I've implemented Elso's comparator-input S/PDIF reciever as seen below except with 10nF input cap and 100k resistor to ground at the noninverting input. I am running it single-ended.
The comparator does pass the S/PDIF signal, but not cleanly. There is a lot of hash around the transitions. If I run it through a buffer it comes out clean. I'm using the usual rat's nest wiring in testing and I suspect that at least part of the problem is that the comparator is picking up and triggering on reflections. Still, this is not ideal. A circuit using 74HC04U triggers cleanly using identical wiring.
I tried connecting a pot tied high to the inverting input and biasing it above ground, but this was not fruitful. I continued to see hash up until the comparator turned off completely. Does this circuit need some hysterisis?
Some background on setup: This circuit is contructed using SOIC package comparator, on a ground plane, bypassed by 100nF and 10uF ceramics, and fed by on-board regulators powered by a lab supply. The rails look completely clean. The outputs are kept well away from the inputs.
Any ideas, Elso, others? Is this a problem with the circuit or with my implementation?
The comparator does pass the S/PDIF signal, but not cleanly. There is a lot of hash around the transitions. If I run it through a buffer it comes out clean. I'm using the usual rat's nest wiring in testing and I suspect that at least part of the problem is that the comparator is picking up and triggering on reflections. Still, this is not ideal. A circuit using 74HC04U triggers cleanly using identical wiring.
I tried connecting a pot tied high to the inverting input and biasing it above ground, but this was not fruitful. I continued to see hash up until the comparator turned off completely. Does this circuit need some hysterisis?
Some background on setup: This circuit is contructed using SOIC package comparator, on a ground plane, bypassed by 100nF and 10uF ceramics, and fed by on-board regulators powered by a lab supply. The rails look completely clean. The outputs are kept well away from the inputs.
Any ideas, Elso, others? Is this a problem with the circuit or with my implementation?
Attachments
Spdf
Is this a problem with the circuit or with my implementation?
Both.
"A circuit using 74HC04U triggers cleanly using identical wiring."
And will sound better in my experience. The SPDIF calls for a 0.5 volt level and you are driving it at logic levels and most likely overdriving the the hell out of the input of the comparator. Comparators are designed to get logic level out puts out of small input signals. Also excesive speed can cause triggering from noise on the interface. A limter amp with a 74HC04U followed by a
74HC86 phase spliter sounds much better in my experience. High speed comparators can cause even very experience designers fits.
For a little more background:http://www.epanorama.net/documents/audio/spdif.html
http://www.analog.com/library/analogDialogue/archives/34-07/comparators/index.html
http://www.analog.com/UploadedFiles/Application_Notes/56232956AN352.pdf
Is this a problem with the circuit or with my implementation?
Both.
"A circuit using 74HC04U triggers cleanly using identical wiring."
And will sound better in my experience. The SPDIF calls for a 0.5 volt level and you are driving it at logic levels and most likely overdriving the the hell out of the input of the comparator. Comparators are designed to get logic level out puts out of small input signals. Also excesive speed can cause triggering from noise on the interface. A limter amp with a 74HC04U followed by a
74HC86 phase spliter sounds much better in my experience. High speed comparators can cause even very experience designers fits.
For a little more background:http://www.epanorama.net/documents/audio/spdif.html
http://www.analog.com/library/analogDialogue/archives/34-07/comparators/index.html
http://www.analog.com/UploadedFiles/Application_Notes/56232956AN352.pdf
Comparator Question
Hi Fred and tiroth,
Fred, glad you are back!
I don't know what is wrong with my circuit.
I have received negative and very positive feedbacks on it.
To me it seems some CDP's have a lot of noise on the ground, making the trippoint for the comparator rather vague. In this case I would put a 2k pot at the inverting input. The pot connected through 820 Ohm resistors to both supply rails. Or use a transformer. Don't ask me which one is the best! Or add hysteresys as Fred suggests. Also note the comparator is a very high speed device that needs proper powersupplydecoupling as close as possible to the device. I got very clean SPDIF signals using my interface with Philips CD-650, Sony CDP-M27 and Sony CDP-X33ES.
Fred I am also glad you answered my old question: which circuit is better? I see you prefer the HCU04 and the 74HC86.
The comparator is not overdriven too much as the 5V max. from the output of the CDP is halved by the arrangement of the 75 Ohm resistors. 3V is the max. input on the comparator. Yes I have blown one up with 5V!
