Hi,
Sorry to hear that. I would suggest returning it and getting your money back.
The problem with most cheap chinese kits and assembled stuff is that they are basically put together by people who have a brain full of audiophile nonsense (which is why they so often omit essential ceramic bypass cap's as they believe "ceramic cap's = bad"), limited if any formal EE training (this includes many people who have an EE degree in china, the standards are revoltingly bad).
And the approach to "design" is best described as "cut and paste" of anything screen-scraped off the net from "famous people" and Chip Makers App-Notes without being able to read and understand the accompanying text but with the application of "improvements" based on the audiophile nonsense they believe in.
The results are often, how shall I say, interesting. Often enough the stuff does not work at all (due to fake parts, missing parts selection etc.) and it then when these chinese Kopy-Kat's turn up on Forums like this and ask for free help, often from the very people who's work they copied badly for personal gain...
Ciao T
Sorry to hear that. I would suggest returning it and getting your money back.
The problem with most cheap chinese kits and assembled stuff is that they are basically put together by people who have a brain full of audiophile nonsense (which is why they so often omit essential ceramic bypass cap's as they believe "ceramic cap's = bad"), limited if any formal EE training (this includes many people who have an EE degree in china, the standards are revoltingly bad).
And the approach to "design" is best described as "cut and paste" of anything screen-scraped off the net from "famous people" and Chip Makers App-Notes without being able to read and understand the accompanying text but with the application of "improvements" based on the audiophile nonsense they believe in.
The results are often, how shall I say, interesting. Often enough the stuff does not work at all (due to fake parts, missing parts selection etc.) and it then when these chinese Kopy-Kat's turn up on Forums like this and ask for free help, often from the very people who's work they copied badly for personal gain...
Ciao T
Agree with ThorstenL,
My whole DAC is a cut&paste job from datasheet schematics (at least for the digital section). Evidently they did not read the accompanying text , not even the part about the bypass caps being close to IC's, as they actually placed them kind-of like in the schematic (ie: soemhing that was farther from IC on the schematic ended up farther on the PCB as well). Go figure!
There are few 'real' high-end manufacturers in China like Audio-GD that appear to design their products, but expensive. They did get excellent reviews is some western magazines. I also rememebr Jungson making a spectacular Class A power amp, but 220v (hence i didi not get one).
My whole DAC is a cut&paste job from datasheet schematics (at least for the digital section). Evidently they did not read the accompanying text , not even the part about the bypass caps being close to IC's, as they actually placed them kind-of like in the schematic (ie: soemhing that was farther from IC on the schematic ended up farther on the PCB as well). Go figure!
There are few 'real' high-end manufacturers in China like Audio-GD that appear to design their products, but expensive. They did get excellent reviews is some western magazines. I also rememebr Jungson making a spectacular Class A power amp, but 220v (hence i didi not get one).
I desoldered the DF1704 daughter board to verify all the settings of the hardware config and install the bypass cap. (there are traces below, and it's not a pin-to-pin daughter board... Result - as per schematics
Any improvenets I can do here ? There is an optonal crystal but not sure if it applies to my dac design.
Thanks
Any improvenets I can do here ? There is an optonal crystal but not sure if it applies to my dac design.
Thanks
Another question:
My DAC uses separate power supplies for the analog and digital portions of both PCM1704.
From the datasheet:
POWER SUPPLIES
The PCM1704 only requires a ±5V supply for operation. Both positive supplies (+VDD and +VCC) should be tied together at a single point and connected to a single +5V analog power supply. Similarly, both negative supplies (–VDD and –VCC) should be tied at a single point and connected to a single –5V analog power supply. No advantage is gained by using separate analog and digital power supplies. It is more important that the analog supplies used to drive these pins are as noise and ripple free as possible to reduce coupling of supply noise to the output.
So would I gain from rewiring the 2 power supplies as one per PCM1704 instead of Vcc and Vdd for both ICs?
Thanks
My DAC uses separate power supplies for the analog and digital portions of both PCM1704.
