belgianbrain
Super Member
Yep. Visually, they are very convincing. I was optimistic when I saw them. But 0 Ohms from pin 1 to pin 5.
Next project: Adcom GDA-700
- Thread starter belgianbrain
- Start date
belgianbrain
Super Member
Got the DF1704 and dropped it in. So far, no sound. BUT, there are some configuration settings. Rudimentary data sheet that was supplied is attached.
Default for the replacement kit is 16 bit standard input format, 20 bit output format, and local reset.
Any of you knowledgeable individuals know if this is how the GDA-700 is working?
Default for the replacement kit is 16 bit standard input format, 20 bit output format, and local reset.
Any of you knowledgeable individuals know if this is how the GDA-700 is working?
SoNic67
Well-Known Member
I can provide a link for the PMD100 datasheet and you can decipher what format was used in your GDA, based on the config pins (grounded or V+).
http://vt52.com/ftp/semi/pmd100.pdf
http://vt52.com/ftp/semi/pmd100.pdf
Last edited:
belgianbrain
Super Member
Thank you. Will be a big help.
Looks like I need to change the 1704 from 16 bit right justified input to 24 bit left justified.
Stay tuned.
Looks like I need to change the 1704 from 16 bit right justified input to 24 bit left justified.
Stay tuned.
Last edited:
belgianbrain
Super Member
Made the change.
Now I have sound, but it is very staticy like an FM station that is not tuned right.
Everything looks good to me now, except I note the PMD100 had a dither input connected to the PLC next to it, and the DF1704 has no such input.
Now I have sound, but it is very staticy like an FM station that is not tuned right.
Everything looks good to me now, except I note the PMD100 had a dither input connected to the PLC next to it, and the DF1704 has no such input.
Last edited:
belgianbrain
Super Member
Good question.
Well, the original PMD100 appeared to be set at Left justified input mode with 20 bit output.
The DF1704 has selectable input modes. I tried them all. Anything right justified just produces white noise. 24 bit left justified input works best, but with static and distortion. I2S input modes also work, but with more static and distortion.
I'm a bit stuck at the moment. The PMD100 has a PLC to the left of it inbetween it and the digital receiver and and it is not entirely clear to me what this is doing.
I also cannot find input receiver Ultra Analog AES 21 AESC003 datasheet. Anyone have one?
Also, the PMD100 had a dither input connected to the PLC output. The DF1704 has no dither input, so this is now not considered by the DF1704.
Any ideas?
Well, the original PMD100 appeared to be set at Left justified input mode with 20 bit output.
The DF1704 has selectable input modes. I tried them all. Anything right justified just produces white noise. 24 bit left justified input works best, but with static and distortion. I2S input modes also work, but with more static and distortion.
I'm a bit stuck at the moment. The PMD100 has a PLC to the left of it inbetween it and the digital receiver and and it is not entirely clear to me what this is doing.
I also cannot find input receiver Ultra Analog AES 21 AESC003 datasheet. Anyone have one?
Also, the PMD100 had a dither input connected to the PLC output. The DF1704 has no dither input, so this is now not considered by the DF1704.
Any ideas?
belgianbrain
Super Member
Thinking about connecting the DOL and DOR of the DF1704 directly to the 1702s and removing the PLC from the path. Considering BCKO and WCKO goes directly to the 1702s, there seems to be no reason for the data signals to go through the PLC.
belgianbrain
Super Member
Interesting. The Parasound PAD2000, which I believe this upgrade is known to work fine on, is designed almost exactly the same as GDA-700 with same chips and everything (except it uses a PCM63).
Looks like ultra analog designed the PAD2000 and Adcom used almost the exact same design.
Maybe time to check all my connections first.
UPDATE: Found a broken trace on my de-emphasis line. Got so excited. Repaired it....and....no difference.
Looks like ultra analog designed the PAD2000 and Adcom used almost the exact same design.
Maybe time to check all my connections first.
