Hello all.
I have installed a DAC board in my Cambridge Audio D100 as a bit of fun, and to find out if the player can be improved.
http://cgi.ebay.co.uk/ws/eBayISAPI....o=DR&its=S%2BI%2BSS&itu=ISS%2BUCI%2BSI&otn=12
The D100 has a hole on the rear which states "upgradeable to digital out" but I can't find where I should take the connection for the DAC. The player had a toslink opticle out which has three wires going towards it. These come from CN1 D OUT, next to that there is CN2 DATA OUT.
Please can anyone help?
I can supply more photos if needed.
I have installed a DAC board in my Cambridge Audio D100 as a bit of fun, and to find out if the player can be improved.
http://cgi.ebay.co.uk/ws/eBayISAPI....o=DR&its=S%2BI%2BSS&itu=ISS%2BUCI%2BSI&otn=12
An externally hosted image should be here but it was not working when we last tested it.
The D100 has a hole on the rear which states "upgradeable to digital out" but I can't find where I should take the connection for the DAC. The player had a toslink opticle out which has three wires going towards it. These come from CN1 D OUT, next to that there is CN2 DATA OUT.
Please can anyone help?
I can supply more photos if needed.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
hallo
you must find out the type of the pcm processor. In most cases it comes from SONY by cambridge audio. If there are CXD2500Q, please refer
http://www.diyaudio.com/forums/showthread.php?threadid=62993
best regards
Andreas
you must find out the type of the pcm processor. In most cases it comes from SONY by cambridge audio. If there are CXD2500Q, please refer
http://www.diyaudio.com/forums/showthread.php?threadid=62993
best regards
Andreas
The black shielded cable that goes from the mainboard to the output board will be the S/P DIF digital output. The cable will be carrying signal, ground and Vcc lines. The thicker black core will be the signal, the red core is probably Vcc and the white core is probably ground. Use a multimeter to check.
If that all checks out, connect signal to the centre pin of the RCA, and ground to the shield. Hopefully the logic levels from the mainboard will be compatible with the DAC's receiver.
If that all checks out, connect signal to the centre pin of the RCA, and ground to the shield. Hopefully the logic levels from the mainboard will be compatible with the DAC's receiver.
Andreas: Thank you for your reply. The Tentlabs XO3 is a nice idea, but it is beyond the budget of this project for the time being.
amc184: Thank you for your reply. This is the info that I was hoping for. I have separated the cable that goes to the Toslink converter, and it looks like the shield was the thicker black wire, also there is a thin white wire, and a thin red wire.
Which is likely to be the digital signal and ground? I have a decent mulitimeter. What should I be looking for?
The red wire is +5v using the shield as ground
The white wire is +2.4v using the shield as ground
amc184: Thank you for your reply. This is the info that I was hoping for. I have separated the cable that goes to the Toslink converter, and it looks like the shield was the thicker black wire, also there is a thin white wire, and a thin red wire.
Which is likely to be the digital signal and ground? I have a decent mulitimeter. What should I be looking for?
The red wire is +5v using the shield as ground
The white wire is +2.4v using the shield as ground
If you're measuring 2.4 volts with a meter, that suggests the actual signal may be more like 5 volts. If so, it would be reasonable to use an attenuator or something to reduce the voltage to normal S/PDIF levels. And possibly a buffer circuit. But check the specs of the DAC; maybe it's happy with 5V digital input signals.
See this thread, which includes the schematic and pictures of the digital output I added to a laserdisc player:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=130129&highlight=
See this thread, which includes the schematic and pictures of the digital output I added to a laserdisc player:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=130129&highlight=
Yeah, thats exactly what I expected.
The white wire is S/P DIF at TTL level (varying extremely fast between 4.8V and 0V, which your meter averages to 2.4V). The optical transmitter requires a TTL level input, so as I said, this is what i expected.
The RXIN input on the DIR9001 is TTL tolerant, so you can put the white (2.4V) wire to pin 20 on the DIR9001, bypassing the input stage. The red (5V) wire you can leave unconnected, that was the supply for the optical transmitter, you won't need it anymore. How the black wire (0V) is connected depends on your grounding scheme. Youcould connect it to ground on the board, but make sure that is the only connection the board has to ground. It's best to avoid a ground loop.
The white wire is S/P DIF at TTL level (varying extremely fast between 4.8V and 0V, which your meter averages to 2.4V). The optical transmitter requires a TTL level input, so as I said, this is what i expected.
The RXIN input on the DIR9001 is TTL tolerant, so you can put the white (2.4V) wire to pin 20 on the DIR9001, bypassing the input stage. The red (5V) wire you can leave unconnected, that was the supply for the optical transmitter, you won't need it anymore. How the black wire (0V) is connected depends on your grounding scheme. Youcould connect it to ground on the board, but make sure that is the only connection the board has to ground. It's best to avoid a ground loop.
amc184: Thanks again.
