Hurtig said:
Being an engineer, I believe measurements do indeed matter, and I spend a great deal measuring and simulating. With that said however, I also avoid feedback if possible, simply because I have found it to sound better. And for the same reason I prefer the AK4396 or WM8740 over the CS4398, they sound better.
It is also interesting to note that you chose a no feedback analog stage becuase it sounds better, yet you chose the CS4398 because it had the lowest THD. A bit contradictory don't you think?
Hurtig said:
Actually, I have talked to a representative at Crystal about the WM8804, and while he unofficially acknowledged that it probably was league above the CS8416, the official response was very similar to yours, that it is a well proven device with a proven track record.
You also have to remember that the CS8416 is old. It is pretty much the same design as the CS8412, which was introduced in 1998. Back then it was very good, the only competitor at the time was YM3623B, which could have as much as 2 ns of jitter (truth be told there was also the SAA7274, which is probably the best SPDIF receiver ever made, although it went out of production almost before it started).
Hurtig said:
Of course you don't have to try other DACs. If you are happy with the performance you are getting now then by all means leave it as it is. However if you are still looking to improve it you should try the Wolfson DACs, or better yet, the new AKM DACs. Trust me on this, those Scotts and Japs know what they are doing.
Hurtig said:
The WM8741 and WM8742 are their two top DACs, and the difference between them should be quite small. The people at Wolfson are very firendly, I do not think you will have any problems getting samples of the WM8804 or any DAC of your choice.
Re: what is the intent of this discussion? schematics, schematics, schematics
Hallo Suchtgutenklang!
As you see I´ve uploaded a few more shots in high rez for you to study. The scematics are not yet complete, mostly because we did a lot of alterations to the construction during develpment, so having updated scematics on the fly, we felt were a bit luxuary. We still did not create the final scematics yet simply.
Btw. I can tell you that the groundplane on the buttom also is used as cooling for the shunts, though we in addition to that guides the heat further down into the buttom of the aluminium casing. This makes a nice and cosy case to touch.
There are still a few small matters to deal with in the decoupling of the PSU´s for the chips.
At this point we i.e. chose to use just film capacitors in decoupling the analog section in the CS4398. We tried to use ceramics, but with no go. Any suggestions besides the Wima cap chosen?
Mostly the whole construction avoids bypass caps, instead we chose to use caps suited for the job where possible, or we did construct us out of the need for caps.
I may call myself a caphater I think.
suchtgutenklang said:
Hallo Suchtgutenklang!
As you see I´ve uploaded a few more shots in high rez for you to study. The scematics are not yet complete, mostly because we did a lot of alterations to the construction during develpment, so having updated scematics on the fly, we felt were a bit luxuary
. We still did not create the final scematics yet simply.Btw. I can tell you that the groundplane on the buttom also is used as cooling for the shunts, though we in addition to that guides the heat further down into the buttom of the aluminium casing. This makes a nice and cosy case to touch.
There are still a few small matters to deal with in the decoupling of the PSU´s for the chips.
At this point we i.e. chose to use just film capacitors in decoupling the analog section in the CS4398. We tried to use ceramics, but with no go. Any suggestions besides the Wima cap chosen?
Mostly the whole construction avoids bypass caps, instead we chose to use caps suited for the job where possible, or we did construct us out of the need for caps.
I may call myself a caphater I think
.Thank you Kurt von Kubik,
now it goes way better to get an impression:
will say: the way over ImageShack works fine ;-) and the quality of the photos is very good
but one more question...maybe a little bit of luxury ;-))
please additionally one more photo in the bird's eye view
then everybody gets nearly a view like on a schematics
would be great!!
cheers
suchtgutenklang
now it goes way better to get an impression:
will say: the way over ImageShack works fine ;-) and the quality of the photos is very good
but one more question...maybe a little bit of luxury ;-))
please additionally one more photo in the bird's eye view
then everybody gets nearly a view like on a schematics
would be great!!
cheers
suchtgutenklang
Here we go I hope!
If you want a good laugh, you can have a look at our completely worn out prototype - the PCB on the side is actually the shunts, as it was not originally designed with shunts.
