Hello every body.
This is the schematic for the new dac project that I and Pred is working on now
Please comment what you like and don´t like!!
But no discussion off OS versus NON Oversampling!!
PCM1798 are dual mono
PCM1794 can also be used
All 22uF caps are Oscons SP
100nF and 10nF are ceramic X7R
TL431 from Onsemi have very low noise
Ferrite beads are from farnell (which to use?)
Are the connections for LRCK,SCLK,SDATA clear enought??
Input CS8416:
http://5577.freewebpage.org/PCM1798.jpg
AD1896->2*PCM179X (in paralel)->Ext clock:
http://5577.freewebpage.org/PCM1798A.jpg
Best regards
Kim
This is the schematic for the new dac project that I and Pred is working on now
Please comment what you like and don´t like!!
But no discussion off OS versus NON Oversampling!!
PCM1798 are dual mono
PCM1794 can also be used
All 22uF caps are Oscons SP
100nF and 10nF are ceramic X7R
TL431 from Onsemi have very low noise
Ferrite beads are from farnell (which to use?)
Are the connections for LRCK,SCLK,SDATA clear enought??
Input CS8416:
http://5577.freewebpage.org/PCM1798.jpg
AD1896->2*PCM179X (in paralel)->Ext clock:
http://5577.freewebpage.org/PCM1798A.jpg
Best regards
Kim
kimschips said:Yes I knew that will come
I have wrote CCS for each shuntleg. A CCS will set the current for each TL431 for about 50mA + Chip current
I can put in resistors-> because each CCS can be a serie reg instead-> up to people
Thank you both
Kim
Hi
Good !. A small resistor + bead in series with the CCS will maintain high impedance beyond the 1st pole of the CCS.
best
kimschips said:Guido could you please explain that a bit
How do you mean with a CCS 1th pole? where CCS impedance stop falling or is it to define a safe work area?
What is the difference between a 47uH inductor and a Ferrite Bead?
Forgot to tell
I am using Left justified format ok?
Best regards
Kim
Hi Kim,
Yes, 1st pole is where output impedance decreases. Inductors are often wirewound, whereas a bead uses les windings (sometimes 1 only) and ferrite material to raise impedance.
best
Thanks Guido
The wirewound itself-> will it limits it´s high frequency performance?
If you got a big 1mH inductor how big will the performance difference be to an choke bead?
Have you tried Murata 22uF ceramic for decoupling (X5R, SMD1206)??
Have you guys any other comments to the schematic??
Kim
The wirewound itself-> will it limits it´s high frequency performance?
If you got a big 1mH inductor how big will the performance difference be to an choke bead?
Have you tried Murata 22uF ceramic for decoupling (X5R, SMD1206)??
Have you guys any other comments to the schematic??
Kim
Yes Jocko look here for the thread
http://www.diyaudio.com/forums/showthread.php?s=&threadid=4562&highlight=
I have tried it, it´s great for decoupling between 100Khz and 1Mhz
Setup->sinus generator that deliver 10Vpp out->470R resistor->non bippolar capacitor->TL431//22uF murata
Kim
http://www.diyaudio.com/forums/showthread.php?s=&threadid=4562&highlight=
I have tried it, it´s great for decoupling between 100Khz and 1Mhz
Setup->sinus generator that deliver 10Vpp out->470R resistor->non bippolar capacitor->TL431//22uF murata
Kim
Hello
I found out, that I have made some mistakes in the PCM1798 settings
It´s corrected to this:
U4 PCM1798 is mono left channel
U9 PCM1798 is mono right channel
The output will be connected together as in the datasheet.
I have also turned the schematic to it´s normally direction.
I have now heard that I can´t use my 27Mhz clock
(upsampling to 105Khz)???
Should Use a 24.5760Mhz clock instaed
(96Khz sampling)???
You still need to rightclick>copy->paste the link in to your adress field in the browser
Input CS8416:
http://5577.freewebpage.org/PCM1798.jpg
AD1896->2*PCM179X (in dual mono)->Ext clock:
http://5577.freewebpage.org/PCM1798A.jpg
Best regards
Kim
I found out, that I have made some mistakes in the PCM1798 settings
It´s corrected to this:
U4 PCM1798 is mono left channel
U9 PCM1798 is mono right channel
The output will be connected together as in the datasheet.
I have also turned the schematic to it´s normally direction.
I have now heard that I can´t use my 27Mhz clock
(upsampling to 105Khz)???
Should Use a 24.5760Mhz clock instaed
(96Khz sampling)
??? You still need to rightclick>copy->paste the link in to your adress field in the browser
Input CS8416:
http://5577.freewebpage.org/PCM1798.jpg
AD1896->2*PCM179X (in dual mono)->Ext clock:
http://5577.freewebpage.org/PCM1798A.jpg
Best regards
Kim
Bon giorno
Hi got back from vacation. Sicily (Siracusa) is nice, good food.