I had done bit of research since my last post and had discovered that coax spdif should only be 1v, so I was worried. It is good to hear that the DIR9001 can deal with 5v.
I found this cicuit that converts TTL to spdif. This was a possibility, but I thought that adding more components to the signal path could not be a great idea:
Then I looked at the cuicuit diagram supplied with the dac, and thought that I could take the two TTL digital signal wires (ignoring the 5v supply) that were entering the optical output of my CDP, and wire them directly to the two wires leaving the digital reciever on my DAC (ignoring its dc supply):
Or, I could just use an optical cable and retain all connections as originally designed.
amc184: your solution seems the best (assuming it works) as it minimises components in the signal path. The only issue is the grounding.
I will let you know how I get on tomorrow, but would also be grateful for further encouragement.
I had done bit of research since my last post and had discovered that coax spdif should only be 1v, so I was worried. It is good to hear that the DIR9001 can deal with 5v.
I found this cicuit that converts TTL to spdif. This was a possibility, but I thought that adding more components to the signal path could not be a great idea:
An externally hosted image should be here but it was not working when we last tested it.
Then I looked at the cuicuit diagram supplied with the dac, and thought that I could take the two TTL digital signal wires (ignoring the 5v supply) that were entering the optical output of my CDP, and wire them directly to the two wires leaving the digital reciever on my DAC (ignoring its dc supply):
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Or, I could just use an optical cable and retain all connections as originally designed.
amc184: your solution seems the best (assuming it works) as it minimises components in the signal path. The only issue is the grounding.
I will let you know how I get on tomorrow, but would also be grateful for further encouragement.
I am not suceeding very well here.
I connected the 2.4v (white wire) to where I thought pin 20 went to, and connected the shield to ground. Nothing.
It is not simple to find where I should attach the wires. Pin 20 goes to a resistor R9. This is difficult to locate as the board is marked under the resistors. I connected it to the one at the bottom of the screen as it was the only one that obviously went to the 74HC00 chip, albeit to a different pin compared to the schematic (attached pdf).
http://www.pic16.com/soft/PCM1793DAC_sch.pdf
I have tempoarily given up and am using a toslink cable to make the connection.
The good news is that it makes a very nice sound.
I connected the 2.4v (white wire) to where I thought pin 20 went to, and connected the shield to ground. Nothing.
It is not simple to find where I should attach the wires. Pin 20 goes to a resistor R9. This is difficult to locate as the board is marked under the resistors. I connected it to the one at the bottom of the screen as it was the only one that obviously went to the 74HC00 chip, albeit to a different pin compared to the schematic (attached pdf).
http://www.pic16.com/soft/PCM1793DAC_sch.pdf
An externally hosted image should be here but it was not working when we last tested it.
I have tempoarily given up and am using a toslink cable to make the connection.
The good news is that it makes a very nice sound.
Hi, dublin,
You might need to disconnect the multiplexer chip (U1) from the digital receiver chip (U4) before you connect your direct TTL S/PDIF input...
btw. the DIR9001 datasheet says that a toslink receiver module's TTL output can be connected directly to the RXIN pin, so there shouldn't be a problem with your signal, and it doesn't need converting (AFAIK).
Also, I'm fairly certain you have the correct TTL S/PDIF signal on the D100 (it looks like the grey cable is for the analog outputs.) I'm assuming the toslink output on the D100 stops working when you unplug the small data cable?
The 74HC00 chip (U1) is used as a multiplexer which lets you choose between toslink or coax input - with no toslink or coax connected, the output of U1 will be left high or low (I haven't taken the time to work out which yet, but it's probably low)...
If you just leave the mux chip connected when you connect your direct signal, the low (or high) output of mux chip will "clash" with your direct TTL input.
What I would do is to desolder one end of R9 (the end connected to pin 8 of U1), then solder your TTL input onto the loose end of R9. This should bypass U1 completely and connect the TTL signal to the DIR9001 (via R9, which should be fine.)
If this doesn't work, then I'm not sure?
OzOnE.
EDIT: Just thought I'd add that if you use a TTL S/PDIF instead of optical or coax, it probably won't make a hugely noticeable improvement to the sound...
The best upgrade would be to bypass the DIR9001 completely and input I2S into the DAC chip (assuming the D100 uses standard I2S). If the D100 uses a different format internally, you'll have to change the FMT pin voltages on the PCM1793 DAC.
You might need to disconnect the multiplexer chip (U1) from the digital receiver chip (U4) before you connect your direct TTL S/PDIF input...
btw. the DIR9001 datasheet says that a toslink receiver module's TTL output can be connected directly to the RXIN pin, so there shouldn't be a problem with your signal, and it doesn't need converting (AFAIK).
