could be that the best thing to do would be to take this topic to another venue where hecklers and trolls are under more constraints to behave.
By the way, evenharmonics is on my ignore list. He will have to stop trolling, stop trying to bait people into his debates, and stop the 'Sea-Lioning' tactics if he wants to get off the ignore list (probably not worth it to him).
If anyone actually has serious questions I would probably be willing to respond.
Regarding input switching, probably best to use gold plated pin headers and jumpers, but if you want automated switching then some high speed CMOS multiplexers in very small SMD packages are probably worth looking at. Some people use CPLDs, but I'm not convinced I like them.
Which multiplexers might be best would depend in part on how many inputs you want to be able to select from.
Also, an important question would be if you want to switch the master clock source. Probably better to use one set of master clocks if possible. Therefore, I would use AK4137 with AK4118 for any SPDIF family inputs. AK4137 can perform ASRC on incoming audio streams to sync them to the local master clock. Since I have used that part before, I can probably help with any questions about how to use it.
On the other hand, if you want to use jumpers and are careful about design then I would say changing master clock sources might be less worrisome.
The thing is, probably best to avoid doing things that will degrade master clock signal quality. Running through unnecessary logic devices would be something I would prefer to avoid if possible.
Hopefully, you understand that layout and sufficient PCB layers are critical for best results in a project of this type. Suggest you read Henry Ott's book, Electromagnetic Compatibility Engineering, if you haven't done so already.
Last edited:
Money is a big motivator. As long as there's money to be made, it will continue on.
In this day and age, there are DACs sold with audible jitter when playing music? Hard to believe.
This is worth a read. I believe all tests were done blind as is Bob's usual
practice.
Grimm Audio: CC1 master clock | Tape Op Magazine | Longform candid interviews with music producers and audio engineers covering mixing, mastering, recording and music production.
T
Exactly right.
Yes, I already pointed in another thread out that AD817 is a crappy and relatively noisy choice for the error amp, but Mark thinks it's been blessed by the golden hand of AKM's resident true-believer.
It's all complete nonsense. He's setting up his commercial venture quite well probably. You know, insist that you can only do it one way, and then release a product that does it exactly that way.
On January 11, 2019 on "quality of new threads going downhill" thread, you responded to me, "You have been on my ignore list before as I recall. Welcome back, and good day.". You've read and responded to my posts number of times since then. You know what happens to you when you cry "wolf" too many times.
Adam, what you really have to be aware of on audio forums is that there are bunch of members posting with business interests. IOW, they are shills. Typical sales pitches are DBT is not accurate, not all audible traits of audio gear are measurable, trust your ears (while their sight is also involved during listening).
The transient performance may have been their reason.
The noise is better than many LDO's, but it is common mode to both + and -
DAC OP's, so will largely be cancelled.
It may also have been chosen purely for subjective reasons, whether you like
it or not, AKM (and other DAC manufacturers) a/ actually do listening tests
b/ consult with certain users WRT subjective properties of their DACs.
TCD
Markw4 does listening tests too. Otherwise how would he know which DAC sounds certain way as he claimed? The important question is how that test was carried out.
It doesn't matter. You will not believe him.
Just listen yourself, and you will find all sound same
AD817 happens to appear quite often in Jung regulators where people were unable to get AD797 stable. It seems to me that the designer picked it out of convenience and as a known quantity, considering the datasheet was last updated in 1995 and there are better alternatives available.
I'm sure AKM has someone who listens or claims to listen. I am sure ESS has someone, too. Cirrus? Maybe... TI and AD? Probably not in an official capacity. If a large portion of my business was selling audio ICs, I'd probably have some golden ear story, too. Do you believe their Velvet Sound filter marketing? The Japanese can be quite eccentric in their designs as well; for a non audio example just look at the PS4's Cell processor. An incredibly flawed design pushed by ego and some kind of perceived need to be different from the competition.
We will never agree on this so it's fine to leave it be, but the notion that a top quality IC linear regulator is not good enough for the typical op-amps used on these boards is just laughable from a technical perspective.
I really do not care what anyone thinks they hear without objective evidence. We have members here who think they have some kind of putty that can change the taste of wine, and that listening to audio passed through leaded solder is harmful. I have no time for this BS.
Last edited:
actually no regs at all is fine sometimes. providing there is enough filtering, of course.
There are people who listen to what happens when changes are made to an audio circuit and there are those who don't. Those who don't listen try to substitute measuring instead which is a mistake, IMHO. One should both listen and measure, in that order.
However, listening is a skill that has to be learned. Neurotic audiophiles haven't learned it. Anyone who wants to is welcome to visit with us here in Northern California and see for yourself. We can teach you a lot about what to listen for, and what equipment makes it easier to do. The skeptics are welcome too.
Enough said on that.
If the S/N ratio around here gets any worse I don't need to be here. This is a hobby for me and not one for which I am willing to take a lot of abuse.
However, listening is a skill that has to be learned. Neurotic audiophiles haven't learned it. Anyone who wants to is welcome to visit with us here in Northern California and see for yourself. We can teach you a lot about what to listen for, and what equipment makes it easier to do. The skeptics are welcome too.
Enough said on that.
If the S/N ratio around here gets any worse I don't need to be here. This is a hobby for me and not one for which I am willing to take a lot of abuse.
Last edited: