I see all multipurpose pins on the headers... so no problem with i2s peripherals.
Awesome.
Do you know where to start to use the MP pins for I2S output?
In the register control?
Cheers!
Same point to start as with every device: Datasheet
And yes, you need to enable the digital i2s in/out and define the pins for data and the clocks.
A somewhat less fruitful experiment, I attempted to get SPDIF input working as well using this board:
AK4118数字接收板 同轴光纤AES转I2S 采样率显示 支持I2S输入！超WM8805 - 音响HiFiDiY - 极DIY
which uses the AK4118A:
https://www.akm.com/content/dam/doc...nterface/ak4118aeq/ak4118aeq-en-datasheet.pdf
This board is configured to drive its I2S clock lines so I had to uncheck the DSP master mode. This resulted in audio output when routing channels 2/3 to the added output channels. However there is crackling distortion when high frequency is present (apart from weirdly if the SPDIF is at 44.1 kHz - wheupon there is no obvious distortion but it just didn't seem to sound right). Thinking about using this board I think I have made a mistake as there is no sample rate conversion between the SPDIF and DSP and so rates apart from 48KHz should not work and even with 48kHz there should be synchronization issues as the boards are not sharing a common clock?
AK4118数字接收板 同轴光纤AES转I2S 采样率显示 支持I2S输入！超WM8805 - 音响HiFiDiY - 极DIY
which uses the AK4118A:
https://www.akm.com/content/dam/doc...nterface/ak4118aeq/ak4118aeq-en-datasheet.pdf
This board is configured to drive its I2S clock lines so I had to uncheck the DSP master mode. This resulted in audio output when routing channels 2/3 to the added output channels. However there is crackling distortion when high frequency is present (apart from weirdly if the SPDIF is at 44.1 kHz - wheupon there is no obvious distortion but it just didn't seem to sound right). Thinking about using this board I think I have made a mistake as there is no sample rate conversion between the SPDIF and DSP and so rates apart from 48KHz should not work and even with 48kHz there should be synchronization issues as the boards are not sharing a common clock?
Attachments
ah it may actually be possible by running the 3E DSP from the clock of the AK4118A board as MCLK is 256xFS which is 12.288MHz which is the usual oscillator frequency for the ADAU1701 when FS=48KHz. So as long as 48kHz SPDIF is used it could be possible.
The 3E datasheet also mention that J2/J3 are in place during master mode but as the board is now a slave I should probably work out what they are doing.
The 3E datasheet also mention that J2/J3 are in place during master mode but as the board is now a slave I should probably work out what they are doing.
J2 connects MP11/BCLK_OUT to MP5/BCLK_IN and J3 connects MP10/LRCLK_OUT to MP4/LRCLK_IN
The I2S input port of the ADAU1701 can only work in slave mode so by applying these jumpers and operating in master mode on the output port LRCLK and BCLK are generated by the ADAU1701 for the input port and it will act as a master. In the case of the SPDIF board its an I2S master so the correct usage will be to remove these jumpers, then the ADAU1701 can act as a master for the DAC.
Messing around with the clock circuit on the ADAU1701 looks a little tricky due to the small parts....
The I2S input port of the ADAU1701 can only work in slave mode so by applying these jumpers and operating in master mode on the output port LRCLK and BCLK are generated by the ADAU1701 for the input port and it will act as a master. In the case of the SPDIF board its an I2S master so the correct usage will be to remove these jumpers, then the ADAU1701 can act as a master for the DAC.
Messing around with the clock circuit on the ADAU1701 looks a little tricky due to the small parts....
The AK4118A board outputs a 24.576 MHz clock on MCLK of its output connector. This is twice the 3E DSPs clock. Looking at the datasheet for the ADAU1701 if pins PLL_MODE0 = 1 and PLL_MODE1 =1 then a fs*512 clock will be accepted by the ADAU1701 rather than the normal fs*256. So hopefully if I remove the crystal on the DSP board, replacing it with the MCLK and change the mode pins the SPDIF board and DSP will be in sync for a 48kHz SPDIF signal.
Attachments
Kipman, thanks for all your development on the 3E Audio DSP board.
I have an HDMI to I2S board, that I could test as an additional ADC, but it still too advanced for me.
Cheers!
I tried my HDMI to I2S but it did not work. Tons of noise and the signal was hear very very very low.
I connected SDATA/DSDL to DSP MP0, BCLK/DCLK to MP5 and LRCK/DSDR to MP4. Probably I am making something wrong or this board was not meant to extract audio from HDMI.
I connected SDATA/DSDL to DSP MP0, BCLK/DCLK to MP5 and LRCK/DSDR to MP4. Probably I am making something wrong or this board was not meant to extract audio from HDMI.
Attachments
You configured the dsp to run in slave mode (digital inputs always run in slave mode). Are the clocks synced in the system? Which module is the master?
Easiest way to connect input peripherals to the dsp over i2s is to set the dsp to be the master and loopback the output clocks to the input (see datasheet of adau1701).