Also the regulators for the digital section was originally designed a lot different from the one in the new design.
On top of that - the kaos is created by a lot of chip decoupling experiments.
I knowit looks lika a time bomb, but relax - it isn´t
Btw. There is actually components on both sides of the PCB.
On the new design it´s just the opamps for the servo and a few surrounding resistors.
The phono receptacles on the first photo,wasn´t soldered, because we wanted to test other models like the Canare 75 Ohm RCA receptacle and maybe Neutrik RCA´s for the analogs.
We also chose to run the 4398 in the power up mode, which means that it doesn´t go of during silence, but stays on.
This results in minor noise during silence, but we think that´s a fair trade for an ever ready chip, which by experience performs best after some hours on.
An externally hosted image should be here but it was not working when we last tested it.
If you want a good laugh, you can have a look at our completely worn out prototype - the PCB on the side is actually the shunts, as it was not originally designed with shunts.
Also the regulators for the digital section was originally designed a lot different from the one in the new design.
On top of that - the kaos is created by a lot of chip decoupling experiments.
I knowit looks lika a time bomb, but relax - it isn´t
An externally hosted image should be here but it was not working when we last tested it.
Btw. There is actually components on both sides of the PCB.
On the new design it´s just the opamps for the servo and a few surrounding resistors.
The phono receptacles on the first photo,wasn´t soldered, because we wanted to test other models like the Canare 75 Ohm RCA receptacle and maybe Neutrik RCA´s for the analogs.
We also chose to run the 4398 in the power up mode, which means that it doesn´t go of during silence, but stays on.
This results in minor noise during silence, but we think that´s a fair trade for an ever ready chip, which by experience performs best after some hours on.
Algar_emi said:Hurtig, no info on the RCA connector yet
http://www.diyaudio.com/forums/showthread.php?postid=1740891#post1740891
Or in the UK.http://www.rapidonline.com/Cables-C...hono/Professional-quality-phono-sockets/28543
Re: hifimaker hello friend ..
Here are a couple of websites that sell preamp controls:
http://electronics.dantimax.dk/Kits/index.html
http://www.twistedpearaudio.com/control/jt.aspx
Twisted Pair Thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=74224&perpage=50&pagenumber=1
Hope this helps in getting you started.
-David
Danilo Almeida said:
Here are a couple of websites that sell preamp controls:
http://electronics.dantimax.dk/Kits/index.html
http://www.twistedpearaudio.com/control/jt.aspx
Twisted Pair Thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=74224&perpage=50&pagenumber=1
Hope this helps in getting you started.
-David
Cauhtemoc said:
Being an engineer, I believe measurements do indeed matter, and I spend a great deal measuring and simulating. With that said however, I also avoid feedback if possible, simply because I have found it to sound better. And for the same reason I prefer the AK4396 or WM8740 over the CS4398, they sound better.
It is also interesting to note that you chose a no feedback analog stage becuase it sounds better, yet you chose the CS4398 because it had the lowest THD. A bit contradictory don't you think?
Actually, I have talked to a representative at Crystal about the WM8804, and while he unofficially acknowledged that it probably was league above the CS8416, the official response was very similar to yours, that it is a well proven device with a proven track record.
You also have to remember that the CS8416 is old. It is pretty much the same design as the CS8412, which was introduced in 1998. Back then it was very good, the only competitor at the time was YM3623B, which could have as much as 2 ns of jitter (truth be told there was also the SAA7274, which is probably the best SPDIF receiver ever made, although it went out of production almost before it started).
Of course you don't have to try other DACs. If you are happy with the performance you are getting now then by all means leave it as it is. However if you are still looking to improve it you should try the Wolfson DACs, or better yet, the new AKM DACs. Trust me on this, those Scotts and Japs know what they are doing.
The WM8741 and WM8742 are their two top DACs, and the difference between them should be quite small. The people at Wolfson are very firendly, I do not think you will have any problems getting samples of the WM8804 or any DAC of your choice.
