The BCLK or DSD CLK digital waveform looks like a fixed frequency square wave.
The left and right channel data waveforms vary with time according to the music waveform. They appear on the I2S bus lines sometime called DATA and LRCK. For one example, when a fixed frequency sine wave is played it might look something like the green trace below on a scope:
There is also a repeating pattern kind of like the above that is an 'idle' or 'zero' waveform that is used when no music is being played. IIRC an 'idle' or 'zero' pattern never has more than two short pulses in a row before a long pulse occurs. So if you see three or more of the shortest time duration positive-going pulses in a row before a longer duration positive pulse occurs, that would tend to indicate its not an idle pattern, so its probably music. Again, this is from memory so it could be wrong.
Also, if you have a digital scope it may be easier to see what the waveform looks like if you reduce the horizontal sweep frequency so you can see a lot of digital pulses on the screen. Then press the 'stop' acquisition button on the scope to show the last horizontal sweep more clearly. You may need to start acquisition again with a slightly different sweep rate if you see too many or not enough pulses after the acquisition is stopped. May take a few tries to get a good sample of pulses on the screen.
The left and right channel data waveforms vary with time according to the music waveform. They appear on the I2S bus lines sometime called DATA and LRCK. For one example, when a fixed frequency sine wave is played it might look something like the green trace below on a scope:
There is also a repeating pattern kind of like the above that is an 'idle' or 'zero' waveform that is used when no music is being played. IIRC an 'idle' or 'zero' pattern never has more than two short pulses in a row before a long pulse occurs. So if you see three or more of the shortest time duration positive-going pulses in a row before a longer duration positive pulse occurs, that would tend to indicate its not an idle pattern, so its probably music. Again, this is from memory so it could be wrong.
Also, if you have a digital scope it may be easier to see what the waveform looks like if you reduce the horizontal sweep frequency so you can see a lot of digital pulses on the screen. Then press the 'stop' acquisition button on the scope to show the last horizontal sweep more clearly. You may need to start acquisition again with a slightly different sweep rate if you see too many or not enough pulses after the acquisition is stopped. May take a few tries to get a good sample of pulses on the screen.
Last edited:
Well then, maybe you should care? If think you are playing music and what your scope shows is a DSD idle signal then you may as well have no signal at all, because you won't have any sound. The idle signal is only to keep the output the DAC at zero volts rather than hitting the negative rail. It means the scope would be indicating you have a problem before the DAC. It can happen 
For most purposes I2SoverUSB may be favored over Amanero. It features galvanic isolation, and NDK SDA clocks. http://jlsounds.com/#:~:text=I2SoverUSB v.,up to 32 bit resolution.
The Amanero has recently been updated & now also offers S/PDIF in and out.
I have tried all the other boards & still find Amanero the best.
Dom provides excellent support, build quality is top notch & the firmware tools make it very flexible.
I use the Amanero board in DAC's I manufacture & find them extremely reliable.
For the record I have no financial connection with Amanero just a very happy customer
I have tried all the other boards & still find Amanero the best.
Dom provides excellent support, build quality is top notch & the firmware tools make it very flexible.
I use the Amanero board in DAC's I manufacture & find them extremely reliable.
For the record I have no financial connection with Amanero just a very happy customer
Have also used Amanero with dedicated 5v power instead of USB power. The mod helps a lot. Thank you @clay-speakers for pointing that out
However, I still find that galvanic isolation makes an audible difference when using large panel ESL speakers.
To illustrate, not long ago we had a visitor who summarized later as follows: "Listening to your systems, the main thing I notice is the speakers. They are extremely low distortion, resulting in a lot of transparency and ability to hear down into the preceding components. Also, the ability to play really loud and clean is astonishing..."
For me what it means is, just gotta be very picky about all components in the system. It can be a pain at times, but the system is used to nitpick over all sorts of little audible issues. Could be most people wouldn't care about most of it. Just kinda what we do around here, is all.
However, I still find that galvanic isolation makes an audible difference when using large panel ESL speakers.
To illustrate, not long ago we had a visitor who summarized later as follows: "Listening to your systems, the main thing I notice is the speakers. They are extremely low distortion, resulting in a lot of transparency and ability to hear down into the preceding components. Also, the ability to play really loud and clean is astonishing..."
For me what it means is, just gotta be very picky about all components in the system. It can be a pain at times, but the system is used to nitpick over all sorts of little audible issues. Could be most people wouldn't care about most of it. Just kinda what we do around here, is all.
