bpetrus said:I just checked the audiochips.net site and they are listed as sold outAny other sources?
yeah, and they aren't going to get more, it seems:
"ATTENTION: AUDIOCHIPS.NET will be closing it's doors affective once all inventory is sold. It's been fun... happy DIY'ing."
Close call - I just got two of these from Audiochips no more than 6 weeks ago
Peter
Has anyone compared the CS8414 to the CS8412 in this application?some observations on 3D TDA1543
I laid out a TDA1543 DAC in the 3D layout that is the theme of this forum. As I usually use the SMD versions of TDA 1543 and of CS8412, this DAC was a bit larger and actually easier to build than my previous projects.
Rather than go through all the details, here are some observations:
1. Bending down pins 1 (BCK) and 2 (WS) of TDA 1543 and bending up pins 11 (FSYNC) and 12 (SCK) of CS8412 allowed these pins to directly touch, and eliminated need for jumper wires. Also, the CS 8412 pins that are grounded were bent upwards and soldered directly to the foil ground plane.
2. For the ground plane I had best results with copper foil extracted from a SOLO air core inductor. (I did not like copper tape at all.) Also, the placement of the different supply grounds with respect to the signal grounds is important in reducing noise.
3. For the jumper connection between SDATA (pin 26 of CS8412) and DATA (pin 3 of TDA 1543) I had best results with 26ga solid core 6N OFC unshielded. This jumper is exposed to a lot of RFI and the sonics are affected by the geometry and material of this jumper: to prove it to myself I used a foil jumper between these two pins: the DAC just hissed and roared.
4. I did not have AN800X regs on hand, so I decided to use the MC78L05ACP and MC78L08ACP regs in TO-92 package from Motorola (I use SOT-89 for my SMD projects), with the Vout pins connected directly to the Vin pins of the respective ICs (actually connected to the 0612 decoupling caps soldered directly to the supply pins and ground points of the ICs). This elimated the need for ferrites and inductors between the reg and the supply pins. The regs do heat up a little, but not excessively with 12V supply. They also have bad PSR rejection, and are not that smooth sonically, but I liked being able to route them so close to the supply pins.
5. I played with PLL filter RC values and finally settled (after reading some posts by kittaylor) on 100R and 4.7uF combination rather than the bog-standard 1k/.047uF combination recommended in the data sheet. The reasons for these values are detailed in kittaylor's thread on pink fish media website.
In comparison with "2D" TDA 1543 DAC with double-sided pcb and much more elaborate Sulzer regs and much more complex PSU, this 3D unit is more direct and up front, and the dynamics are actually better, despite use of conservative data sheet standard Vref and I/V values. I think that it is a good layout if you are using BGs (as I am) since BGs tend to be a little soft on percussion, and this layout adds a bite and attack to percussion instruments that balances out the softness of BGs.
Last thing: I did an I/V resistor shoot-out on this DAC, and my top 5 list (with winner at top) is the following:
1. Caddock TF020
2. Riken
3. Vishay TX2352
4. Caddock MK-132
5. NOS Holco
The ranking is subjective, of course, but if your layout can tolerate the large size of the TF020s, I would try them out.
I laid out a TDA1543 DAC in the 3D layout that is the theme of this forum. As I usually use the SMD versions of TDA 1543 and of CS8412, this DAC was a bit larger and actually easier to build than my previous projects.
Rather than go through all the details, here are some observations:
1. Bending down pins 1 (BCK) and 2 (WS) of TDA 1543 and bending up pins 11 (FSYNC) and 12 (SCK) of CS8412 allowed these pins to directly touch, and eliminated need for jumper wires. Also, the CS 8412 pins that are grounded were bent upwards and soldered directly to the foil ground plane.
2. For the ground plane I had best results with copper foil extracted from a SOLO air core inductor. (I did not like copper tape at all.) Also, the placement of the different supply grounds with respect to the signal grounds is important in reducing noise.
3. For the jumper connection between SDATA (pin 26 of CS8412) and DATA (pin 3 of TDA 1543) I had best results with 26ga solid core 6N OFC unshielded. This jumper is exposed to a lot of RFI and the sonics are affected by the geometry and material of this jumper: to prove it to myself I used a foil jumper between these two pins: the DAC just hissed and roared.
4. I did not have AN800X regs on hand, so I decided to use the MC78L05ACP and MC78L08ACP regs in TO-92 package from Motorola (I use SOT-89 for my SMD projects), with the Vout pins connected directly to the Vin pins of the respective ICs (actually connected to the 0612 decoupling caps soldered directly to the supply pins and ground points of the ICs). This elimated the need for ferrites and inductors between the reg and the supply pins. The regs do heat up a little, but not excessively with 12V supply. They also have bad PSR rejection, and are not that smooth sonically, but I liked being able to route them so close to the supply pins.