Hi Fred and tiroth,
Fred, glad you are back!
I don't know what is wrong with my circuit.
I have received negative and very positive feedbacks on it.
To me it seems some CDP's have a lot of noise on the ground, making the trippoint for the comparator rather vague. In this case I would put a 2k pot at the inverting input. The pot connected through 820 Ohm resistors to both supply rails. Or use a transformer. Don't ask me which one is the best! Or add hysteresys as Fred suggests. Also note the comparator is a very high speed device that needs proper powersupplydecoupling as close as possible to the device. I got very clean SPDIF signals using my interface with Philips CD-650, Sony CDP-M27 and Sony CDP-X33ES.
Fred I am also glad you answered my old question: which circuit is better? I see you prefer the HCU04 and the 74HC86.
The comparator is not overdriven too much as the 5V max. from the output of the CDP is halved by the arrangement of the 75 Ohm resistors. 3V is the max. input on the comparator. Yes I have blown one up with 5V!
Hmm...frustrating. I will probably cut the comparator out then and piggyback a small board to replace it. I figured comparator should be better than the 04U but experience says no.
Elso, my first thought was ground or rail bounce, but my bypass caps are right on the pads of the comparator, and the bypassing is well in excess of what should be required. I also tried varying the trip point without much success.
I think the comparator is just too fast and too sensitive. Jocko told me that, I should have listened.
Thanks for your input.
Elso, my first thought was ground or rail bounce, but my bypass caps are right on the pads of the comparator, and the bypassing is well in excess of what should be required. I also tried varying the trip point without much success.
I think the comparator is just too fast and too sensitive. Jocko told me that, I should have listened.
Thanks for your input.
This is the best circuit I have found to date:
http://www.fet.uni-hannover.de/~purnhage/dat/spdif.txt
http://www.fet.uni-hannover.de/~purnhage/dat/spdif.txt
Comparator Question
Hi tiroth,
I am still puzzled as the DAC side of my circuit also works with the Sony SCD-1, Sony CDP-XA30ES and Genesis Digital Lens SPDIF output, I just remember. (Not my own equipment BTW)
May I ask what CD-player you use?
And Jocko should have told me! We exchanged emails about the SPDIF interface!!
Elso/ not using SPDIF anymore....
Hi tiroth,
I am still puzzled as the DAC side of my circuit also works with the Sony SCD-1, Sony CDP-XA30ES and Genesis Digital Lens SPDIF output, I just remember. (Not my own equipment BTW)
May I ask what CD-player you use?
And Jocko should have told me! We exchanged emails about the SPDIF interface!!
Elso/ not using SPDIF anymore....
how then to drive cs8412 differentially?
I bought a comparator with a view to making Elso's circuit to drive my home-made dac. At present I am driving the cs8412 with a home-wound pulse transformer which Elso's circuit was meant to replace.
From the previous posts in this thread it seems that using the comparator is not as straightforward as it sounds. But how do you use the 74HC04U circuit to drive the 8412 differentially??
I bought a comparator with a view to making Elso's circuit to drive my home-made dac. At present I am driving the cs8412 with a home-wound pulse transformer which Elso's circuit was meant to replace.
From the previous posts in this thread it seems that using the comparator is not as straightforward as it sounds. But how do you use the 74HC04U circuit to drive the 8412 differentially??
Fred's Way/Circuit
Just look here.
http://www.diyaudio.com/forums/showthread.php?postid=90851#post90851
And the Pass DAC D1 for the idea with the 74HC86.
/ not using SPDIF anymore.....
Hi Philw,PhilW said:
Just look here.
http://www.diyaudio.com/forums/showthread.php?postid=90851#post90851
And the Pass DAC D1 for the idea with the 74HC86.
/ not using SPDIF anymore.....
SPDIF Interface
Hi tiroth,
I received the following feedback from Mike:
"
I finally tried your AD8561 SPDIF interface (just the receiver part of the circuit) in my TDA1543 non-oversampling (NOS) DAC and it worked great. There was quite an improvement in sound quality over having the SPDIF input connected through two caps directly to the CS8412. The CS8412 connects to four TDA1543s. The interface also solved a digital lock problem I had with my TDA1543 NOS DAC. Before adding your AD8561 SPDIF interface, I almost always had to wait 15 to 30 minutes for the DAC to warm up before it would lock on to the digital out from my transport. I'm also using Wildmonkeysect's loopfilter on the CS8412. I guess Jocko was right about the CS8412's Schmitt-trigger input dumping noise back into the SPDIF cable.