Now let's go back to the project.
TL431:
I would put always the capacitor over the "higher" resistor in the TL431 power supplies (R3 and R20 in the receiver schematics). It makes a unity gain circuit at higher frequencies and has a lower noise (look at TNT measurements). Since the circuit works properly it can be left out (some people think that this circuit "sounds" better without but engineer in me would put it). Do we really need 3 capacitors since the circuit will be so close to the chip? As I sad a succesfull project is full of compomises and you can leave them since I do not have to install them. I thought we would use OSCONS for C15 and C7 in the same schematics? Do OSCONS exist as SMD? Finally sinking 50mA through TL is an overkill IMHO. Minimum 2mA is needed so 10 including the 1k 330Ohm divider should be enough. Since we are going to use CCS for feeding (it is wired which I like ) the DAC I think they should be on the power supply separate PCB. In that sence has someone knows a good way for wiring the PCBs together at least for PS? I always used soldering but when you try to change somethinh it becomes a pain. Recently did tube preamp and had some hum problems. As a "good boy scout" I have made all connections very short. Imagine the problem of turning the PCB soldered on both sides with wires 
CS8416:
I like it exept choosing inputs. I would like to have three inputs. If I have time I'd add an USB input. I saw on some shematics that people are using the logical buffer between switch and the circuit. I do not know if it is necessary, does somebody knows? I see you have a button for the reset. I am not sure, but like through the fog, I remember that this chip should be reseted after power up. If so I propose we make a simple NE555 reset circuit using monostable multivibrator.
PCMs:
I would leave the configuration pins to the switches. When I do test ASIC PCBs I make what we call "solder switches". It is like a 0805 SMD support. Just the distance is smaller, in fact a little bigger than the minimum distance between the metalic lines of the PCB process. When you want to connect the two togeter (make a jumper) you just simply solder it down. It cuts costs and it is easy. So you "preview" a default configuration that you think will work and than solder if it is necessary. In this way we could test parallel as well as dual mono configuration. That means you have to leave the outputs not connected together. And since it is a current output it should not be the problem to connect them on the I/v conversion board.
AD1896:
I thought we could go 196k with 27MHz clock? I am not sure about your question and will look into it. We did not discussed any clock solution. Do you plan to leave a place only for Guido's XO on the PCB and someone wants to use another clock should connect it externaly? The same thing for config should be applied here in order to change something if necessary. Any error in configuration that might appear in the design period for mis understanfing of the datasheet could cost us a lot and cutting end changing small lines on these small nasty chip will be pain!
Long post. Sorry for that. As I read it through I see I have more questions than answers. I like this middle, strait to the head approach on the DAC design. After I hope we solve PS and than develop 2-3 ideas for I/V conversion
Pred
Hi got back from vacation. Sicily (Siracusa) is nice, good food.
Now let's go back to the project.
TL431:
I would put always the capacitor over the "higher" resistor in the TL431 power supplies (R3 and R20 in the receiver schematics). It makes a unity gain circuit at higher frequencies and has a lower noise (look at TNT measurements). Since the circuit works properly it can be left out (some people think that this circuit "sounds" better without but engineer in me would put it). Do we really need 3 capacitors since the circuit will be so close to the chip? As I sad a succesfull project is full of compomises and you can leave them since I do not have to install them. I thought we would use OSCONS for C15 and C7 in the same schematics? Do OSCONS exist as SMD? Finally sinking 50mA through TL is an overkill IMHO. Minimum 2mA is needed so 10 including the 1k 330Ohm divider should be enough. Since we are going to use CCS for feeding (it is wired which I like
) the DAC I think they should be on the power supply separate PCB. In that sence has someone knows a good way for wiring the PCBs together at least for PS? I always used soldering but when you try to change somethinh it becomes a pain. Recently did tube preamp and had some hum problems. As a "good boy scout" I have made all connections very short. Imagine the problem of turning the PCB soldered on both sides with wires CS8416:
I like it exept choosing inputs. I would like to have three inputs. If I have time I'd add an USB input. I saw on some shematics that people are using the logical buffer between switch and the circuit. I do not know if it is necessary, does somebody knows? I see you have a button for the reset. I am not sure, but like through the fog, I remember that this chip should be reseted after power up. If so I propose we make a simple NE555 reset circuit using monostable multivibrator.
PCMs:
I would leave the configuration pins to the switches. When I do test ASIC PCBs I make what we call "solder switches". It is like a 0805 SMD support. Just the distance is smaller, in fact a little bigger than the minimum distance between the metalic lines of the PCB process. When you want to connect the two togeter (make a jumper) you just simply solder it down. It cuts costs and it is easy. So you "preview" a default configuration that you think will work and than solder if it is necessary. In this way we could test parallel as well as dual mono configuration. That means you have to leave the outputs not connected together. And since it is a current output it should not be the problem to connect them on the I/v conversion board.
