“Final Q: why???
OPs are not aligned.😢”
Would this be a complaint about a freely shared design?
OPs are not aligned.😢”
Would this be a complaint about a freely shared design?
Last edited:
I'm not complaining. I am grateful. I have a stereo OP board on 2 mono OP for AD1865 and I can't use it due to non-alignment. No problem, Nixi will make another one.
The uf-l pads go through the register circuitry. For a direct I2S you will need to strap the register pads or hack by injecting the I2S signals after them on the pcb. If wires are far from eachothers for that reason you migth find a very close gnd return path for each of the three wires... So 6 wires.
About PCM56 it is not only about thd or snr...there is something that sounds rigth with him. However few have made iso listening test at iso layout. Andrea Ciuffoli prefered the ad1865 over the tda1541A or the other pcm.
But the few that prefered here the pcm63 over the ad1862, it would be fun to have more inputs from whom having tied the pcm52 and not only the ad1865. I just remember one input IIRC.
About PCM56 it is not only about thd or snr...there is something that sounds rigth with him. However few have made iso listening test at iso layout. Andrea Ciuffoli prefered the ad1865 over the tda1541A or the other pcm.
But the few that prefered here the pcm63 over the ad1862, it would be fun to have more inputs from whom having tied the pcm52 and not only the ad1865. I just remember one input IIRC.
He has only 64 pcs left. He is selling it from 2007 and my thread will speed up his sale 🤣
Use the I2S input header pins:
2 - LRCK (LR_CLK_I2S - 15)
4 - BCK (BCLK_I2S - 11)
1 or 3 - GND (10, 12 ...)
... if shift registers are not installed you don’t need to cut the JP1 and JP2
DL - Data Left (DATA_I2S - 13)
DR - Data Right (SPDIF - 9)
... you can use non-isolated (manual page 3: 1. Bus‐powered option) or isolated (manual page 3: 2. Bus-powered USB side) configuration in jlsounds;
... H1.3 pin must be connected to H1.1 through external R1 4.7kΩ configuration resistance;
... configuration table on the manual page 6 (20-bit PCM63, J2:Open J3:Open J4:Close, B1-5:Open);
OpAmps are not aligned from the first version (it was by mistake) and I did not correct it even for AD1865, all my DACs are the same (your stereo OP board should be compatible with all), something like a signature LOOL 🤣🤣
I was made one dac with PCM56 long time ago... Cant remeber was it with PMD100 in os mode or with CS8414 in nos? Riv followed with some diskrete OP adopted from Hafler DH110 preamp section. BC327/337, Darlington outpu... Anyway, I was suprised how it was good. That I will remember about These PCM56. Very specific and simply good, direct, rich sound.
The most true is 2... But that is for transsmition line type not the dac. Older studio models had ballanced line but in the output, with transformer.
.
4. should be mesured with stopped clock and inverted only data, but whole line but only data segment - everything else should stay on "0" and signal integrity with ground bounces should be threated.
If it not You will have double amount of noise...
.
3. in most cases higher output signal is counter-productive. Output levels are alredy way too high even for the direct input on amplifiers which getting full power, at half or less then half of DACs output P-P.
Heres another illustrative data point for NOS users interested in the issue of US images folding back into the audible mid-range via a R2R ladder + analogue stage IMD noise.
This is a torture test with full level 20KHz tone into a discrete resistor ladder DAC from Gustard. The IMD's are multiples of ~4kHz ( 44kHz sampling?) with NOS in blue and OS in red. OS deceases IMD by as much as 60dB and pushes the noise floor down 20dB.
Would any DIY NOS DAC depending on the IV and analogue IMD performance perhaps sound better with software OS done right?
These PCM56 looks like a straight ladder DAC:
Will glitch and bipolar noise be an issue compared to later designs?
Would a double stack differential design mitigate the low bit depth and glitch/bipolar issues?
If not faked, perhaps affordable PCM56 eg 10 for ten bucks might be ideal for a double stacked differential design using eight chips with two spare chips ?
This looks to me like an analog stage in origin IMD spur:
1) 44.1kHz sample rate - 20kHz test tone = 24.1kHz image tone
2) 24.1kHz image tone - 20kHz test tone = 4.1kHz IMD spur
Then launching a series of analog stage generated distortion harmonics, with that 4.1kHz IMD spur acting as the fundamental. At least, that‘s what it appears is happening to me without full details of the test procedure. I know of no digital sampling phenomena which folds DAC image-band frequencies to below Nyqvist, down to the audio base-band. A related phenomena does occur via aliasing, where ADC input frequencies above Nyqvist do fold down in to the audio base-band.