From the datasheet:
POWER SUPPLIES
The PCM1704 only requires a ±5V supply for operation. Both positive supplies (+VDD and +VCC) should be tied together at a single point and connected to a single +5V analog power supply. Similarly, both negative supplies (–VDD and –VCC) should be tied at a single point and connected to a single –5V analog power supply. No advantage is gained by using separate analog and digital power supplies. It is more important that the analog supplies used to drive these pins are as noise and ripple free as possible to reduce coupling of supply noise to the output.
So would I gain from rewiring the 2 power supplies as one per PCM1704 instead of Vcc and Vdd for both ICs?
Thanks
Audio GD also did no design but cut and paste, just they are not cut and paste from data sheet or application notes, but clone serveral brands then put everything together, maybe a ML digital section + Krell analog stage, still cut and paste.
Took apart the wm8805 PCBs and found that the crystal is powered at 3.3v so only missing bypass caps.
After removing the buffer circuit from the dac the wm8805 receiver module doesn't seem to lock well onto the spdig. There is a 75 ohm resistor at the RCA jack (signal to ground). Only one .01uF before a direct connection to the Wm8805 input pin.
Added bypass caps to the pcm1704s and must have overheated a leg, since one of them is getting hot and mo output. I'm well above average at soldering, I must have been inattentive.
Wondering why would the wm8805 have trouble with the spdif. Tried two different players, same. Error is intermittent.
My scope is only 20mhz and I see spdif at about 1v p-p but no detail on the waveform.
After removing the buffer circuit from the dac the wm8805 receiver module doesn't seem to lock well onto the spdig. There is a 75 ohm resistor at the RCA jack (signal to ground). Only one .01uF before a direct connection to the Wm8805 input pin.
Added bypass caps to the pcm1704s and must have overheated a leg, since one of them is getting hot and mo output. I'm well above average at soldering, I must have been inattentive.
Wondering why would the wm8805 have trouble with the spdif. Tried two different players, same. Error is intermittent.
My scope is only 20mhz and I see spdif at about 1v p-p but no detail on the waveform.
Hi,
The SPDIF grounds are grounded.
You need to confirm the SPDIF input(s) are operating in "SPDIF" Mode, not CMOS.
Sorry to hear about the dead PCM1704, SMD soldering is not always for the faint hearted...
Ciao T
The SPDIF grounds are grounded.
You need to confirm the SPDIF input(s) are operating in "SPDIF" Mode, not CMOS.
Sorry to hear about the dead PCM1704, SMD soldering is not always for the faint hearted...
Ciao T
Yes I have noticed most of this designers ideas are taken from either Diyaudio or directly from the actual component data sheet , I totally agree with cut and paste comment
.I have got mine working for now over two hours and then it loses signal lock so hopefully some more playing around will get this thing locking on , the seller/designer will not give me the complete schematic to fault find very annoying .
Hi,
That could be software related. The WM8805 has many "undocumented" features and correct software implementation is a major PITA.
Ciao T
That could be software related. The WM8805 has many "undocumented" features and correct software implementation is a major PITA.
Ciao T
Hello,
ThorstenL: Thanks, not much left to check since the settings are set via an atmel chip... The cricuit seems ok. Must be a bad setting in the registers.
paul_peraic:
I have the full schematics, but the product delivered is a bit different (single regulator, missing bypass caps in some places, no 75R on inputs, different crystal).
Problem is that all the settings are set via an atmel chip, so no way of knowing if a bad setting is causing the lock issue (and there are lots of settings). Other than the settings, it may be defective, but I doubt it. I have 2 units, both lock but have intermitent errors (error led lights up).
There isn't anything between the input pins and the IC.
PM me and I'll email you the schematics.
ThorstenL: Thanks, not much left to check since the settings are set via an atmel chip... The cricuit seems ok. Must be a bad setting in the registers.
paul_peraic:
I have the full schematics, but the product delivered is a bit different (single regulator, missing bypass caps in some places, no 75R on inputs, different crystal).
