I must be thick, mate
Can someone explain why it needs to be complex? All the complexity is in the brick wall that proceeds the conversion.
Right, which means that you will still get tons of ultrasonic crap out of this stage.
patwen:
Is this what you were expecting in terms of performance?
Jocko
Can someone explain why it needs to be complex? All the complexity is in the brick wall that proceeds the conversion.
Right, which means that you will still get tons of ultrasonic crap out of this stage.
patwen:
Is this what you were expecting in terms of performance?
Jocko
Re: RE: sinx/x filter:
Jocko,
I see that i split the thread up so that the original thread can go on.
edit: i have to copy the thread as a few posts have to remain in both threads to avoid confusion. Stored links to posts may not work.
Ren,
do you know what a Wadlbeisser is? this is a small cowardly dog trying to bite the lower legs of bystanders but only from behind.
Jocko,
good idea!
I see that i split the thread up so that the original thread can go on.edit: i have to copy the thread as a few posts have to remain in both threads to avoid confusion. Stored links to posts may not work.
Ren,
do you know what a Wadlbeisser is? this is a small cowardly dog trying to bite the lower legs of bystanders but only from behind.
Kudos.....Dice Man!
That is Ren to a tee. His main raison d'etre is to torment me. He is constantly giving me unsolicited advice, which I always ignore.
Nothing riles him more than being ignored.
(Actually, he is a competent designer. He is just a bigger smart-*** than I am. Hard to believe, but true.)
Yip ! Yip!
Jocko
That is Ren to a tee. His main raison d'etre is to torment me. He is constantly giving me unsolicited advice, which I always ignore.
Nothing riles him more than being ignored.
(Actually, he is a competent designer. He is just a bigger smart-*** than I am. Hard to believe, but true.)
Yip ! Yip!
Jocko
simple complexity
Jocko Homo>
Regarding ultra sonic noise, I am using a tube amp, so I don't have a problem with that.
I did try non-oversampling ("NOS") TDA1541A without ANY filtering at all (directly passive I/V only). To me, the sound is very good, no ultra sonic noise whatsoever. The main character of the NOS TDA1541A is "musical", it is so much more musical than normal OS DAC, although it doesn't sound "hi-fi", so this may turn some audiophile off. The only problem is a bit lack of treble, which should be corrected by the sin(x)/x filter.
Over the years, I have realized that (at least in my case) the fewer number of components in the signal chain and the more simple the circuit, the better the sound. When I look back, all the upgrades I have done and currently still staying in my system are the ones that simplified the circuit or signal path. Every time, the improvement is very similar, more transparency, more details and more musical.
For me, the source (CD or LP) is the best you can get, everything down the chain will not improve on what you already have (as we are not adding an new information to it - information theory), the best that we can do it to minimize losses caused by the later stages (filtering or amplification, whatever).
As no component is perfect, so the fewer the number of components down the chain, the lesser the imperfection and the lesser the contamination to the sound (that¡¦s one of the reason I switch from SS to tube, they are so much more simple and elegant). Of course, one still need to design things correctly, but just try to do it with fewer but best quality components.
The above is my general impression, by no means golden rule. For example, power supply design is slightly different, one must use quite a few number of components to get it right. But I guess the ¡§simple is better¡¨ rule can still apply. Like instead of using many stages of RCRCRC filtering, a proper designed LC stage filter will sound much better.
So, for me, the most simple can accomplish the most complex.
Jocko Homo>
Regarding ultra sonic noise, I am using a tube amp, so I don't have a problem with that.
I did try non-oversampling ("NOS") TDA1541A without ANY filtering at all (directly passive I/V only). To me, the sound is very good, no ultra sonic noise whatsoever. The main character of the NOS TDA1541A is "musical", it is so much more musical than normal OS DAC, although it doesn't sound "hi-fi", so this may turn some audiophile off. The only problem is a bit lack of treble, which should be corrected by the sin(x)/x filter.
Over the years, I have realized that (at least in my case) the fewer number of components in the signal chain and the more simple the circuit, the better the sound. When I look back, all the upgrades I have done and currently still staying in my system are the ones that simplified the circuit or signal path. Every time, the improvement is very similar, more transparency, more details and more musical.
For me, the source (CD or LP) is the best you can get, everything down the chain will not improve on what you already have (as we are not adding an new information to it - information theory), the best that we can do it to minimize losses caused by the later stages (filtering or amplification, whatever).
As no component is perfect, so the fewer the number of components down the chain, the lesser the imperfection and the lesser the contamination to the sound (that¡¦s one of the reason I switch from SS to tube, they are so much more simple and elegant). Of course, one still need to design things correctly, but just try to do it with fewer but best quality components.