THANKS for the help.
"
So I believe it can work.
Hi tiroth,
I received the following feedback from Mike:
"
I finally tried your AD8561 SPDIF interface (just the receiver part of the circuit) in my TDA1543 non-oversampling (NOS) DAC and it worked great. There was quite an improvement in sound quality over having the SPDIF input connected through two caps directly to the CS8412. The CS8412 connects to four TDA1543s. The interface also solved a digital lock problem I had with my TDA1543 NOS DAC. Before adding your AD8561 SPDIF interface, I almost always had to wait 15 to 30 minutes for the DAC to warm up before it would lock on to the digital out from my transport. I'm also using Wildmonkeysect's loopfilter on the CS8412. I guess Jocko was right about the CS8412's Schmitt-trigger input dumping noise back into the SPDIF cable.
THANKS for the help.
"
So I believe it can work.
Using 74hc86 to drive 8412 differentially
>Just look here.
>http://www.diyaudio.com/forums/show...90851#post90851
>And the Pass DAC D1 for the idea with the 74HC86.
I looked, but as far as I can see the Pass DAC D1 uses a transformer coupled into the 8412 by two capacitors. The 74hc86 in the diagram is used for something completely different.
I am still no further on with the 74hc86 - how do we wire it to drive differentially please?
>Just look here.
>http://www.diyaudio.com/forums/show...90851#post90851
>And the Pass DAC D1 for the idea with the 74HC86.
I looked, but as far as I can see the Pass DAC D1 uses a transformer coupled into the 8412 by two capacitors. The 74hc86 in the diagram is used for something completely different.
I am still no further on with the 74hc86 - how do we wire it to drive differentially please?
Why not use a decent 75 Ohm pulse transformer at the input of CS8412 ?
The results with a transformer are very good and it is a simple solution compared to the electronics I see in your schematic. I tried 2 real "digital" transformer-types and they offer a clear improvement over direct coupling. Apart from the quality of the transformer not much can go wrong.
The only problem is finding decent transformers, I know
Wombat, I did some experiments with capaitors at the input of the Crystal and filmtypes ( MKP, MKT etc ) are not beneficial in this application. Best results were with Kemet X5R ceramics.
I once tried a Wima 100 nF MKT and there was audible distortion introduced. I did not care for any measurements as I wanted to listen so I changed it immediately for a good ceramic cap and everything was OK again.
Not a scientific approach but very effective
The results with a transformer are very good and it is a simple solution compared to the electronics I see in your schematic. I tried 2 real "digital" transformer-types and they offer a clear improvement over direct coupling. Apart from the quality of the transformer not much can go wrong.
The only problem is finding decent transformers, I know
Wombat, I did some experiments with capaitors at the input of the Crystal and filmtypes ( MKP, MKT etc ) are not beneficial in this application. Best results were with Kemet X5R ceramics.
I once tried a Wima 100 nF MKT and there was audible distortion introduced. I did not care for any measurements as I wanted to listen so I changed it immediately for a good ceramic cap and everything was OK again.
Not a scientific approach but very effective
Pulsetransformer for SPDIF
Yah, Jean-Paul,
Why not share your knowledge of what you think is a decent pulse transformer?
Fred rolls his own but did not want to share his design.
I addition I can report that my interface at the DAC side with the AD8561 comparator also works very well with the JK-Acoustics transport.http://www.jkacoustics.nl/Ref CD-transport.htm
It is a Philips CDM12pro.
jean-paul said:Why not use a decent 75 Ohm pulse transformer at the input of CS8412 ?
The results with a transformer are very good and it is a simple solution compared to the electronics I see in your schematic. I tried 2 real "digital" transformer-types and they offer a clear improvement over direct coupling. Apart from the quality of the transformer not much can go wrong.
The only problem is finding decent transformers, I know
Yah, Jean-Paul,
Why not share your knowledge of what you think is a decent pulse transformer?
Fred rolls his own but did not want to share his design.
I addition I can report that my interface at the DAC side with the AD8561 comparator also works very well with the JK-Acoustics transport.http://www.jkacoustics.nl/Ref CD-transport.htm
It is a Philips CDM12pro.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.