Having researched this, concensus is that dual DACs are better than balanced, although its all pretty subjective (Dynamics and soundstage improve, as reported by a number of DIY TDA1541 implementations). Single ended output is also more versatile, and the amp I'll use has unbalanced inputs. I'll look into a configuration that would allow you to do both dual & balanced switchable.
Regarding tubes, I definately want to use a tube output as an option. I'm comfortable with digital and solid state but have no experience with tubes, Jean-Paul can you nominate a design that you would like to see (buffer with sinx/x compensation).
Not sure about gold plating, elib is checking at the moment with the supplier about the costs, so I'll add that request to the list. I have seen silver rub on powder that can be applied after the board is done but I assume this is not the same sonically as gold plating?
Regards,
Dean
Regarding tubes, I definately want to use a tube output as an option. I'm comfortable with digital and solid state but have no experience with tubes, Jean-Paul can you nominate a design that you would like to see (buffer with sinx/x compensation).
Not sure about gold plating, elib is checking at the moment with the supplier about the costs, so I'll add that request to the list. I have seen silver rub on powder that can be applied after the board is done but I assume this is not the same sonically as gold plating?
Regards,
Dean
PCB layout
Brian,
The hardest part of this exercise is getting the PCB layout correct to ensure minimum ground loops and maximum noise performance. I'm happy for you to work on the PCB as I have just downloaded Eagle and still coming to grips with the package (eg. I add text to the schematic and it adds a small crosshair to the botton left of each line of text!!).
Regarding the schematic, would it be easier if I continued the schematic in Eagle & attach the files to this thread? I have all the schematics on paper or based on schematic snippets from others on this site and the Internet, but it will take me a week or so to create a complete design schematic.
Anyone interested in a USB receiver using the PCM2902?
Regards,
Dean
Brian,
The hardest part of this exercise is getting the PCB layout correct to ensure minimum ground loops and maximum noise performance. I'm happy for you to work on the PCB as I have just downloaded Eagle and still coming to grips with the package (eg. I add text to the schematic and it adds a small crosshair to the botton left of each line of text!!).
Regarding the schematic, would it be easier if I continued the schematic in Eagle & attach the files to this thread? I have all the schematics on paper or based on schematic snippets from others on this site and the Internet, but it will take me a week or so to create a complete design schematic.
Anyone interested in a USB receiver using the PCM2902?
Regards,
Dean
Re: PCB layout
As for the noise performance, I see how this is a challenge.
I can read Eagle schematic files. I have been using Protel to do layouts.
As for the PCM2902, I was just viewing Mark William's website:
http://www.overclockers.com.au/~mwp/dac3/
where he uses this chip in his dac, which he recently finished. I don't see that chip having any place in this project, but it is an interesting chip.
--
Brian
deandob said:
As for the noise performance, I see how this is a challenge.
I can read Eagle schematic files. I have been using Protel to do layouts.
As for the PCM2902, I was just viewing Mark William's website:
http://www.overclockers.com.au/~mwp/dac3/
where he uses this chip in his dac, which he recently finished. I don't see that chip having any place in this project, but it is an interesting chip.
--
Brian
Brian,
I want to use this DAC to connect to a computer audio source, so a USB input as well as SPDIF is an option I'd like to try. Anyway with the concept of a modular DAC you can choose what source input best suits needs, although I agree most people would not want to use a DAC this way. I downloaded the datasheet for the PCM2902 and the DAC component of the chip is not good enough, and it outputs SPDIF but not i2s, more research is needed (which I'll take offline).
Peter,
Do you use a transformer between the source & the DAC to provide better isolation that the caps can provide? Interesting that you find the caps sounded better as the signal is digital, maybe your transformer was not setup correctly?
Regards,
Dean.
I want to use this DAC to connect to a computer audio source, so a USB input as well as SPDIF is an option I'd like to try. Anyway with the concept of a modular DAC you can choose what source input best suits needs, although I agree most people would not want to use a DAC this way. I downloaded the datasheet for the PCM2902 and the DAC component of the chip is not good enough, and it outputs SPDIF but not i2s, more research is needed (which I'll take offline).
Peter,
Do you use a transformer between the source & the DAC to provide better isolation that the caps can provide? Interesting that you find the caps sounded better as the signal is digital, maybe your transformer was not setup correctly?
Regards,
Dean.
Peter Daniel said:
I am curious to know what effect using an digital isolator for the digital input, such as an optoisolator, or the new analog devices iCouplers.
"The ADuM1100 is a digital isolator based on Analog Devices’
iCoupler technology. Combining high speed CMOS and monolithic
air core transformer technology, this isolation component
provides outstanding performance characteristics superior to
alternatives such as optocoupler devices.
25 Mbps and 100 Mbps."
--
Brian
A high end DAC needs isolation from the source using either a transformer or another form of isolator that can pass 5Mhz - 10Mhz digital. The analog isolator Brian mentions looks to fit the bill.
In theory as long as the digital signal can pass without too much distortion to cause bit errors at the receiver, the type of isolation should not matter. Peter's experiments are interesting as they show that there might be other forces at work where he can hear a difference between different coupling methods.
Peter, what about ordering the Analog isolator Brian mentions above as a sample & try it out to see if it makes a difference.
In theory as long as the digital signal can pass without too much distortion to cause bit errors at the receiver, the type of isolation should not matter. Peter's experiments are interesting as they show that there might be other forces at work where he can hear a difference between different coupling methods.
Peter, what about ordering the Analog isolator Brian mentions above as a sample & try it out to see if it makes a difference.
