It's unclear from the User Guide, but it means >=150mA per channel (>=300mA in total). I'll chage the User Guide.
I have Advanced/Upgrated Amanero with Crystek CCHD-957. My opinion - it is noticeably better than the regular version. It has better clocks, 7 LDO (instead of one in standard Amanero) and the ability to power it from external power sullpy (not from USB 5V). I've not tested the last option (powering from external 5V).
No need in R130 and C9 for stock transformers. It's a snubber surcuit to reduce oscillations, just for fine tuning. If you decide to use LL1684 - datasheet is recommended 1K + 3nF. No recommendations for LL1527XL.
R65: stock transformers have the best square wave results (according to Pavel's measurements) when load to ~ 5K-7K. Load = R65 || input impendance of your amp. For example, my amp has 100K input impendance, if R65 = 6K, the load will be 5,66K (1/Rload = 1/100K+1/6K). In most cases you can use in R65 any value between 6-7K. As you can see, it's depends on your amp input impendance.
C (schematics for 1527XL) - no need for stock transformers (see recommended shematics in User Guide). It's only for LL1684 and 1527XL, try C 0,1uF - 1uF in addition to datasheet's schematics, you may prefer it (or not).
Pavel uses XLR and prefer the following configuration (as I know): no ferrit bead (or shorted by wire), J8 - open, J13 - open, J10 shorted (XLR). It's fully isolated output.
I use RCA and have better measurements with - no ferrit bead (shorted by wire), J8 - shorted, J13 - shorted, J10 - open (no need in XLR).
Pavel used this ferrit bead and then decided just to short it by wire.
There are some recommendations in datasheets for 1684 and 1527XL. I can add some in User Guide, but it's very subjective.
For example in datasheet for LL1684 is recommended 10K||1K + 3nF. I prefer 10K as a load and C = 0,1uF Wima FKP (see schematics for LL1527 XL in User Guide, last page).
Last edited:
I'm building this project with all in one.
PSU using the Talema with LT3045 LDO for 4 PS Different.
Amanero isolated by ISO7640.
R FIR can use the Dip or Melf MMA0204.
OPT can use Lundahl, Edcor, Permalloy taobao or Jensen and Cinemag.
Layout match for double two floor to trie by dual jump 2.54mm.
I hope I will received received good result.
Top Viewer:
1. TOP (Gerber View).pdf - Google Drive
Bottom View:
2. BOT (Gerber View).pdf - Google Drive
PSU using the Talema with LT3045 LDO for 4 PS Different.
Amanero isolated by ISO7640.
R FIR can use the Dip or Melf MMA0204.
OPT can use Lundahl, Edcor, Permalloy taobao or Jensen and Cinemag.
Layout match for double two floor to trie by dual jump 2.54mm.
I hope I will received received good result.
Top Viewer:
1. TOP (Gerber View).pdf - Google Drive
Bottom View:
2. BOT (Gerber View).pdf - Google Drive
Attachments
Last edited:
I saw a few people talking about memory speed being important for this project and dsd. I just learned the other day that formatting memory sticks and hard drives with the largest block size increases read and write speeds and after trying it (64kb block size instead of standard 4kb) it has a significant effect. Minimum file size is 64kb however, or whatever block size you use.
Is it possible to I2S signal straight into DSC (Pi, Minnowboard, etc) or will I still need Amanero or XMOS usb board? Goal would be to setup a network only dac.
DSC DAC works with a DSD stream only. Therefore, the I2S stream needs to be converted to a DSD format. To do this, you can use converters based on AK4137, like this AK4137 | eBay
AK4137 will make it possible to achieve your goal very simply. But AK4137 does not allow you to get the highest quality sound that can be obtained by using HQP or other software players with built-in PCM conversion to DSD stream.
For more details you can read the DSC 2.6 review:
PureDSD
This is my target goal from ages (unfortunately I'm an electronic goat so unable to implement it) ... 😱
Few pages ahead there's a discussion about the usage of a BeagleBoneBlack as a NAA directly connected to DAC via I2S. I already experienced HQP and NAA setup ( via USB connection not direct I2S) but to be honest I consider it too complicated, my idea is for a single unit as a playout device. BBB/Raspy could be an option but they are too limited in terms of processing power.