I'm not really sure about your intentions in Audio. My goal is tho have the best possible sound. To me it sounds like you are more into choosing the components you think is best, and then live with the sound it produces.
Question is: If for examble: What if you listened to our DAC, and accepted that it was the best you ever heard. Would you then be able to live with the fact that it was based on IC's from Crystal?? Or would you comprimise the sound, to get a Wolfson based DAC??
I'm not trying to take sides in this DAC chip argument but I thought I'd check to see what chip the Bryston BDA-1 DAC uses as this seems to be getting praised as a fantastic new product. As you can see HERE it uses the crystal CS-4398. The BDA-1 cost's $1995 and I've heard of people swapping $5000 DAC's in preference of the BDA-1, so the CS-4398 surely can't be that bad.
Hurtig said:
Then you have misunderstood me, my goal is certainly also to have the best possible sound. I try to evaluate each component and choose the one I think sounds best.
Hurtig said:
If this were to happen, I would try to find out what makes your DAC the best, and if this turns out to be the CS4398, I would no longer consider it inferior to the WM8740 for example, and I would of course not "downgrade" to the Wolfson DAC.
DavidJE said:
I never said it was bad, in fact it is far from bad, I just said that the WM8740 or AK4396 was better.
Well I don´t think it is possible to choose anything that might make everyone happy really!
From my viewpoint the CS4398 is a chip as cool as they get these days.
I know it is possible to twist its sound pretty much hardly doing nothing, just altering the decoupling of i.e. the PSU for the analog section, and suddenly you´ll find it sound very different.
If you then hook it up with an analog stage of the kind we did, I guarantee that you can hear it was tweaked.
To me it is indeed very difficult to discuss wether this or this chip might be better, as the context is also different.
Looking back on the listening experience with different chips I´d never think of Wolfson as an alternative, neither AKM. It would be Burr Brown, but as already mentioned, the BB is more difficult to implement, and it might actually be, that the superior specs of 1794A precisely is acomplished because it does not contain the circuitry that is extra needed - just an idea.
The 32 bits sample length in both Wolfson and AKM I consider not of any interest to us. We have 16 bit information on CD´s, and most high rez. recordings is done 24/192 of which both 8416 and 4398 is perfectly capable, we do not intend to do any upsampling either.
In other words I´d not give any attention to these chips, mostly because the rest of the DACs components is of bigger importance, and next because I never really liked them.
I also do need another reason than "it sounds better", because I need to know what better?
Generally the attention is always attracted to the points on which the light shines, very seldom to where the problem really is.
Exactly that is the reason why we do not implement a 120.000µF electrolytic bank, heavy 750 VA RC transformers, 8 DAC chips per channel and a single transistor output stage with a 1000$ cap.
A lot would like that and even be impressed, we´ve seen the quest from a somewhat different viewpoint, where low supply noise, low impedance. a capacitorfree signalpath, carefully designed filtration and most of all a no feed back design has been attended thorougly.
If not close to end of life the Burr Brown PCM1704 easily could have found its way into our project. It was the target chip in the beginning, but it was at a critical point NRND, although it actually has turned into an active component again. But life is certainly a lot easier implementing complete stereo systems with onboard filters than 4 DACs and one NPC filter chip.
Anyways - I think Hurtig might give the Wolfson reciever a try sometime even though I´d rather not.
From my viewpoint the CS4398 is a chip as cool as they get these days.
I know it is possible to twist its sound pretty much hardly doing nothing, just altering the decoupling of i.e. the PSU for the analog section, and suddenly you´ll find it sound very different.
If you then hook it up with an analog stage of the kind we did, I guarantee that you can hear it was tweaked.
To me it is indeed very difficult to discuss wether this or this chip might be better, as the context is also different.
Looking back on the listening experience with different chips I´d never think of Wolfson as an alternative, neither AKM. It would be Burr Brown, but as already mentioned, the BB is more difficult to implement, and it might actually be, that the superior specs of 1794A precisely is acomplished because it does not contain the circuitry that is extra needed - just an idea.