Last edited:
I can't think of a mechanism whereby simply going to OS would reduce the noise floor by 20dB. OS with noise shaping would though. Or perhaps the change in noise is as a result of the AES17 brickwall filter being used in the NOS case - is it an opamp-based active filter?
Thanks! Thats what I was trying to say. Sorry for my poor choice of words. I dont have your technical expertise. Thats why I said it was an illustration - like a painting a picture. I thought the only practical way the audio band could be distorted was from IMD? Ignorance is bliss.
I read that could happen but thought it was sampling theory 101 with no practical implications and dismissed it? Does this ever happen? Are there any other ways ultrasonic images can theoretically "fold" (is that a precise technical term?) down into the audible band purely in the digital domain creating contaminated data in a file before it even goes off to a DAC?
Yep. Thats what I assumed but the measurement site is a bit light on test conditions. They use an AP though. I linked the page if your curious.
Looks like the Gustard has embedded software OS-filter with "PCM filters" but you would know better about the output stage then me sorry.
Sorry. Im not an EE. I only know when OS is done right for me I like it but there are plenty of NOS lovers who have tried OS and don't like it.
No worries, nor any place for apologies. I feel that we all are here to on occasion be teacher, and to on occasion be student. 🤔 🙂
“Fold”, while not technical, I see as an acceptable usage. Anyway, yes, digital sampling folding phenomena does occur, but the nature of folding depends on whether we are converting an analog signal to a digital signal, or converting a digital signal to an analog signal. A sampled system’s Nyquist frequency (which is always 1/2 the sample rate) serves as a sort of pivot point for the folding behavior.
Avoiding all of the math - with ADCs, the folding phenomena relocates analog signal input frequencies which wrongly happen to be HIGHER than the Nyquist frequency, to become falsely encoded as a digital signal that’s LOWER than Nyquist frequency. For CD, that means it’s folded somewhere below 22.05kHz. Down in to the audio band, so, it’s audible. Such folded down frequencies have a generic name, which is, ‘aliases’. Again, avoiding all of the math - DACs, on the other hand, produce repeating copies of the digital audio signal frequencies which can only manifest HIGHER than the Nyquist frequency. Such folded DAC frequencies have the generic name of, ‘images’. For CD sample rate and higher, all DAC image frequencies are ultrasonic, and so cannot be heard except maybe by the very young.
The main point above is that, ADCs only fold aliased frequencies down to LOWER than the system Nyquist frequency, while DACs only fold image frequencies up to HIGHER than the system Nyquist frequency.
While I don’t know what form of OS the Gustard utilizes, that chart shows that OS greatly reduces it’s analog stage distortion through effective suppression of the (44.1 -20) = 24.1kHz image frequency. With a suppressed 24.1kHz image frequency, there is a suppressed (24.1 - 20) = 4.1kHz IMD spur. With a suppressed 4.1kHz IMD spur, there is a suppressed series of resulting distortion harmonics.
Last edited:
I already went there and found it was in Chinese so used Yandex to translate. I'm guessing its the same guy who posts measurement threads on ASR.
The AES17 is a filter applied to facilitate measurements of kit putting out high levels of ultrasonics, nothing much to do with the DUT.
I have had some time living with my copy of a Miro AD1862 DAC and have a question for Miro DAC owners that have tried filters on the output. In my copy that has no filter I find the sound to be very clean, detailed, and natural with a really nice sound stage. If you have implemented a filter stage :
1) How/what kind of filter did you implement ?
2) How do you feel it changed the sound/presentation ?
Thanks
1) How/what kind of filter did you implement ?
2) How do you feel it changed the sound/presentation ?
Thanks
My Miro PCM63 version is not working. Whether the problem is the direct connection to JLSound or the defective PCM63 chips, I would not know. I checked all the voltages and everything is OK. When I put PCM63 I get DC at the output. About 0.7V on pins 2/3 on the base for the OP.
I was hoping to listen to music for the weekend.
Back to AD1865.
I was hoping to listen to music for the weekend.
Back to AD1865.
@skrstic tbh, that doesn't look anything like the pcm63p-k chip here. And Mika Y version looks different.
https://www.diyaudio.com/community/threads/real-or-fake-pcm63.121552/#post-1488201
I'll be receiving originals from ruach in a few weeks, and i can take a pic as well.
https://www.diyaudio.com/community/threads/real-or-fake-pcm63.121552/#post-1488201
I'll be receiving originals from ruach in a few weeks, and i can take a pic as well.
I hope your PCM63 are OK? It would be good to know where you got the PCM63 from? The top surface in your photo looks very smooth.
Is it as smooth on top as on the circular depression when you look at it under side lighting?
Or is the top rougher like below?