troubleshooting my new dac

Status
Not open for further replies.
Jump to Latest
D

DaKi][er

Member
Joined 2003
#1
I’ve been working on a dac and finally got the circuit board for it and I’ve now soldered it altogether

board
schematic

Being a bit new to all this I don’t really know what to do now that it doesn’t exactly work 🙁

There were a few things that troubled me with something shorting the 5v digital line to ground and I was removing parts trying to track that down, and ended up finding it at C65 on the board (in the pic linked above)

Next after that, the reg102's use for regulating the digital side voltage weren’t playing nice
The 5v one was outputting 4.75v with 5.6v feeding into it
As well as the 3.3v ones outputting 4.1v and 4.25v with them being fed from the 5v (that was at 4.75v)

Playing with some of the space chips of these I still couldn’t get them to output the voltage that they should have been even under no or little load out of circuit

Just using an lm7805 and an lm317t bypassing the reg102's onboard, it is now being fed the correct 5v and 3.3v now

With that going now, I plug in an optical signal from a cd player in it and all I am getting is white noise and static out of it, pausing or removing the signal stops the sound so something is getting through

I'm thinking that the df1704 and pcm1704 wants a 24bit wordlength and its only getting 16bit, in setting up the pins on the chips I wasn’t exactly sure what ones I was meant to set them to so I chose the ones that seemed to be the standard values

I am sort of stumped and where to go now, I don’t have an oscilloscope or fancy test equipment to play with, so if anyone could look over what I have done and point out anything I have missed I would be very grateful

Thanks,
 
#3
Hi DaKi][er,
I have in my hands the PCM1704 EVM Board (and must say that it sounds really good, I use it as reference DAC..), and I see that it uses more or less the same components than your project.

On the EVM there are 26 jumpers/switches in the digital section, to try all possible settings. I made a picture, if this can help you, of the right settings for receiving normal SPDIF signals from CD players and soundcards:

An externally hosted image should be here but it was not working when we last tested it.


You can find the settings meaning in the PCM1794 EVM manual.

Bye,
Paolo

PS: in your post both the links go to the PCB image..
This is the schematic link: http://www.mypage.tsn.cc/dakiller/dac1/dacs.png
 
#4
Ahh yes, the schematic is here - http://www.mypage.tsn.cc/dakiller/dac1/dacs.png (i just copied the board hyperlink and didn’t change the url 🙁 )

Thankyou very much for that pic of the evaluation board paologatto, I did look through the documentation on it but it didn’t explain it any better than the datasheets

I’ll have a look over it in detail, not looking forward to changing pins on the df1704; soldering (as well as unnecessarily de-soldering) was fun enough
 
#5
paologatto said:
You can find the settings meaning in the PCM1794 EVM manual.
[/B]
Click to expand...

I meant PCM1704 EVM..

Just tell me also if you want that I try some settings change according with your schematic. I have not deeply examined all the settings changes in the EVM, but I remember that with some combinations the effect was similar to what you describe as digital "crackling" noise.

Since I have the EVM board now mounted and working, it's not a problem for me to make some experiments..

bye


PS: and.. YES! it accepts the SPDIF out of the PCM2902! 😉
 
#6
The only thing I can see that I got was the IW0 pin I had set high for 24bit input and it should be low for 16bit, I probably set it to that thinking I would get a 24bit dac overall and it would have just worked with 16 bit as well

So I managed to lift that pin on the df1704 and pulled it low but now i am getting nothing at all, it’s getting late here so I’ve left it till tomorrow

The original sound I wouldn’t call it 'crackling', it was a white noise like sound but you could hear changes in it that might have been quiet parts in the music or such
 
#7
DaKi][er said:
The only thing I can see that I got was the IW0 pin I had set high for 24bit input and it should be low for 16bit, I probably set it to that thinking I would get a 24bit dac overall and it would have just worked with 16 bit as well

So I managed to lift that pin on the df1704 and pulled it low but now i am getting nothing at all, it’s getting late here so I’ve left it till tomorrow

The original sound I wouldn’t call it 'crackling', it was a white noise like sound but you could hear changes in it that might have been quiet parts in the music or such
Click to expand...

I have not checked in your schematic how you configured the in/out formats of the DF1704 and of the PCM1704, but hope that the following info can be useful.

Here are some very experimental results of settings changes.. try to understand what is more near to your case:

- Receiving 16 bit PCM data from a CD player, listening to Alexander OST by Vangelis 🙂

Modifiying the following bits:

---------------------------------
[1]

IIS = H
IW0 = L
IW1 = L
OW0 = H
OW1 = H

Sounds perfect
---------------------------------
[2]

IIS = L
IW0 = L
IW1 = L
OW0 = H
OW1 = H

hear like white digital noise. Silence if I stop the CD
---------------------------------
[3]

IIS = H
IW0 = H
IW1 = L
OW0 = H
OW1 = H

Sounds perfect, like case [1]
---------------------------------
[4]

IIS = H
IW0 = L
IW1 = H
OW0 = H
OW1 = H

Digital noise like case [2]
---------------------------------
[5]

IIS = H
IW0 = L
IW1 = L
OW0 = L
OW1 = H

or

OW0 = H
OW1 = L

or

OW0 = L
OW1 = L


Sounds very bad and/or very attenuated, like you were playing with a 2 bit DAC
---------------------------------

I tried also changing the SW101 that chenges the PCM1704 20/24 bit data interface. When on 20 bit (and configuration [1] above) the output is very distorted. On H (24 bit) works fine.