Problem is that all the settings are set via an atmel chip, so no way of knowing if a bad setting is causing the lock issue (and there are lots of settings). Other than the settings, it may be defective, but I doubt it. I have 2 units, both lock but have intermitent errors (error led lights up).
There isn't anything between the input pins and the IC.
PM me and I'll email you the schematics.
Hi,
Maybe restore the Input Amplification with a 3.3V Supply? I can tell you off line how to fix that circuit properly (I do not want to give the CKM's [Chinenes Kopy Mao's] something to Kopy that actually work's).
But if the IC generally locks to a 0.5V P-P SPDIF Signal the register setting is likely proper and the unlock issue is somewhere else.
Ciao T
Maybe restore the Input Amplification with a 3.3V Supply? I can tell you off line how to fix that circuit properly (I do not want to give the CKM's [Chinenes Kopy Mao's] something to Kopy that actually work's).
But if the IC generally locks to a 0.5V P-P SPDIF Signal the register setting is likely proper and the unlock issue is somewhere else.
Ciao T
My input spdif + goes from RCA connector to a relay (toslink/spdif) via 3cm of twisted wire from RCA to PCB then to the IC input. The ground is grounded on the PCB. The 75R is right at the RCA connector.
Should I bring the ground of the SPDIF closer to the WM8805?
I wonder if in default hardware mode, spdif in, 12Mhz crystal, i2s out, the WM8805 would work better. Most likely the default settings would be conductive to 'standard' operation.
Should I bring the ground of the SPDIF closer to the WM8805?
I wonder if in default hardware mode, spdif in, 12Mhz crystal, i2s out, the WM8805 would work better. Most likely the default settings would be conductive to 'standard' operation.
Last edited:
The SPDIF input of a DAC is a primary source of common-mode noise and the routing of its gnd potentially has a large effect on sound quality. Therefore I'd recommend not bringing the ground from the RCA input on to the pcb, rather it needs routing away to a star ground. Just this week a DAC I'm playing with kept making intermittent hissing noises, I eventually tracked it down to the receiver losing lock because I'd put a ferrite bead on the ground to the SPDIF input in a misguided attempt to reduce common-mode noise. I've now changed to using a transformer on the input, a resistor to ground to soak up some of the noise and the problem hasn't recurred.
Hi,
Transformers can help a lot, but they carry their own problems. Before Jocko Homo took his leave of this place for his own good reasons he could have told you some of this...
Ciao T
Transformers can help a lot, but they carry their own problems. Before Jocko Homo took his leave of this place for his own good reasons he could have told you some of this...
Ciao T
The voltage regulator for the PLL could be a problem - it's about a 1.7mA load and some vregs need 3.5mA-10mA minimum load current to be stable (i.e. the LM317 from the OP or LT1086 or AMS1117-3.3). I've seen the vreg oscillate or be extra noisy with higher Vout than expected. Not that I would use one of these noisy beasts for top performance.
The output divider adds about 4.2mA load and you could add a resistor to up the load. I find a dedicated vreg for every pwr pin on every IC can be overkill.
The output divider adds about 4.2mA load and you could add a resistor to up the load. I find a dedicated vreg for every pwr pin on every IC can be overkill.
My CS8414 to WM8805 converter board loses lock after about an hour and a half the lock light starts flickering gradually and crackling starts in the right channel then intensifies after this it totally loses lock the SPDIF input does not have a transformer.
Then I shuffle through the on board input selector and it is good for another hour and a half very strange .
It is very difficult to go back the CS8414 after getting a taste of the WM8805 receiver it has so much more of everything .
Then I shuffle through the on board input selector and it is good for another hour and a half very strange .
It is very difficult to go back the CS8414 after getting a taste of the WM8805 receiver it has so much more of everything .
Last edited:
Yes I've followed Jocko's writings with immense interest. He seems to me to put undue emphasis on waveform fidelity and ignore common-mode noise. So he disses (for example) the low interwinding capacitance designs of SC because such buy their low capacitance at the expense of higher leakage inductance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.