The 32 bits sample length in both Wolfson and AKM I consider not of any interest to us. We have 16 bit information on CD´s, and most high rez. recordings is done 24/192 of which both 8416 and 4398 is perfectly capable, we do not intend to do any upsampling either.
In other words I´d not give any attention to these chips, mostly because the rest of the DACs components is of bigger importance, and next because I never really liked them.
I also do need another reason than "it sounds better", because I need to know what better?
Generally the attention is always attracted to the points on which the light shines, very seldom to where the problem really is.
Exactly that is the reason why we do not implement a 120.000µF electrolytic bank, heavy 750 VA RC transformers, 8 DAC chips per channel and a single transistor output stage with a 1000$ cap.
A lot would like that and even be impressed, we´ve seen the quest from a somewhat different viewpoint, where low supply noise, low impedance. a capacitorfree signalpath, carefully designed filtration and most of all a no feed back design has been attended thorougly.
If not close to end of life the Burr Brown PCM1704 easily could have found its way into our project. It was the target chip in the beginning, but it was at a critical point NRND, although it actually has turned into an active component again. But life is certainly a lot easier implementing complete stereo systems with onboard filters than 4 DACs and one NPC filter chip.
Anyways - I think Hurtig might give the Wolfson reciever a try sometime even though I´d rather not.
Hi KvK and Hurtig,
What I can conclude from this thread is that you guys designed nice peace of audio hardware. Later on you realized that low ground noise is the name of the game and absolute prerequisite to achieve high quality sound. In other words – no doubt that 5-6 years spent on this project has ripped benefits.
What I don’t understand is the intention of this thread:
We are all more or less aware of your findings that took you 5-6 years or R&D.
You can not really puzzle most of us by saying that you have a fantastic design.
If you don’t provide the circuit diagram and PCB layout, YOU and US are both at loss. We are both wasting time.
YOU are not going to receive our suggestions on improving your design (which you may or may not discard a priori – up to your egos…).
US will not learn absolutely anything from this thread.
This is DIY site – either share for mutual benefit, or make it commercial thru different channels.
Best Regards,
Boky
What I can conclude from this thread is that you guys designed nice peace of audio hardware. Later on you realized that low ground noise is the name of the game and absolute prerequisite to achieve high quality sound. In other words – no doubt that 5-6 years spent on this project has ripped benefits.
What I don’t understand is the intention of this thread:
We are all more or less aware of your findings that took you 5-6 years or R&D.
You can not really puzzle most of us by saying that you have a fantastic design.
If you don’t provide the circuit diagram and PCB layout, YOU and US are both at loss. We are both wasting time.
YOU are not going to receive our suggestions on improving your design (which you may or may not discard a priori – up to your egos…).
US will not learn absolutely anything from this thread.
This is DIY site – either share for mutual benefit, or make it commercial thru different channels.
Best Regards,
Boky
Kurt von Kubik said:
I will stop arguing now. I have given you my opinion about the DAC and SPDIF receiver, you have given me yours, let's leave it at that and move on.
On another note, I think a lot of people would like to see your schematics. It would be most interesting to see how you implemented your analog stage or the PSU for example.
I see that a lot of users are intersted in the schematic. Currently I am still working on these.
But I don't really see why they are so interesting
As I have mentioned several times, it is not a matter of using the right schematic or the best components. It's "simply" a matter of "matching" the schematic, components and real-life implemnetation.
Change one transistor to a different type, and you may end up with a crappy DAC. Change the PCB layout slightly, and you may see the same result...
You should consider the art of building audio gear, as building a violin I have not yet seen the cookbook of building the perfect Stradivarius. Simply because it is pure craftmanship.
But I don't really see why they are so interesting
As I have mentioned several times, it is not a matter of using the right schematic or the best components. It's "simply" a matter of "matching" the schematic, components and real-life implemnetation.
Change one transistor to a different type, and you may end up with a crappy DAC. Change the PCB layout slightly, and you may see the same result...
You should consider the art of building audio gear, as building a violin I have not yet seen the cookbook of building the perfect Stradivarius. Simply because it is pure craftmanship.
- Status
- Not open for further replies.