Last edited:
Good afternoon I am sorry for my English.
Couldn't get past your post. For many years, like you, I listened to only one frequency grid 44.1. For many years, since the purchase of Amanero boards, 48 did not work at all.
I specifically found the order confirmation email and specifically found the gift of purchase. Just like you, I bought 2 boards from the author through a group purchase on the forum in November December 2013.
As I said earlier, 48 kHz does not work in both. Both are green on thin textolite.
Just last night, I decided to unsolder the crystal oscillator. But before warming up with a soldering station (I even applied a flux), I decided to check the contacts of the quartz oscillators. The oscillators have small contact pads on the side walls. I was able to connect to them with a multimeter with needles.
There was no power plus on the 24 kHz oscillator. I did not lift it from the board, but took a very thin core of the occasion and soldered it to the site and the element next to it - by analogy with a 22 kHz oscillator.
And everything worked as expected!
I finished everything at about 2 o'clock in the morning and listened to music until 4 )) And then I finally saw a message. And I can't call it anything other than mysticism.
The second board is not next to me now. But I think there is the same defect.
I will attach a couple of photos, maybe not very good ones. But I think you understand what I'm talking about.
Please let me know how it all worked out for you.
Attachments
Hi all. I really wanted to ask for advice about the firmware. The DAC lay idle for a long time and I stopped keeping track of the firmware versions and what's new. Advise what version to flash the converter? DAC I have is not very new on the AD1853. can only PCM. if some of the firmware can convert DSD to pcm it would not be bad but not necessary. I listen mainly to Tidal.
Hello to all Amanero fans.
Question: has anyone managed to get Amanero to work correctly in the (RJ24, Slave) format? According to the
Domenico(Amanero) post for this mode you need CPU = "firmware_1096d_RJ" and at configuration
tab set field "Bit offsee"=7. However, checking with CLPD_1080 (similar to slave_1080) with CPU firmware
"firmware_1096d_RJ" and active configuration fields: Slave Mode, Stream format= Right Justified and Word
Length=24bit when setting "Bit offset"=7 then at the Amanero DATA output in both channels the digital data is
completely distorted and differently in the channels! If you specify "Bit offset"=8, then the original data is indeed shifted to the
right in each channel , but it seems that this shift is larger than necessary for 24-bit data.
P.S. there seems to be a connection with JonnyQuest's previous upper post
Question: has anyone managed to get Amanero to work correctly in the (RJ24, Slave) format? According to the
Domenico(Amanero) post for this mode you need CPU = "firmware_1096d_RJ" and at configuration
tab set field "Bit offsee"=7. However, checking with CLPD_1080 (similar to slave_1080) with CPU firmware
"firmware_1096d_RJ" and active configuration fields: Slave Mode, Stream format= Right Justified and Word
Length=24bit when setting "Bit offset"=7 then at the Amanero DATA output in both channels the digital data is
completely distorted and differently in the channels! If you specify "Bit offset"=8, then the original data is indeed shifted to the
right in each channel , but it seems that this shift is larger than necessary for 24-bit data.
P.S. there seems to be a connection with JonnyQuest's previous upper post
This link will work if the above one doesn't (post #2198) - https://www.diyaudio.com/community/threads/usb-to-i2s-384khz-dsd-converter.216474/post-4849059
Hi.
I have got this not genuine board, i would like to powered from better supply.
xmos xu208
The problem is the pc don't always recognize it I tried all solution. When the usb supplies it, there is no problem. Any idea?
Thank you.
I have got this not genuine board, i would like to powered from better supply.
xmos xu208
The problem is the pc don't always recognize it I tried all solution. When the usb supplies it, there is no problem. Any idea?
Thank you.
There is some handshaking that goes on when a USB device is first connected to USB bus. IIUC that protocol may be initiated by the USB board when it senses power from the USB bus. If running from an external power supply, a signal that USB power is present may need to be detected by the USB board, even if the USB board only uses that signal for handshaking purposes, not for power.
This is exactly what is happening. The XMOS hardware design guide is pretty explicit about it. You can power the XMOS by an outboard power supply, but you need to connect the 5V coming through the USB interface to the XMOS (through some circuitry to be exact) in order for everything to be running smoothly. I have implemented my XMOS design exactly in this way and had no problems with enumeration.