AD1896:
I thought we could go 196k with 27MHz clock? I am not sure about your question and will look into it. We did not discussed any clock solution. Do you plan to leave a place only for Guido's XO on the PCB and someone wants to use another clock should connect it externaly? The same thing for config should be applied here in order to change something if necessary. Any error in configuration that might appear in the design period for mis understanfing of the datasheet could cost us a lot and cutting end changing small lines on these small nasty chip will be pain!
Long post. Sorry for that. As I read it through I see I have more questions than answers. I like this middle, strait to the head approach on the DAC design. After I hope we solve PS and than develop 2-3 ideas for I/V conversion
Pred
I am wondering about the power supply also. For example, your last capacitor to the pcm1798 is 220uf, which is much higher than value suggested by Texas Instrument.
I will not parallel black gate caps.
I may consider (but this is your project so you are free to do whatever you please) to have a large capacitor say 2200 uf right after the ferrite to set up a low-pass filter to the shunt.
I may consider just copy the capacitor values from the manufacturer data sheet.
I may consider just having separate V+/GND solder pads for each ICs. The solder pads are connected to the capacitors as suggested by the datasheet. This way people can use whatever powersupply that they want on a daughter board.
I will not parallel black gate caps.
I may consider (but this is your project so you are free to do whatever you please) to have a large capacitor say 2200 uf right after the ferrite to set up a low-pass filter to the shunt.
I may consider just copy the capacitor values from the manufacturer data sheet.
I may consider just having separate V+/GND solder pads for each ICs. The solder pads are connected to the capacitors as suggested by the datasheet. This way people can use whatever powersupply that they want on a daughter board.
Re: Re: Bon giorno
I think somebody already explained this. Also you were looking which one. I will use Murata BLM31AJ601SN1L as Guido has recomended on his site. In farnell it is 581094 reference (0.4 Euro for more than 10 pieces). Also has found some SMD oscons SVP series they are 70mOhm esr compared to 60mOhm ESR of the SC and SA series. They are like 0.9EUR more expensive than standard lead types so if somebody asks me I would stick to standard lead type (not SMD) capacitors at this place. I looked at ZA from rubicon that are advertized as Ultra low ESR and I see that they have around 200mOhm (5 times lower) if I am not mistaking.
Yes I know about K&K audio transformer and resistor approach. Aikido is "hype" these days... I do not want to express my feelings on the subjest and spam this thread but 2 Lundahl's in the output will set me back around 120EUR if I am not mistaking ( do not get me wrong I love my tube preamp). Tubes + good PS in my opinion will go in to the 300EUR figure easy. If you read last Cary CD review in stereophile you will see that even a great tube lover as Art Dudley is, he admited that he could not wauch for better sound performance of the tube output in comparison to SS (of course THD is worse for tube but for me that does not matter much). He liked it just slightly more. The "slightly" I am not ready to pay 300EUR at least not at the begininng. I want reply any more on this subject in future I would like that we make something instead...
Pred
kimschips said:
I think somebody already explained this. Also you were looking which one. I will use Murata BLM31AJ601SN1L as Guido has recomended on his site. In farnell it is 581094 reference (0.4 Euro for more than 10 pieces). Also has found some SMD oscons SVP series they are 70mOhm esr compared to 60mOhm ESR of the SC and SA series. They are like 0.9EUR more expensive than standard lead types so if somebody asks me I would stick to standard lead type (not SMD) capacitors at this place. I looked at ZA from rubicon that are advertized as Ultra low ESR and I see that they have around 200mOhm (5 times lower) if I am not mistaking.
agent.5 said:
Yes I know about K&K audio transformer and resistor approach. Aikido is "hype" these days... I do not want to express my feelings on the subjest and spam this thread but 2 Lundahl's in the output will set me back around 120EUR if I am not mistaking ( do not get me wrong I love my tube preamp). Tubes + good PS in my opinion will go in to the 300EUR figure easy. If you read last Cary CD review in stereophile you will see that even a great tube lover as Art Dudley is, he admited that he could not wauch for better sound performance of the tube output in comparison to SS (of course THD is worse for tube but for me that does not matter much). He liked it just slightly more. The "slightly" I am not ready to pay 300EUR at least not at the begininng. I want reply any more on this subject in future I would like that we make something instead...
Pred
pred said:
i don't have a problem with this either. My point really is to make the PCB simple (and as close to the datasheet as possible). This way if people want to experiment, they can.
And I don't think you need to pay that much for transformers either. Just check out the pricing in the Cinemag group buy thread.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.