Thank you very much! I understand you. But in the second version pcb (ak4399&ak4118, see "4399V2-lt1085.jpg") use all three of the capacitor. As I understand you removed the capacitor, but I can add the missing two capacitors 100nf in parallel capacitors 100uf. As I understand it, they do not really need. But the worst of them will not. Important that they were all or not. Thank you!
Sorry for the bad English (I use Google).
I'll try to explain.The pic ground1.jpg green line marked by grounding trace that starts from the transformer and ends at the two capacitors after the diodes. Purple line marks the grounding trace that starts from the two capacitors are used for op-amps. In pic ground2.jpg blue line indicates the grounding trace that used to lm317 / lm337. Grounding trace indicated by the green line is connected to the grounding trace indicated by the purple line after about two capacitors diodes. Pic ground3.jpg shows the grounding trace for the op-amp.
Attachments
You can see clearly on the pictures that star-ground is used. Two RCA return tracks to star ground and each channel's opamp have their own traces the same star point.
Given the current design, I don't think ground tracks of analog stage psu can be improved. The blue tracks, from regulator adjustment circuits to ground, are going straight to the smoothing caps. Ideally they would connect to the star point , but indirectly they do because of the relatively short purple track on topside of the board.
Given the current design, I don't think ground tracks of analog stage psu can be improved. The blue tracks, from regulator adjustment circuits to ground, are going straight to the smoothing caps. Ideally they would connect to the star point , but indirectly they do because of the relatively short purple track on topside of the board.
Thank you very much! May also make the grounding trace a power DAC? Cut off the ground after regulators (marked in black, see cut-groud-dac.jpg).
IMHO, possible in the analog stage psu cut off pink grounding trace and ground RCA socket connected to the ground after two large capacitors diodes (see cut-groud-psu.jpg). It will be right? Or pointless?
IMHO, possible in the analog stage psu cut off pink grounding trace and ground RCA socket connected to the ground after two large capacitors diodes (see cut-groud-psu.jpg). It will be right? Or pointless?
Attachments
You're welcome .No, this first option won't work. If you do those black cuts, the regulators won't have their ground pins connected. The whole point is to make this pin sense the ground potential at the load side, for better regulation.
As I said, I see no point in changes of the ground tracks in the analog stage.
Regards, Nino
.No, this first option won't work. If you do those black cuts, the regulators won't have their ground pins connected. The whole point is to make this pin sense the ground potential at the load side, for better regulation.As I said, I see no point in changes of the ground tracks in the analog stage.
Regards, Nino
Hello,
my logic analyser is here, I have tested it with the original and the Arduino controller.
The start-up sequence of the Aurduino controller looks like this (sample rate is 250 MHz)
http://www.diyaudio.com/forums/[url...ata/500/thumbs/Start_Sequence.png[/IMG][/url]
The first 3 SPI commands of the AK4118 look like this
I do not see any problems. These commands are specified in datasheet, but the receiver has no lock.
I think that the very short changes of the low/high state (for example of the AK4118 CS channel) are the ADC Errors. The original controller has this "switches" too.
my logic analyser is here, I have tested it with the original and the Arduino controller.
The start-up sequence of the Aurduino controller looks like this (sample rate is 250 MHz)
http://www.diyaudio.com/forums/[url...ata/500/thumbs/Start_Sequence.png[/IMG][/url]
The first 3 SPI commands of the AK4118 look like this
I do not see any problems. These commands are specified in datasheet, but the receiver has no lock.
I think that the very short changes of the low/high state (for example of the AK4118 CS channel) are the ADC Errors. The original controller has this "switches" too.
Last edited:
Hi pca01,
Thanks a lot for testing the start-up sequence. On the second picture, the magnified one, the spikes are only on the CS line, not on CLK or MOSI lines. So I don't think it's ADC errors of your analyzer, the spikes on the CS line are likely to be the cause of your DAC controller not working. When state of CS line changes during transmission, the sent/received byte isn't valid. The CS line only suffers from the spikes. What pull-up resistor value does it have? My DAC has 4k7 resistors on each line.
You could also try to add a small cap across ground and CS line, to filter the spikes.
Nino
Hi Nino,
I think it is not a problem of the DAC controller, but it's a problem of this analyser.
The original controller has the same spikes, but it works.
I have found in the internet the same descriptions and screenshots from other peoples. I have already buy an another analyser (from Zeroplus) and would like to return this analyser to seller.
I think it is not a problem of the DAC controller, but it's a problem of this analyser.
The original controller has the same spikes, but it works.
I have found in the internet the same descriptions and screenshots from other peoples. I have already buy an another analyser (from Zeroplus) and would like to return this analyser to seller.
Attachments
If the Arduino doesn't lock: Check the INT0 pin, when the receiver locks to a signal it must go LOW. If the pin is LOW, but still no lock, the frequency isn't correctly read from the register of the AK4118. Then check your MISO line. The controller outputs audio only when a signal with valid fs is supplied.
What about DIYINHK voltage decoupling modification in this thread?
http://www.diyaudio.com/forums/digital-line-level/229513-akm-ak4399-dac-pcb-ic-ebay.html
http://www.diyaudio.com/forums/digital-line-level/229513-akm-ak4399-dac-pcb-ic-ebay.html
You mean this message:
http://www.diyaudio.com/forums/digital-line-level/229513-akm-ak4399-dac-pcb-ic-ebay.html
IMHO this mod done on the board.
Attachments
Ok, thanks, you're welcome. The Arduino controller is a nice project. terranigma is also using it, but has afaik sometimes start-up problems. For me it's working 100% now (testing every day, no problems). Keep the pull-up resistors on MOSI and CLK (at the DAC side that is). I'll post new software probably soon, with small adjustments to the start-up procedure of the DAC. Keep us posted when you've got results!
Nino
Nino
I do not understand, why not? The foto shows that the contacts between 24-25, 26-27, 29-30, 31-32 soldered smd-capacitors. According to the scheme dual ak4399 between contacts 24-25, 29-30 soldered smd-capacitors and connected contacts 25-26 and 30-31 and further connected to the electrolytic capacitors. Contact 27, 32 connections to the ground. So I think this mod is made, plus electrolytic capacitors.
![http://www.diyaudio.com/forums/[url...ata/500/thumbs/Start_Sequence.png[/IMG][/url]](http://www.diyaudio.com/forums/[url=http://www.diyaudio.com/forums/gallery/showphoto.php/photo/12248][IMG]https://origin.dastatic.com/forums/gallery/data/500/thumbs/Start_Sequence.png[/IMG][/url]){kind=link}
Improve the power supply circuit. As here.
L.K.S Audio MH-DA003 ES9018 * 2 Dual-Channel Decoder Amanero USB DSD DAC Crystek Femto Clock - Professional Audio Store - Shenzhenaudio.com
"5. The use of Class A discrete analog portion of supply Toshiba sound supervisor C2240 / A1930 composition.
6. Low voltage, digital, analog circuit using LT317 and LT1763 regulator ultra low noise, and each functional component has independent regulator supply."
I will be compiled lt1763 apply to both digital and analog power DAC. I will Lt1963 + lt1763. Separate stabilizers digital and analog portion of each chip ak4399.
To power operational amplifiers I ordered this board.
MisterTao: Leading Taobao Agent - Taobao Product - Super Power Supply Kit Parallel Group
Since the power is adjustable, to increase the power to 15V. Operational amplifier will oрa1611. hdam modules I liked less.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.