My dream solution should be a mini-ATX board Pc (i3/i5) with an I2S PCI board (something like PinkFaun bridge), the crytical point it's a well implemented mute function to avoid pop/clicks (I already destroyed 6,000$ speakers and I would like to avoid a second run).
I plan to leverage HQPlayer to convert pcm/redbook to DSD and send it along to NAA hooked up to the DAC.
Yeah, I have the same issue with not being good with DIY
.The other option that I have been thinking about is Minnowboard which is Intel based SOC. They are bit more expensive but according to the compare portion it does support I2S.
It actually doesn’t matter which of the R1-7 are soldered: you can use the jumpers on the PCB to route the mute signal to pin 11 (J4-9).
I also accounted routing the DSDEN (pin 7), but I think it’s obsolete - I simply wound up using the JP1 to +3V3.
I was think something like this sits on DSC2:
u.fl for dsd1 dsd2 dsd_clk
a relay for the dsd_on
Last edited:
A mechanical relay does not have enough speed for mute. And you forgot about the very unpleasant effect - the bounce contacts. That's why I use electronic switch.
Last edited:
Today I've tested new Amanero firmware. CPLD_for_1080, firmware_2004be, oemtool118.
DSD512 artifacts is still here 🙂 (strong distortion in 5-15 minutes from the beginning of playback).
Yes, the firmware had been reworked - for 2003be the distortions did not disappear until the playback stopped. On 2004be they pass themselves in 3-5 seconds. During 30 minutes I've noticed ~ 4-5 artifacts. Ran back to 2003be_71A.
I hope this is not the final version of 2004be.
DSD512 artifacts is still here 🙂 (strong distortion in 5-15 minutes from the beginning of playback).
Yes, the firmware had been reworked - for 2003be the distortions did not disappear until the playback stopped. On 2004be they pass themselves in 3-5 seconds. During 30 minutes I've noticed ~ 4-5 artifacts. Ran back to 2003be_71A.
I hope this is not the final version of 2004be.
Last edited:
Hi Vit
What format does the Amanero expect? Is it raw DSD which it just passes through, or does it convert DoP to raw DSD?
If it's raw DSD, what format is expected? U32LE or U32BE?
Thanks
What format does the Amanero expect? Is it raw DSD which it just passes through, or does it convert DoP to raw DSD?
If it's raw DSD, what format is expected? U32LE or U32BE?
Thanks
Download oemtool117 from
Amanero Technologies
and unzip it.
Read README.pdf with how to flash info. Among others, 2004be is available in "CPU Firmware Selection" of oemtool.
Amanero Technologies
and unzip it.
Read README.pdf with how to flash info. Among others, 2004be is available in "CPU Firmware Selection" of oemtool.
Last edited:
[QUOTE = wengtech; 5453011] hello,
anyone tried with NXP 74HC595D ??
what is the different with 74AHCT595D? : confused:
thanks. [/ QUOTE]
For NXP ICs 74AHCT595D [/ B] vs 74HC595D the main differences are:
1. Maximum frequency SHCP / STCP: 170MHz [/ B] vs 50MHz
2. HIGH-level input voltage: 2V [/ B] vs ~3V (74HC595D is less stable with a 3.3V Input logic level signal)
3. Different power consumption
Sincerely, Alexey Galkin.
anyone tried with NXP 74HC595D ??
what is the different with 74AHCT595D? : confused:
thanks. [/ QUOTE]
For NXP ICs 74AHCT595D [/ B] vs 74HC595D the main differences are:
1. Maximum frequency SHCP / STCP: 170MHz [/ B] vs 50MHz
2. HIGH-level input voltage: 2V [/ B] vs ~3V (74HC595D is less stable with a 3.3V Input logic level signal)
3. Different power consumption
Sincerely, Alexey Galkin.