Would it be possible to use a (good) CATV double galvanic isolator? Frequentie 5 to 1 GHz.
http://www.dkt.dk/download/pdf/14-2e Isolatorer.pdf
Regards,
Peter K
http://www.dkt.dk/download/pdf/14-2e Isolatorer.pdf
Regards,
Peter K
Peter K,
Your isolators may work and are convenient being in a small box with connectors and the SPDIF impedance is the same as TV (75 ohms). You may end up with larger losses at lower frequencies (SPDIF is 2 - 5 Mhz for 44Khz redbook) with these devices as their freq response is quoted as 5Mhz and up.
In another thread there was a discussion about using the SPDIF transformer at the source not the DAC so an external box like this might be the way forward. I have no idea what it would sound like, maybe Peter Daniels has a TV isolator lying around that he can connect to his DAC for listening tests.
Your isolators may work and are convenient being in a small box with connectors and the SPDIF impedance is the same as TV (75 ohms). You may end up with larger losses at lower frequencies (SPDIF is 2 - 5 Mhz for 44Khz redbook) with these devices as their freq response is quoted as 5Mhz and up.
In another thread there was a discussion about using the SPDIF transformer at the source not the DAC so an external box like this might be the way forward. I have no idea what it would sound like, maybe Peter Daniels has a TV isolator lying around that he can connect to his DAC for listening tests.
PCB
I'm still working on the DAC schematic (or maybe not getting time to work on it is more accurate) as there is a reasonable amount of complexity with all the features discussed here. I'll have more time on the weekend.
There is another thread for a TDA1543 design & PCB where we may be able to use parts of the design & PCB layout to save re-inventing the wheel.
I'm still working on the DAC schematic (or maybe not getting time to work on it is more accurate) as there is a reasonable amount of complexity with all the features discussed here. I'll have more time on the weekend.
There is another thread for a TDA1543 design & PCB where we may be able to use parts of the design & PCB layout to save re-inventing the wheel.
Deja Vu
The SPDIF interface really needs a bandwidth of about 50KHz to at least 20MHz to 50Mhz.
The schematic at:
http://www.ndh.net/home/kboehm/T-DAC-P2.htm
is illegible for the most part. I did see a transformer for a 75 ohm interface loaded with what appears to be 100 ohms.
The iCoupler won't work since the input voltage is TTL levels I believe and SPDIF is smaller than this. Even if you used a big enough signal the results would be very likely to be poor sonically. The conversion rates for passing a signal through an air core transformer spell RFI issues in an interface that is incredibly sensitive to RF noise. The use of a Schmitt input on this interface
tends to launch noise into this interface and often cause signal reflections and poor sonics. I think some research into the basics of how the SPDIF interface works is going to be required before you "improve it."
The level of understanding I see here for the most part, will not even get a signal trough much less a good sounding one. BTW this is a very twitchy interface to get the best sonics out of and took Jocko and I over a year ((working seperately)t o get to sound good and over 2 years to sound really good and we both knew enough RF design to know what we were up against. Yes, you can hear the difference between PP caps from different manufactures in the interface,
but I never like the MITs for this.
The SPDIF interface really needs a bandwidth of about 50KHz to at least 20MHz to 50Mhz.
The schematic at:
http://www.ndh.net/home/kboehm/T-DAC-P2.htm
is illegible for the most part. I did see a transformer for a 75 ohm interface loaded with what appears to be 100 ohms.
The iCoupler won't work since the input voltage is TTL levels I believe and SPDIF is smaller than this. Even if you used a big enough signal the results would be very likely to be poor sonically. The conversion rates for passing a signal through an air core transformer spell RFI issues in an interface that is incredibly sensitive to RF noise. The use of a Schmitt input on this interface
tends to launch noise into this interface and often cause signal reflections and poor sonics. I think some research into the basics of how the SPDIF interface works is going to be required before you "improve it."
The level of understanding I see here for the most part, will not even get a signal trough much less a good sounding one. BTW this is a very twitchy interface to get the best sonics out of and took Jocko and I over a year ((working seperately)t o get to sound good and over 2 years to sound really good and we both knew enough RF design to know what we were up against. Yes, you can hear the difference between PP caps from different manufactures in the interface,
but I never like the MITs for this.
Re: Deja Vu
Well, if I spent two years improving it, I would probably find better caps as well. But I only dedicated one evening for that, using whatever I had available.
Some additional info (never mind the cats): http://www.diyaudio.com/forums/showthread.php?s=&postid=27144#post27144
Fred Dieckmann said:
Well, if I spent two years improving it, I would probably find better caps as well. But I only dedicated one evening for that, using whatever I had available.
Some additional info (never mind the cats): http://www.diyaudio.com/forums/showthread.php?s=&postid=27144#post27144
TDA1541A
I had been trying to buy a single TDA1541A for several weeks, with no luck. Anyway, I ended up buying 400 of them to get the one I wanted. I now have 399 that are available for sale. I hope it is OK to post such a message here. If anybody is interested, send me an e-mail. I'm not quite sure how much these are worth, so if anybody has an idea, please let me know. I know I've seen these for over 20.00, but I don't think that is reasonable. These are the plain part, no crowns.
Thanks,
jon@cusackmusic.com
I had been trying to buy a single TDA1541A for several weeks, with no luck. Anyway, I ended up buying 400 of them to get the one I wanted. I now have 399 that are available for sale. I hope it is OK to post such a message here. If anybody is interested, send me an e-mail. I'm not quite sure how much these are worth, so if anybody has an idea, please let me know. I know I've seen these for over 20.00, but I don't think that is reasonable. These are the plain part, no crowns.
Thanks,
jon@cusackmusic.com
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.