Also, I'm fairly certain you have the correct TTL S/PDIF signal on the D100 (it looks like the grey cable is for the analog outputs.) I'm assuming the toslink output on the D100 stops working when you unplug the small data cable?
The 74HC00 chip (U1) is used as a multiplexer which lets you choose between toslink or coax input - with no toslink or coax connected, the output of U1 will be left high or low (I haven't taken the time to work out which yet, but it's probably low)...
If you just leave the mux chip connected when you connect your direct signal, the low (or high) output of mux chip will "clash" with your direct TTL input.
What I would do is to desolder one end of R9 (the end connected to pin 8 of U1), then solder your TTL input onto the loose end of R9. This should bypass U1 completely and connect the TTL signal to the DIR9001 (via R9, which should be fine.)
If this doesn't work, then I'm not sure?
OzOnE.
EDIT: Just thought I'd add that if you use a TTL S/PDIF instead of optical or coax, it probably won't make a hugely noticeable improvement to the sound...
The best upgrade would be to bypass the DIR9001 completely and input I2S into the DAC chip (assuming the D100 uses standard I2S). If the D100 uses a different format internally, you'll have to change the FMT pin voltages on the PCM1793 DAC.
Ozone makes a very good point. When I said to connect the CDP S/P DIF line to the receiver IC, I also meant that you should disconnect the existing input, as like Ozone said, they certainly will interfere with each other.
I don't think TTL to optical to TTL is a good permanent solution. With the optical connection, there is a voltage threshold where the transmitter outputs light, and a luminance threshold where the receiver outputs a voltage. If these thresholds are even slightly different from transition to transition (and they will be), phase noise (also known as jitter, I hate that name) is introduced.
Phase noise is bad.
Some of the phase noise from the S/P DIF input will get to the recovered clock output, no matter how good the S/P DIF receiver is.
Anton
I don't think TTL to optical to TTL is a good permanent solution. With the optical connection, there is a voltage threshold where the transmitter outputs light, and a luminance threshold where the receiver outputs a voltage. If these thresholds are even slightly different from transition to transition (and they will be), phase noise (also known as jitter, I hate that name) is introduced.
Phase noise is bad.
Some of the phase noise from the S/P DIF input will get to the recovered clock output, no matter how good the S/P DIF receiver is.
Anton
Hi, Fred,
The easiest way to confirm the type of audio data is to use an oscilloscope. If you don't have an o'scope, you can usually download the datasheet for the DAC in your device, then work out which audio standard is in use.
Apparently, the 970HD uses the Cirrus Logic CS4360 DAC. Here is the datasheet...
http://www.cirrus.com/en/pubs/proDatasheet/CS4360.pdf
Page 22 shows which pins select the data format, and page 23 shows the timing diagrams for each format - The audio data format can be selected by connecting the DIF0 and DIF1 pins to a high or low voltage during power-on ("high" will be 3.3V or 5V depending on the supply voltage / DVD player design).
Caution: don't connect these pins to anything yourself, as they are very likely set (soldered) by design!
The problem is, the audio format can also be selected by an external controller (CPU), and might be set by the player's CPU itself.
So, without using a scope, you could still connect the audio data signals of the Oppo's DAC to an external DAC and see if you get any audio...
You will need to make sure your external DAC is compatible with the signal voltages from the Oppo (ie. 3.3V or 5V). Also, turn the volume way down first! (incorrect data formats can produce extremely loud noise and could damage your speakers and hearing!
)
btw, since the CS4369 DAC in the Oppo is a 6-channel DAC, the front left / right audio data is probably on pin 4 (SDIN1) of the DAC. You will also need SCLK, LRCK, and possibly MCLK (if your external DAC doesn't have it's own clock.)
(If you need help, don't hesitate to ask.)
OK, the info below isn't strictly necessary, but I thought I'd add it anyway...
To the best of my knowledge (I'm still learning!), with "standard" I2S, there are usually 32 bits per channel which are clocked into the DAC (even if not all of those bits contain actual audio data.)
The problem is, some CD players / DACs "clock-in" 32 bits per channel (64fs), and some only "clock-in" 24 bits per channel (48fs). AFAIK, for I2S, 64fs is considered to be "standard I2S", but this is not always set in stone.
For I2S, either 16 or 24 of these bits actually contain the audio data in most things I've seen, the "empty" bits are generally ignored by the DAC. If you only input 16-bit audio data into a 24-bit capable DAC, there may be an option to set the DAC to ignore the "empty" 16 bits (in case any spurious data is received I suppose?)
Then there are the "left-justified" and "right-justified" formats... The main difference between left-justified and "standard" I2S is that I2S has a delay of one bit-clock cycle (BCK) before the audio data bits are input (you can see this on the timing diagram.)