5. I played with PLL filter RC values and finally settled (after reading some posts by kittaylor) on 100R and 4.7uF combination rather than the bog-standard 1k/.047uF combination recommended in the data sheet. The reasons for these values are detailed in kittaylor's thread on pink fish media website.
In comparison with "2D" TDA 1543 DAC with double-sided pcb and much more elaborate Sulzer regs and much more complex PSU, this 3D unit is more direct and up front, and the dynamics are actually better, despite use of conservative data sheet standard Vref and I/V values. I think that it is a good layout if you are using BGs (as I am) since BGs tend to be a little soft on percussion, and this layout adds a bite and attack to percussion instruments that balances out the softness of BGs.
Last thing: I did an I/V resistor shoot-out on this DAC, and my top 5 list (with winner at top) is the following:
1. Caddock TF020
2. Riken
3. Vishay TX2352
4. Caddock MK-132
5. NOS Holco
The ranking is subjective, of course, but if your layout can tolerate the large size of the TF020s, I would try them out.
dac component sources
ebay user panda_99 has a bunch of DAC parts. CS8412, CS8414, TDA1541A-R1, TDA1543, TDA1545A at pretty reasonable rates. CS8412 is available for around $10, 1543 for $4. International shipping for any number of chips is $5. He ships out of Taiwan. If you don't find them in his listings don't despair; he usually relists in a day or two.
I just bought some of these off him, he's very prompt with email responses and shipping. Still haven't got them since my chips will be taking the scenic route round the world, with a friend visiting the US playing courier. I'm paranoid about the stuff getting stuck or stolen by Indian customs/post office. Of course you hear bad things about USPS too... you wouldn't want your stuff to get into Newman's hands, now, would you? Luckily he operates in NY
bpetrus said:I just checked the audiochips.net site and they are listed as sold out
ebay user panda_99 has a bunch of DAC parts. CS8412, CS8414, TDA1541A-R1, TDA1543, TDA1545A at pretty reasonable rates. CS8412 is available for around $10, 1543 for $4. International shipping for any number of chips is $5. He ships out of Taiwan. If you don't find them in his listings don't despair; he usually relists in a day or two.
I just bought some of these off him, he's very prompt with email responses and shipping. Still haven't got them since my chips will be taking the scenic route round the world, with a friend visiting the US playing courier. I'm paranoid about the stuff getting stuck or stolen by Indian customs/post office. Of course you hear bad things about USPS too... you wouldn't want your stuff to get into Newman's hands, now, would you? Luckily he operates in NY
In order to have some reference, I got hold of ML31.5 Reference transport and ML35 DAC. The transport is from 1996, the DAC is dated 1992. It's using CS8412 receiver, PMD100 HDCD filter and Ultra-Analog dual DACs with discreet output stage.
I was pretty much impressed by the transport, as it seems to be another level above my CD-Pro implementation. The bass notes are better defined, rock solid, with more control and overall it provides more musicality.
I was pretty much impressed by the transport, as it seems to be another level above my CD-Pro implementation. The bass notes are better defined, rock solid, with more control and overall it provides more musicality.
Attachments
I was not impressed by the DAC initially, but later I used better support (dedicated platform) and warming up helped definitely. I got better results using Tosling than RCA connection (and I expect balanced line to be even better).
I spent whole afternoon yesterday comparing ML DAC and my battery operated NOS implementation od TDA1543.
I was using ML380S for switching and both DACs were connected to the preamp using Kimber Silver Streak cable. For amp connection, I used an interconnect sent to me by Steve Eddy, and it appears to be very good (and I prefer it now to Silver Streak).
My DAC is using passive output, with Riken resistors.
The was basically no difference in volume level between two DACS.
I was pretty much surprised with the outcome. The bass on TDA DAC was actually better, deeper with more punch. Overall tonality was very close. ML DAC seemed to have a bit etched highs, with more analytical signature, but overall it was very listenable, but I preferred it only on some selected material.
The non oversampling TDA wasn't at all any inferior to ML dac. It had it's own character, which could be characterised by more smoothness and more analog sound. It also seemed to be a bit more natural. It was actually better on more complex recordings (I preferred ML dac on some simple, female vocal recordings, where not too many instruments were involved).
I could sense that the small DAC could be improved by using better coupling caps (I had BG N 4.7u) as clarity wasn't the best (comparing to ML DAC) and could be definitely improved.
