Is there any reason/solution found for this case? Ran today into same problem. No heating issues on chips, dac keeps running smoothly after display freezes. Board is obviously first version.
I'm sorry to hear that, knoberik. Maybe your board is different indeed.
What value pull-up resistors does your pcb have? The array is located next to the connector for the controller cable. The controller board uses also pull-up resistors. What value are they? Can you identify the micro controller type?
What value pull-up resistors does your pcb have? The array is located next to the connector for the controller cable. The controller board uses also pull-up resistors. What value are they? Can you identify the micro controller type?
Pull-up resistors are measured 10k, all of them and both on main board and on display board (on display board are silk markings 4k7*8, but those are really 10k arrays intalled).
Controller is marked STC 8952RC.
BTW the display will not freeze without input signal, it reacts for buttons as supposed to. It freezes only with input signal applied, within few seconds.
How you can separate the different versions? All boards I see in this branch fotos look very much the same. My board has red solder mask and looks similar like other boards with red solder mask, photographed here.
Before any mods the dac ran stable, not freezing for 24h in row. After that one of LT1085 (R Analog supply) went bad, as it had bad thermal connection with cooler, didn't notice this first. R-channel went mute because of this. Those regulators showed as second hand, so I replaced all of them.
Last edited:
Weiliang version:
Assembled New Dual AK4399 AK4113 DAC with LCD software 192K - $116.00 : Enjoy in your hi-fi project, diy tube amp, amplifier diy, amp diy
yuan-jing version:
Yuan Jing Audio - 32Bit / 192KHz LCD DAC Decoder AK4399+WM8805+NE5532 Optical, Coaxial - $105.00
I'm not sure, but maybe YJ-version was just me.
What mods to do?
In this case it's the red board version.
+5V supplies star groundings mod, 4113 ground plane separation, separate 3.3V for 4113 A and D parts (added separate 3.3V low noise regulator for analog part of 4113). VTDD mod to +5V.
Bringing VTDD back to +3.3V does not affect the display freezing, it still freezes fast.
Unfortunately, I not badly understood you. In what this mod consists?
Unfortunately, I not badly understood you. In what this mod consists? The chip ak4113 in my opinion has no analog part.
Yes, see: http://www.diyaudio.com/forums/digi...g-dual-x2-ak4399-dac-lcd-101.html#post4268390
Please take a look at post #981 for example. And there is some more, follow NinoSimona posts.
Sorry for disturbing, I really didn't pay much attention there is difference between Weiliang and Yuan Jing versions of DAC's, they look very much similar. Obviously I'm dealing with Yuan Jing product. But are there some meaningful differences at all? One seems pretty much copy of another.
BTW I solve the display freezing problem by bridgeing back small ground plain under AK4113. Despite there is another gnd route going to 5V PSU GND, there is need for some reason to be those ground planes connected.
Other mods described in this branch earlier you can do for Yuan Jing board, if you feel they help, without problem.
BTW I solve the display freezing problem by bridgeing back small ground plain under AK4113. Despite there is another gnd route going to 5V PSU GND, there is need for some reason to be those ground planes connected.
Other mods described in this branch earlier you can do for Yuan Jing board, if you feel they help, without problem.
Attachments
knoberik: Glad you found a solution. 
Still I don't really understand why the ground plains connection is needed in your case... I don't see big differences in the layout of the board. Controller is SM8952AC, 5V version, so TVDD mod is a good idea.
Pull-up resistors at main board are 4k7 in my case. If I would replace them with 10k, my Arduino controller won't be able to communicate with both dacs anymore. Especially CLK line seems to be susceptible to noise. Maybe the connection of the ground plains provides a better return path for the CLK line, making it less sensitive to noise generated by the dacs (?), but it's also affecting performance of the left channel dac chip. Anyway, ground routing around AK4113 is done very badly, most probably auto-route. Guess we have to live with that...
On the red pcbs no WM8805 is used as receiver, the yuan-jing discription is clearly incorrect (I guess they copied it from single AK4399 version).
Did you have good progress sound-wise after the mods you did?
Nino
Still I don't really understand why the ground plains connection is needed in your case... I don't see big differences in the layout of the board. Controller is SM8952AC, 5V version, so TVDD mod is a good idea.Pull-up resistors at main board are 4k7 in my case. If I would replace them with 10k, my Arduino controller won't be able to communicate with both dacs anymore. Especially CLK line seems to be susceptible to noise. Maybe the connection of the ground plains provides a better return path for the CLK line, making it less sensitive to noise generated by the dacs (?), but it's also affecting performance of the left channel dac chip. Anyway, ground routing around AK4113 is done very badly, most probably auto-route. Guess we have to live with that...
On the red pcbs no WM8805 is used as receiver, the yuan-jing discription is clearly incorrect (I guess they copied it from single AK4399 version).
Did you have good progress sound-wise after the mods you did?
Nino
Hi Nino,
The resistor array on my main board is 10K*8. Should I replace it with 4K7 one according to your statememnt above?
In this article circuits reduce noise are described:
http://www.edn.com/Pdf/ViewPdf?contentItemId=4422750
In this article it is approved:
"Since most of the preamp/VCO applications where these sorts of noise reduction circuits tend to be applied are class A, they tend to consume a constant amount of supply current."
This phrase belongs to this dac? Whether is correct what is consuming of current in digital part of DAC and analog part is invariable (constant). It is possible to use these circuits in this dac?
http://www.edn.com/Pdf/ViewPdf?contentItemId=4422750
In this article it is approved:
"Since most of the preamp/VCO applications where these sorts of noise reduction circuits tend to be applied are class A, they tend to consume a constant amount of supply current."
This phrase belongs to this dac? Whether is correct what is consuming of current in digital part of DAC and analog part is invariable (constant). It is possible to use these circuits in this dac?
Most benefit seem to come from 5V real star grounding mod. Other mods for my board are under work, nothing special to report yet.
About display freezing of Yuan Jing (red) board - there still is possibility for display freezing accure, if input data flow interrupts. For example by switching channels a lot and fast on digital TV-box receiver and if the data flow interrupts in this situation.
When data flow does not interrupt, the display will not freeze, at least not for tested 50 hours of continuous worktime.
Sound possibilities of this double-AK4399 feel very promising, very worth of studying how to get utmost out of this DAC.
About display freezing of Yuan Jing (red) board - there still is possibility for display freezing accure, if input data flow interrupts. For example by switching channels a lot and fast on digital TV-box receiver and if the data flow interrupts in this situation.
When data flow does not interrupt, the display will not freeze, at least not for tested 50 hours of continuous worktime.
Sound possibilities of this double-AK4399 feel very promising, very worth of studying how to get utmost out of this DAC.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.