So, I suggest you to investigate IIS, IW1, IW0, OW1 and OW2.

Let me know..

Bye,
Paolo
 
#8
it was originaly configured for your case [3] at the start and i was getting noise and changing the IW0 to then be your case [1] then just gave me silence

i cant really play with jumpers, it is just directly soldered to the board and changeing the IW0 was hard enough
 
#9
DaKi][er said:
it was originaly configured for your case [3] at the start and i was getting noise and changing the IW0 to then be your case [1] then just gave me silence

i cant really play with jumpers, it is just directly soldered to the board and changeing the IW0 was hard enough
Click to expand...


uhm... have you checked that the CS8416 is configured for I2S format? Actually in the EVM there is a CS8414 but the behavior should be quite similar..
 
#15
paologatto said:
Try to write here the complete configuration set of the DF1704 and I'll try to check with the PCM1704 EVM if works correctly.

By the way, what do you use as signal source? I mean, a soundcard, a CD player...?
Click to expand...
Combinations of -

I2S - H
IW0 - H/L
IW1 - H/L
MODE - L
MD - L
MC - L
ML - H
DEM - L
SF0 - L
SF1 - L
OW0 - H/L
OW1 - H/L

I’m using a dvd player with a cd playing via optical out and an iRiver h120 optical out

I’m not getting anything out now, no static at all and just silence so I don’t think it is to do with the df1704 anymore

tobias_svensk said:
If it makes you feel a little bit better the board looked nice 🙂
Click to expand...
it looks even nice in person, just doesn’t sound as good as it looks :bawling:
 
#16
Bump.

A bit of an update now, I managed to get a scope onto it to see what was happening (one of my mothers friends husband is an ex telecom engineer who was happy to help what he could)

Plugging in a Yamaha CD player sp-dif out

Coming out of the cs8416 -
LR clock - 44.0989 MHz
Bit clock - 2.82233 MHz
Data out - data coming out in packets with 1/2 the packet filled up and the other is blank at the same rate of the Bit clock
Master clock - 11.28932 MHz

This is fed into the DF1704, and then the output of that is -
Bit clock - 11.28930 MHz
Word clock - 352.791 kHz
Data L and R - on the scope what looked like 32 bit packets with 24 bits of them used

This then went through the isolators unchanged and into the pins of the dac chips

Now the problem is that there is nothing coming out of the dac's at all, 0V and 0A, we tried to put a 1-10k resistor from the dac output pin to ground and there was nothing with that either
This was the same result with both of the chips here

I’m thinking that there is something wrong with the chips but the fact that both aren’t working is unlikely that I got a bad chip on both
My next thought is to replace them but they will cost me around $100 aus to get them I don’t want to do it if it isn’t going to fix it, so I’m asking if anyone else has a better suggestion???:bawling:
 
#17
DaKi][er said:
Bump.

Now the problem is that there is nothing coming out of the dac's at all, 0V and 0A, we tried to put a 1-10k resistor from the dac output pin to ground and there was nothing with that either
This was the same result with both of the chips here

I’m thinking that there is something wrong with the chips but the fact that both aren’t working is unlikely that I got a bad chip on both
My next thought is to replace them but they will cost me around $100 aus to get them I don’t want to do it if it isn’t going to fix it, so I’m asking if anyone else has a better suggestion???:bawling:
Click to expand...


did you check if DF output / PCM input format match ?
 
#18
Guido Tent said:
did you check if DF output / PCM input format match ?
Click to expand...
How would that be done?

All the pins for the output format of the cs8416 and in/ out formats of the DF are correct, and from what I understand of the datasheets what is being fed into the DAC is in the correct format and the DF is doing its job of 8X OS and splitting the L and R
 
#20
kevinkr said:
Wicked dumb question, but as data format sounds valid have you checked that the dacs are actually getting power? I have done worse believe me...😀 Sometimes it is something really simple that gets you, but you are moving in the right direction.

Kevin
Click to expand...
Yes, the voltages to the dac's measured +4.902 and -5.015 which is in the power supply requirements range (±4.75 to ±5.25) but one thing in the datasheets said under the Absolute Maximum ratings "Supply Voltage Differences ±0.1V" and I'm wondering if it could be be between the +ve and the -ve rails accuracy but I would see it as the difference between the analog and digital voltage inputs on the chip but they are both fed off the same voltage supply here
 
Status
Not open for further replies.
Top Bottom