All in all, I had pretty hard time picking the better DAC, but eventually decided on nos TDA1543. It sounded less digital and was more forgiving to less properly recorded material.
Of course both DACs could be still improved: ML could be tweaked and better passive parts installed (sa well as FIFO buffer as seen on later designs), TDA1543 is still not finalized, as I'm still experimenting here.
However since it was my first time listening to actually good oversampling DAC, I now believe that oversampling can bring great results when properly implemented.
I was also nicely surprsed that those two DACs were not that far apart sonically
Here's the inside pic of ML dac.
I spent whole afternoon yesterday comparing ML DAC and my battery operated NOS implementation od TDA1543.
I was using ML380S for switching and both DACs were connected to the preamp using Kimber Silver Streak cable. For amp connection, I used an interconnect sent to me by Steve Eddy, and it appears to be very good (and I prefer it now to Silver Streak).
My DAC is using passive output, with Riken resistors.
The was basically no difference in volume level between two DACS.
I was pretty much surprised with the outcome. The bass on TDA DAC was actually better, deeper with more punch. Overall tonality was very close. ML DAC seemed to have a bit etched highs, with more analytical signature, but overall it was very listenable, but I preferred it only on some selected material.
The non oversampling TDA wasn't at all any inferior to ML dac. It had it's own character, which could be characterised by more smoothness and more analog sound. It also seemed to be a bit more natural. It was actually better on more complex recordings (I preferred ML dac on some simple, female vocal recordings, where not too many instruments were involved).
I could sense that the small DAC could be improved by using better coupling caps (I had BG N 4.7u) as clarity wasn't the best (comparing to ML DAC) and could be definitely improved.
All in all, I had pretty hard time picking the better DAC, but eventually decided on nos TDA1543. It sounded less digital and was more forgiving to less properly recorded material.
Of course both DACs could be still improved: ML could be tweaked and better passive parts installed (sa well as FIFO buffer as seen on later designs), TDA1543 is still not finalized, as I'm still experimenting here.
However since it was my first time listening to actually good oversampling DAC, I now believe that oversampling can bring great results when properly implemented.
I was also nicely surprsed that those two DACs were not that far apart sonically
Here's the inside pic of ML dac.
Attachments
I have a version of this with AD8561 comparator, single AD817 as I/V and buffer, and LM317 regs.
http://pinkfishmedia.net/forum/showthread.php?s=&threadid=4869
Great sound for not a lot of money. Open, good timing, very visceral and dynamic, clear and slightly dry, with somewhat soft tonality that I haven't heard in commercial gear but has a positive effect.
The AD817 seems an underappreciated opamp, give it a go if you're in the mood for experiemntation.
http://pinkfishmedia.net/forum/showthread.php?s=&threadid=4869
Great sound for not a lot of money. Open, good timing, very visceral and dynamic, clear and slightly dry, with somewhat soft tonality that I haven't heard in commercial gear but has a positive effect.
The AD817 seems an underappreciated opamp, give it a go if you're in the mood for experiemntation.
Peter:
you picked a pretty high-caliber OS DAC to compare with your project! With ML DAC, I was impressed by pinpoint, spatial presentation, and clarity of mids.
"I was pretty much surprised with the outcome. The bass on TDA DAC was actually better, deeper with more punch."
I also noticed precisely this superiority of 3D NOS TDA 1543 to other DACS (Theta, Parasound, CEC) besides ML.
"I could sense that the small DAC could be improved by using better coupling caps (I had BG N 4.7u) as clarity wasn't the best (comparing to ML DAC) and could be definitely improved."
Lots of factors could contribute to a lack of clarity, but : What coupling caps are more desirable for the NOS 1543 DAC in your opinion?
"Of course both DACs could be still improved: ML could be tweaked and better passive parts installed (sa well as FIFO buffer as seen on later designs)"
Agreed, overall, the passive parts in most ML equipment are not that impressive.
"TDA1543 is still not finalized, as I'm still experimenting here."
With what?
Kittaylor:
"single AD817 as I/V and buffer"
"The AD817 seems an underappreciated opamp, give it a go if you're in the mood for experiemntation."
How are the sonics different with AD 817 as opposed to passive I/V?
Has anyone rigged up this design "double mono" with two TDA1543 for that extra 3dB gain? Please describe ...
Best regards
you picked a pretty high-caliber OS DAC to compare with your project! With ML DAC, I was impressed by pinpoint, spatial presentation, and clarity of mids.
"I was pretty much surprised with the outcome. The bass on TDA DAC was actually better, deeper with more punch."
I also noticed precisely this superiority of 3D NOS TDA 1543 to other DACS (Theta, Parasound, CEC) besides ML.