UPDATE: Found a broken trace on my de-emphasis line. Got so excited. Repaired it....and....no difference.
Last edited:
belgianbrain
Super Member
Well, I connected the 1704 DOL and DOR directly to the PCM1702 chips. Output connection between DF and D/A is now direct and doesn't concern anything else.
I still have exactly the same problem. Static and distortion in music. I have tried all input modes of the DF1704 and 24 bit left justified is the one that works best.
AES21 worked fine until I tried the DF swap. But I'm still flying blind not knowing what it outputs and not having a datasheet
I still have exactly the same problem. Static and distortion in music. I have tried all input modes of the DF1704 and 24 bit left justified is the one that works best.
AES21 worked fine until I tried the DF swap. But I'm still flying blind not knowing what it outputs and not having a datasheet
belgianbrain
Super Member
I have sent the seller several emails. All I got was a response saying "we have not previously had anyone use this in a GDA-700" . Seller has not been helpful so far, but it is a holiday weekend so I will give him a few more days before drawing any conclusions about their quality of service...
I also sent out emails to Adcom, Parasound, and Wadia asking if they have datasheets....
I also sent out emails to Adcom, Parasound, and Wadia asking if they have datasheets....
belgianbrain
Super Member
SO, it seems AES21 is just a CS8412 with a PLL implemented in front of it. So, you can learn a lot about AES21 by looking at CS8412 data sheet.
CS8412 was a very popular receiver chip and seems to be implemented so at the DF, DIN should be sampled on the falling edge of BCKIN.
BUT, DF1704 (made by same Burr Brown company) is configured to sample DIN on the rising edge of BCKIN and I don't think you can change this.
As far as I can tell now, I need to break the data clock trace and implement an inverter.
CS8412 was a very popular receiver chip and seems to be implemented so at the DF, DIN should be sampled on the falling edge of BCKIN.
BUT, DF1704 (made by same Burr Brown company) is configured to sample DIN on the rising edge of BCKIN and I don't think you can change this.
As far as I can tell now, I need to break the data clock trace and implement an inverter.
SoNic67
Well-Known Member
I think the DF-PMD module was supposed to work directly:
http://audiotuning.de/english/index.html
http://audiotuning.de/english/index.html
Last edited:
belgianbrain
Super Member
I would think a manufacturer would design its own receiver to work directly with its own DF without requiring intervening circuitry. But, that's just me maybe...
Anyway, got it working!
Found a National DM74LS86N Quad 2Input Exclusive OR gate chip here and configured it to work as an inverter and invert the bitclock. Works!
From initial listening, YES, the DF1704 absolutely murders the PMD100. They're not even in the same league.
Also added some OSCON caps to the power supplies on the AES21 and DF1704 while I was at it.
Stay tuned for more. This thing is sounding VERY good now and definitely fighting with the GDA-600 for the crown.
Anyway, got it working!
Found a National DM74LS86N Quad 2Input Exclusive OR gate chip here and configured it to work as an inverter and invert the bitclock. Works!
From initial listening, YES, the DF1704 absolutely murders the PMD100. They're not even in the same league.
Also added some OSCON caps to the power supplies on the AES21 and DF1704 while I was at it.
Stay tuned for more. This thing is sounding VERY good now and definitely fighting with the GDA-600 for the crown.
Last edited:
belgianbrain
Super Member
GDA-700 sounding VERY, VERY good now.
Compared to GDA-600, I'd say it has a tad less bass, but a bigger and more refined sound stage than the 600. Right now, I'm starting to prefer my refined 700.
This is with OPA827 installed. Will try some 627 soon to compare.
Compared to GDA-600, I'd say it has a tad less bass, but a bigger and more refined sound stage than the 600. Right now, I'm starting to prefer my refined 700.
This is with OPA827 installed. Will try some 627 soon to compare.
belgianbrain
Super Member
Got 6 OPA627AP to try. Two have same lot code and measure 52K from pin 1 to pin 5. Other four have same lot code and measure 68k from pin 1 to pin 5.