The above is my general impression, by no means golden rule. For example, power supply design is slightly different, one must use quite a few number of components to get it right. But I guess the ¡§simple is better¡¨ rule can still apply. Like instead of using many stages of RCRCRC filtering, a proper designed LC stage filter will sound much better.
So, for me, the most simple can accomplish the most complex.
I still feel the reason this non-o/s stuff has caught on is largely due to the SAA7220. It has a zero in the response, to cancel out one of the 3 poles Philips uses in the analog filter. Looking at these on a spectum analyzer tells a very interesting story.
Non-o/s probably sounds a lot like the older Wadia convertors. It had what appeared to be nothing more than a Bessel filter, done in the digital domain. Looking at one of those on a spectrum analyzer told an even more ugly story.
I would not attempt to build the sort of system you have, but if your system is indeed free from the effects of stuff above 20 kHz, well ok. In that case, there is an almost endlessly number of variations of second-order Sallen-Key types filters......with peaking....... that you can try to get some treble boost.
I guess to each his own.........
Jocko
Non-o/s probably sounds a lot like the older Wadia convertors. It had what appeared to be nothing more than a Bessel filter, done in the digital domain. Looking at one of those on a spectrum analyzer told an even more ugly story.
I would not attempt to build the sort of system you have, but if your system is indeed free from the effects of stuff above 20 kHz, well ok. In that case, there is an almost endlessly number of variations of second-order Sallen-Key types filters......with peaking....... that you can try to get some treble boost.
I guess to each his own.........
Jocko
Sallen-Key filters
"There is not much Sallen-Key types filters materials on the net "
WHAT?!!!!!!!!!!!!!!!!!!!!!!!
Cathode followers
http://www.google.com/search?hl=en&ie=UTF-8&oe=UTF-8&q=Cathode+followers&btnG=Google+Search
Sallen-Key
http://www.google.com/search?hl=en&lr=&ie=UTF-8&oe=UTF-8&q=Sallen-Key+&btnG=Google+Search
jfet followers
http://www.google.com/search?hl=en&lr=&ie=UTF-8&oe=UTF-8&q=jfet+followers&btnG=Google+Search
Passive components:
Resistors, capacitors, inductors:
Active components:
Tubes, op amps, transistors ICs
Capise?
You are just funning me to get a rude respons right?
Art for art's sake money for godsake.
"There is not much Sallen-Key types filters materials on the net "
WHAT?!!!!!!!!!!!!!!!!!!!!!!!
Cathode followers
http://www.google.com/search?hl=en&ie=UTF-8&oe=UTF-8&q=Cathode+followers&btnG=Google+Search
Sallen-Key
http://www.google.com/search?hl=en&lr=&ie=UTF-8&oe=UTF-8&q=Sallen-Key+&btnG=Google+Search
jfet followers
http://www.google.com/search?hl=en&lr=&ie=UTF-8&oe=UTF-8&q=jfet+followers&btnG=Google+Search
Passive components:
Resistors, capacitors, inductors:
Active components:
Tubes, op amps, transistors ICs
Capise?
You are just funning me to get a rude respons right?
Art for art's sake money for godsake.
Thanks ....
Please kindly re-read my phase:
“There is not much Sallen-Key types filters materials on the net which are using PASSIVE components”
I have used the word “PASSIVE”, so I was looking for materials which talk about Sallen-Key types filters using passive components, and ONLY passive components.
Many links in from the Google searches (I did do many searches) are using opamps or other active components.
So please don’t be rude and don’t make any pre-crime judgment like Mr.Bush.
[Joke]
Thanks for your help anyway.
Please kindly re-read my phase:
“There is not much Sallen-Key types filters materials on the net which are using PASSIVE components”
I have used the word “PASSIVE”, so I was looking for materials which talk about Sallen-Key types filters using passive components, and ONLY passive components.
Many links in from the Google searches (I did do many searches) are using opamps or other active components.
So please don’t be rude and don’t make any pre-crime judgment like Mr.Bush.
[Joke]Thanks for your help anyway.
Filters that I have used:
NPC5813.....does not support I2S,
NPC5843....can support I2S with a few extra parts,
DIF1704.....supports I2S, and most everything else.
Have lots of the first two extra, if anyone wants to play with them.
Now......back to analog filters.
A simple first-order, passive only, is not going to help you in the least. Period.
A second-order CAN be made with only one JFET, and a few parts around it. Try it sometime.
Jocko
NPC5813.....does not support I2S,
NPC5843....can support I2S with a few extra parts,
DIF1704.....supports I2S, and most everything else.
Have lots of the first two extra, if anyone wants to play with them.
Now......back to analog filters.
A simple first-order, passive only, is not going to help you in the least. Period.
A second-order CAN be made with only one JFET, and a few parts around it. Try it sometime.