Both the ADAU1701 need to to be running from the same clock but the clock on the HDMI inpput board is 24MHz (I think from reading the top). However if you read the ADAU1701 datasheet the only acceptable clocks are: 64*FS, 256*FS, 384*Fs and 512*Fs. So its not posible to run the ADAU1701 from the same clock as the HDMI board unless the HDMI has another clock that is phase locked with its sample clock that is acceptable for the ADAU1701. (assuming you wanted to clock the ADAU1701 from the HDMI board).
I can tell you the ADAU1701 green board from Aliexpress is quite silent. I tested it with a CD which is around 105db sensitive and with a JBL 2226H which i think its around 95db.
The 3E audio I have is runnning my SW so i haven't tested it with a CD. I bought another 3E that should be arriving withing the next days, I can report if you want.
Right now I am testing an ADAU 1452 from Aliexpress. As soon as I manage to learn how to use it, I can report.
From all the tested, miniDSP 2x4 Balanced is by far the noisiest, but it's probably because I haven't configured correctly the gains.
I would certainly appreciate you letting me know after you’ve had to chance to get those going. I’ve got a project that’s on hold right now because I just don’t know which board to get. I sit close (8 feet) and listen at low volumes so the SEOS based speakers I’m building certainly won’t be forgiving of a noisy DSP. And for reasons I probably don’t need to get into I need to have the DSP between the pre amp and the power amp. Thanks mga2009!
I just tested the ADAU1452 that I've got from Aliexpress, with my compression driver (Denovo DE150) and seems very silent. Now, consider I am not an expert and I am comparing the boards very subjectively, just putting my ear very close to the compression driver (inside the horn). The compression driver is conntected directly to my amplifier (ICEpower 50ASX2SE and Anaview Abletec 110-2300). As source I am using my laptop's 3.5mm output jack, with Windows 10 volume to the top, amplifier without volume knob, and feeding the DSP with 5 volts from my laptops USB.
If you want some more technical answer, and I can help you with that, just tell me. I am willing to learn and provide you with the results.
Cheers!
Last edited:
That's great to know! The ADAU1452 looks nice. Could you suggest which one to buy? Here's a few.
ADAU1452_DSP Development Board, Learning Board (STM32F103) Has a Schematic Diagram|Air Conditioner Parts| - AliExpress
ADAU1452 / 1466 Development Kit, USBi Plus 1452/1466 Development Board|Personal Care Appliance Parts| - AliExpress
ADAU1452 Development Board / 78 Routines / Full Chinese Video Tutorial / EVAL ADAU1452REVBZ|Fingerprint Recognition Device| - AliExpress
ADAU1701 development board (new version) /ADAU1401/ADAU1702/ADAU1761 development board|board board|board development - AliExpress
Or if someone knows their stuff and wants to put together a nice unit for me you can reply here or send me a PM and we can talk.
Thanks!
Shawn
ADAU1452_DSP Development Board, Learning Board (STM32F103) Has a Schematic Diagram|Air Conditioner Parts| - AliExpress
ADAU1452 / 1466 Development Kit, USBi Plus 1452/1466 Development Board|Personal Care Appliance Parts| - AliExpress
ADAU1452 Development Board / 78 Routines / Full Chinese Video Tutorial / EVAL ADAU1452REVBZ|Fingerprint Recognition Device| - AliExpress
ADAU1701 development board (new version) /ADAU1401/ADAU1702/ADAU1761 development board|board board|board development - AliExpress
Or if someone knows their stuff and wants to put together a nice unit for me you can reply here or send me a PM and we can talk.
Thanks!
Shawn
Last edited:
The biggest improvement will be to attenuate the DAC output as the noise floor is approximately constant. When you put the attenuation in the noise floor is reduced as well as the signal but then you will increase the output level of the DAC increasing the SNR.
I really don't know which one to recommend. I can tell you that in the links there is an ADAU1701 which I would not recommend at that price.
I bought this one:
Lusya ADAU1452_DSP development board learning board DSP board with CODEC F9 006|Replacement Parts & Accessories| - AliExpress
And there is a cheaper one, which does not come with a microcontroller (IDK the implications of this):
https://www.aliexpress.com/item/33027816077.html?spm=a2g0s.8937460.0.0.816c2e0ehNgNsg
Please consider that you need a FreeUSBi or similar device to program it.
Kipman, thanks for your recommendation. So, If I understand correctly, the best way to lower noise is to reduce the gain, right on the DAC's (DSP in this case?) outputs?
I'm sorry but I'm confused, not because of your explanation but my lack of experience. Are these two separate things, to attenuate the DAC output, which attenuates the noise floor (good) but also the output signal (not necessarily good) and to increase the output level of the DAC which increases SNR (good)?
The very best thing to do would be to reduce the gain of the power amplifiers however this often has a minimum due to stability considerations. The next best thing would be to attenuate the signal between the power amp and the DAC.