Right-justified means that the bits are obviously shifted right (or delayed) so that the LSB of the audio data is always in the same place (directly before the next channel is clocked in.)
I personally prefer 64fs I2S because a lot of devices use this and it generally makes it easier to interface things. (I often play with DSP programming and Dolby / DTS decoders).
Again, this is just what I've learnt so far, so I might not be 100% correct, and I don't want to start a flame war over naming conventions etc.
OzOnE.
The easiest way to confirm the type of audio data is to use an oscilloscope. If you don't have an o'scope, you can usually download the datasheet for the DAC in your device, then work out which audio standard is in use.
Apparently, the 970HD uses the Cirrus Logic CS4360 DAC. Here is the datasheet...
http://www.cirrus.com/en/pubs/proDatasheet/CS4360.pdf
Page 22 shows which pins select the data format, and page 23 shows the timing diagrams for each format - The audio data format can be selected by connecting the DIF0 and DIF1 pins to a high or low voltage during power-on ("high" will be 3.3V or 5V depending on the supply voltage / DVD player design).
Caution: don't connect these pins to anything yourself, as they are very likely set (soldered) by design!
The problem is, the audio format can also be selected by an external controller (CPU), and might be set by the player's CPU itself.
So, without using a scope, you could still connect the audio data signals of the Oppo's DAC to an external DAC and see if you get any audio...
You will need to make sure your external DAC is compatible with the signal voltages from the Oppo (ie. 3.3V or 5V). Also, turn the volume way down first! (incorrect data formats can produce extremely loud noise and could damage your speakers and hearing!
)btw, since the CS4369 DAC in the Oppo is a 6-channel DAC, the front left / right audio data is probably on pin 4 (SDIN1) of the DAC. You will also need SCLK, LRCK, and possibly MCLK (if your external DAC doesn't have it's own clock.)
(If you need help, don't hesitate to ask.)
OK, the info below isn't strictly necessary, but I thought I'd add it anyway...
To the best of my knowledge (I'm still learning!), with "standard" I2S, there are usually 32 bits per channel which are clocked into the DAC (even if not all of those bits contain actual audio data.)
The problem is, some CD players / DACs "clock-in" 32 bits per channel (64fs), and some only "clock-in" 24 bits per channel (48fs). AFAIK, for I2S, 64fs is considered to be "standard I2S", but this is not always set in stone.
For I2S, either 16 or 24 of these bits actually contain the audio data in most things I've seen, the "empty" bits are generally ignored by the DAC. If you only input 16-bit audio data into a 24-bit capable DAC, there may be an option to set the DAC to ignore the "empty" 16 bits (in case any spurious data is received I suppose?)
Then there are the "left-justified" and "right-justified" formats... The main difference between left-justified and "standard" I2S is that I2S has a delay of one bit-clock cycle (BCK) before the audio data bits are input (you can see this on the timing diagram.)
Right-justified means that the bits are obviously shifted right (or delayed) so that the LSB of the audio data is always in the same place (directly before the next channel is clocked in.)
I personally prefer 64fs I2S because a lot of devices use this and it generally makes it easier to interface things. (I often play with DSP programming and Dolby / DTS decoders).
Again, this is just what I've learnt so far, so I might not be 100% correct, and I don't want to start a flame war over naming conventions etc.
OzOnE.
Hi,
@fredlock, no problem at all if you want to PM me. Again, I'm still picking things up all the time, but I'm glad to help.
@amc184, I was never sure that I2S always had the same number of bits per frame - When people say "I2S", is it generally accepted that it is always 64fs (32 bits per "frame")?
If so, then yes, I2S is I2S, and the other formats (left and right justified) are different and can also have different bits per frame / channel?
Thanks,
OzOnE.
@fredlock, no problem at all if you want to PM me. Again, I'm still picking things up all the time, but I'm glad to help.
@amc184, I was never sure that I2S always had the same number of bits per frame - When people say "I2S", is it generally accepted that it is always 64fs (32 bits per "frame")?
If so, then yes, I2S is I2S, and the other formats (left and right justified) are different and can also have different bits per frame / channel?
Thanks,
OzOnE.
If memory serves the D10 uses some little 8-pin Crystal DAC. Mine does anyway.
Interesting though - the mainboard in this D100 looks different to mine. Mine is smaller, and to be honest the assembly is really really poor, with cheap capacitors used and a 7808 regulator that gets very hot. That transformer isn't a toroid - if you pull at the copper can, it eventually comes off and you will find a small EI transformer underneath.
Interesting though - the mainboard in this D100 looks different to mine. Mine is smaller, and to be honest the assembly is really really poor, with cheap capacitors used and a 7808 regulator that gets very hot. That transformer isn't a toroid - if you pull at the copper can, it eventually comes off and you will find a small EI transformer underneath.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.