Jocko
Sallen-Key
"I was looking for something more simple (passive filter and tubes)in the signal path to ensure the sound quality'
I don't think it is possible to make a Sallen-Key with only passive parts. I guess I also was under the impression that you wanted a simple filter built around cathode followers (tubes)which should be straight forward with some research in the links Mr. Nyos supplied.
"Of course, one still need to design things correctly, but just try to do it with fewer but best quality components."
What is it that you are actually trying to design and what is the desired response from the circuit?
H.H.
"I was looking for something more simple (passive filter and tubes)in the signal path to ensure the sound quality'
I don't think it is possible to make a Sallen-Key with only passive parts. I guess I also was under the impression that you wanted a simple filter built around cathode followers (tubes)which should be straight forward with some research in the links Mr. Nyos supplied.
"Of course, one still need to design things correctly, but just try to do it with fewer but best quality components."
What is it that you are actually trying to design and what is the desired response from the circuit?
H.H.
HarryHaller>
>I don't think it is possible to make a Sallen-Key with only >passive parts.
This is exactly the info I need, so I can stop looking.
>I guess I also was under the impression that you wanted a >simple filter built around cathode followers (tubes)which should
Exactly, a simple ALL PASSIVE filter with a gentle peak of 3dB around 20KHz, than roll off on a 1st or 2nd slope. The phase of the filter should be keep constant (Bessel filter?), and lagging in phase.
Then a very simple tube output stage (No SRPP) to amplify the voltage.
Optional:
1. A transformer I/V stage
2. a couple more dips for the filter at the multiples of the sampling freq to suppress digital images a bit.
Basically, I would like to design somethign which Thorsten did here, but for TDA1541A instead.
http://groups.yahoo.com/group/Thunderstone_technical/files/TL Circuits/NOS-DAC-Analogue/
http://pub4.ezboard.com/ffakeidsfrm1.showMessage?topicID=361.topic
Any suggestions?
Is there an filter design program that I can use? (Pspice is a bit too complex to pick up again, I did used it back when I was a student in Uni, but now nearly forget everything.)
Thanks.
>I don't think it is possible to make a Sallen-Key with only >passive parts.
This is exactly the info I need, so I can stop looking.
>I guess I also was under the impression that you wanted a >simple filter built around cathode followers (tubes)which should
Exactly, a simple ALL PASSIVE filter with a gentle peak of 3dB around 20KHz, than roll off on a 1st or 2nd slope. The phase of the filter should be keep constant (Bessel filter?), and lagging in phase.
Then a very simple tube output stage (No SRPP) to amplify the voltage.
Optional:
1. A transformer I/V stage
2. a couple more dips for the filter at the multiples of the sampling freq to suppress digital images a bit.
Basically, I would like to design somethign which Thorsten did here, but for TDA1541A instead.
http://groups.yahoo.com/group/Thunderstone_technical/files/TL Circuits/NOS-DAC-Analogue/
http://pub4.ezboard.com/ffakeidsfrm1.showMessage?topicID=361.topic
Any suggestions?
Is there an filter design program that I can use? (Pspice is a bit too complex to pick up again, I did used it back when I was a student in Uni, but now nearly forget everything.)
Thanks.
Any suggestions?
I made all the suggestions that I can think of ie. a Sallen Key filter
built around a Cathode follower or tupe gain stage. I guess you will have to read a bit about Sallen-Key filters first but it should be pretty easy to adapt the second order filter in the shown op amp schematic which is a Sallen-Key filter.
"Exactly, a simple ALL PASSIVE filter with a gentle peak of 3dB around 20KHz, than roll off on a 1st or 2nd slope.'
First order filters don't peak.
"A couple more dips for the filter at the multiples of the sampling freq to suppress digital images a bit."
What kind of filter? What frequency bands? How much attenuation? How much in band effects are you willing to tolerate from these filters?
Is there an filter design program that I can use?
Yes there are many Spice programs besides PSpice that have demos on the WEB. Also searh on the Forum for some of the favorites.
H.H.
I made all the suggestions that I can think of ie. a Sallen Key filter
built around a Cathode follower or tupe gain stage. I guess you will have to read a bit about Sallen-Key filters first but it should be pretty easy to adapt the second order filter in the shown op amp schematic which is a Sallen-Key filter.
"Exactly, a simple ALL PASSIVE filter with a gentle peak of 3dB around 20KHz, than roll off on a 1st or 2nd slope.'
First order filters don't peak.
"A couple more dips for the filter at the multiples of the sampling freq to suppress digital images a bit."
What kind of filter? What frequency bands? How much attenuation? How much in band effects are you willing to tolerate from these filters?
Is there an filter design program that I can use?
Yes there are many Spice programs besides PSpice that have demos on the WEB. Also searh on the Forum for some of the favorites.
H.H.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.