"I could sense that the small DAC could be improved by using better coupling caps (I had BG N 4.7u) as clarity wasn't the best (comparing to ML DAC) and could be definitely improved."
Lots of factors could contribute to a lack of clarity, but : What coupling caps are more desirable for the NOS 1543 DAC in your opinion?
"Of course both DACs could be still improved: ML could be tweaked and better passive parts installed (sa well as FIFO buffer as seen on later designs)"
Agreed, overall, the passive parts in most ML equipment are not that impressive.
"TDA1543 is still not finalized, as I'm still experimenting here."
With what?
Kittaylor:
"single AD817 as I/V and buffer"
"The AD817 seems an underappreciated opamp, give it a go if you're in the mood for experiemntation."
How are the sonics different with AD 817 as opposed to passive I/V?
Has anyone rigged up this design "double mono" with two TDA1543 for that extra 3dB gain? Please describe ...
Best regards
Slawney
You can read about my ongoing adventures with the AD817 here:
http://pinkfishmedia.net/forum/showthread.php?s=&threadid=7385
In short, AD817 is more open and has wider dynamic range, though this is with my speakers very far apart to make sure the sound is not too fierce or disorderly. Almost touching the sidewalls in my case.
Passive I/V (with 1% 1206 SMD resistors) sounds somewhat compressed. Not inherently a bad thing, as I can get a satifyingly visceral sound at lower volumes. It's also a bit thicker and warmer.
You can read about my ongoing adventures with the AD817 here:
http://pinkfishmedia.net/forum/showthread.php?s=&threadid=7385
In short, AD817 is more open and has wider dynamic range, though this is with my speakers very far apart to make sure the sound is not too fierce or disorderly. Almost touching the sidewalls in my case.
Passive I/V (with 1% 1206 SMD resistors) sounds somewhat compressed. Not inherently a bad thing, as I can get a satifyingly visceral sound at lower volumes. It's also a bit thicker and warmer.
slawney said:Lots of factors could contribute to a lack of clarity, but : What coupling caps are more desirable for the NOS 1543 DAC in your opinion?
With what?
Actually the clarity is not really big problem, the sound is very clean a properly defined, it's only when comparing with oversampling ML DAC it seems to be a bit less. But ML seems to be etched and slightly digital sounding, that may give false impression of clarity. As good as BG caps are, I still think that there are better film caps out there. The big size of those caps is an inconvenience in that type of design, but I very much like Siemens MKV caps. They sound better to me than any other I tried so far, including Hovlands, MIT RTX and Jensens Copper foil. An interesting caps comparison can be found here: http://home.zonnet.nl/geenius/Cap.html and it seems like Mundirf Supreme Silver are the ones to go with
I never tried them though.As to the passive I/V, it didn't lack in any respect to what ML DAC offered with it's discreet output buffer (same buffer as used on ML30.6). Actually, dynamics and bass performance seemed better on my small DAC and ML DAC seemed thin at times (by comparison).
What I plan on improving is revised layout, getting rid of ferrite beads, using better caps for battery bapass, better input coupling caps and better overall bypass scheme. I also want to try those bigger Caddocks resistors (what value are you using here?)
Did you use input buffer chip?
I assume both of you are running your DACs at 8V?
Peter,
Bottom line: Mundorfs are definitely worth trying.
I use MCaps ZN with my DDDAC and I found them to integrate best into my system. They are very detailed and have excellent treble resolution. The BG N types sounded definitely more compressed and ...well... 'tubifying' (we did a cap shootout with this DAC at one point, the ZN being the clear winner). I guess it all depends on the system/room/...-context.Peter Daniel said:
[...] As good as BG caps are, I still think that there are better film caps out there. The big size of those caps is an inconvenience in that type of design, but I very much like Siemens MKV caps. They sound better to me than any other I tried so far, including Hovlands, MIT RTX and Jensens Copper foil. An interesting caps comparison can be found here: http://home.zonnet.nl/geenius/Cap.html and it seems like Mundirf Supreme Silver are the ones to go with
Bottom line: Mundorfs are definitely worth trying.
I only used MCap Supremes once, as XO caps in the Epos ES11. In that application they managed to gel the treble (slightly sizzling before) nicely with the already fine midband. Very detailed, but without any harshness.
I haven't heard the Supremes Silver/Oil yet, and I'd probably be quite reluctant to try them as they are twice as expensive as the 'normal' Supremes.
I haven't heard the Supremes Silver/Oil yet, and I'd probably be quite reluctant to try them as they are twice as expensive as the 'normal' Supremes.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.