IMO, to put it simply, non oversamling avoids the use of the SAA7220 upsampling filter chip in the signal path, which for various reasons, results in "bad" sound.
When this chip is bypassed most people notice an improvement in the sound.
The problem is the noise generated by the power requirements of the SAA7220 ( 180mA draw at 5V ). So yes, cutting out this chip does little to settle the question of whether OS or NOS sounds better - its influence on the power supplies (rather than the signal path) is what knackers the sound. I noticed a big improvement in the sound myself when the SAA7220 chip was removed - much lower sibilance.
I also was not going to enter the debate about NOS vs OS, only point out that in this case, it sounds better when removed 🙂
Last edited:
Discrete sounds better also than any Opamp(6 kinds) I used in different players sony555 and diy dacs. My discrete i/v has way more details and doesn’t curtain the sound. You can hear distinctly every words with all the voices of a choir singing along. Tda1541 is the best dac I experienced, following with cirus logic, sabre, akm. To me the sabre and akm are not 'in same league' but the cirus is awesome. The pcm56k is kind of good but too electronic sounding with an overemphasis of middle bass. The cirus has incredible highs but so far lacks the instruments separation a tda1541 has.
What would be the best way for connecting the DAC to the power amp?
DAC->IV Stage->Volume Pot->Power Amp
or
DAC->IV Stage->Volume Pot->Preamp->Power Amp
If I need a preamp I can integrate it into the DAC enclosure, I think I want to use tubes for the preamp what would be the best circuit for driving the power amps (10K impedance) directly.
DAC->IV Stage->Volume Pot->Power Amp
or
DAC->IV Stage->Volume Pot->Preamp->Power Amp
If I need a preamp I can integrate it into the DAC enclosure, I think I want to use tubes for the preamp what would be the best circuit for driving the power amps (10K impedance) directly.
anybody
You need to determine whether you need the pre amp stage. If the PA is sensitive enough and/or the the output of the I/V stage is high enough, then no need for a preamp, or just have a passive one - which is a stupid 'hifi' name for a vol pot.
However, in a nos design you will have to filter the output before it gets to the PA and the opamps are the simplest option.
Suck it and see!
What is DEM Re-Clocking exactly and does it improve the sound quality?
Can I add DEM Re-Clocking to the AnalogMetric Board by copying the setup in the first picture I uploaded.
I have found two output stages that seem to be really good, Which one has the lowest distortion, noise, and best SQ?
for Pedja Rogic DDNF IV Stage can i sub the transistors for BC560C,BC550C and the JFET for a BF256?
can remove the DC blocking cap from "DDNF IV Stage" if it goes through a 10K volume and into the Pass Labs B1 Buffer stage?
Can I add DEM Re-Clocking to the AnalogMetric Board by copying the setup in the first picture I uploaded.
I have found two output stages that seem to be really good, Which one has the lowest distortion, noise, and best SQ?
for Pedja Rogic DDNF IV Stage can i sub the transistors for BC560C,BC550C and the JFET for a BF256?
can remove the DC blocking cap from "DDNF IV Stage" if it goes through a 10K volume and into the Pass Labs B1 Buffer stage?
Attachments
What is DEM Re-Clocking exactly and does it improve the sound quality?
It improves EVERYTHING!
Yes it works fine.Can I add DEM Re-Clocking to the AnalogMetric Board by copying the setup in the first picture I uploaded.
I have found two output stages that seem to be really good, Which one has the lowest distortion, noise, and best SQ?
Neither of them filter out the sampling frequency.
Have you even built the DAC yet? Respectfully, as advice, you ask a lot of questions, and many questions you ask will involve subjective replies. It may be time to build some things and make up your own mind. Then you can let us know what you think.
Last edited:
can remove the DC blocking cap from "DDNF IV Stage" if it goes through a 10K volume and into the Pass Labs B1 Buffer stage?
What do you think will happen if you do this?
what happen? I don't really want to know 🙂. I've tried ECdesign's IV with 2SK170-BL, here I did manage to get rid of output capacitors using adjustable Vref, the sound is very transparent, next best thing after tube output imo..without high voltage hassles
Im sorry that I ask so many questions, i never build a DAC before and it pretty confusing. I ordered the board from analogmetric like a week ago and I got a TDA1541 R1 from a broken Philips CD Player and I bought the CS8412 a couple days ago i'm just waiting for the stuff to arrive.
It improves EVERYTHING!
ERi HI
Tell me about ' everything ' please ?😀
I've done advanced re clocking and am looking to do some sort of DEM re clock mod VERY soon - just a few resistors to come.
How did you do yours ?
Is it really that noticeable.....I want you to say ' yes ' so I keep going with this idea.
I keep err - ing as everything is sounding real nice with a few tweaks..
I kept the 7220 for OS but silenced it's terrible contribution by giving it it's own 11.289 and it's own dedicated power supply and transformer. Same for 7310 and RAM.
TDA has it's own 5.6 and 11.289 and I really want to do something with pins 16 and 17
I've seen and been told of a few ways to deal with it but just wanted your opinion as you've obviously done DEM re clocking based on your comments.
Thanks
Andrew
Im sorry that I ask so many questions, i never build a DAC before and it pretty confusing. I ordered the board from analogmetric like a week ago and I got a TDA1541 R1 from a broken Philips CD Player and I bought the CS8412 a couple days ago i'm just waiting for the stuff to arrive.
1541 or 1541a? There are some minor differences, eg the functioning of pin 4
ERi HI
Tell me about ' everything ' please ?😀
I've done advanced re clocking and am looking to do some sort of DEM re clock mod VERY soon - just a few resistors to come.
Andrew,
I've listened to TDA1541A for years in many different CD players and DACs.
if you just use the 470pf cap on pins 16,17. its a nice sounding DAC, warm, musical and detailed. Where it lacks is that ulimate crisp HF transparency you can get with *some* modern delta sigma DAC chips. Depending on the output stage, it can sound a little bit rounded in the bass, a little bit lacking in overall definition, and an overall sense of "fuzzyness" over the entire sound. Not knocking it, as even like this IMO, it still is a prefereable sound to the IMO, "wierd unnatural" sound of *most* delta sigma DACs. (Obviously this is my own subjective opinion)
Reclocking using a 74HC04 does this:
Bass becomes tighter and conveys tone better, leading edges of notes are faster and more defined, High frequenices loose the slight fuzzyness and become crystal clear. High hats sound like real high hats, brassy and fast. Overall detail improves vastly over the stock sound using the cap. Background sounds and "atmosphere" are easilly heard. The prominent midrange sound of the TDA1541 goes away. The sound is more balanced from top to bottom. Bass goes deeper and more realistic. highs seem to go higher and cleaner. There is NO downside whatsoever. Like I said earlier, everything gets better
Is "advanced reclocking" the same as asyncronously reclocking? I thought that was no good becuse you are dropping the odd sample because the separate clock is not syncronised to the incoming SPDIF signal.
Or are you talking about a CD player?
How did you do yours ?
Pretty much the same as the post on the other page, and I attached it again. EC designs used a 74HC02, becuse that is what he had lying around. I used a 74HC04. becuse that is what I had lying around. I presume either one will work as well as the other.
Is it really that noticeable.....I want you to say ' yes ' so I keep going with this idea.
I keep err - ing as everything is sounding real nice with a few tweaks..
Yes, its noticable like day and night. Once you reclock you will NEVER go back.
I kept the 7220 for OS but silenced it's terrible contribution by giving it it's own 11.289 and it's own dedicated power supply and transformer. Same for 7310 and RAM.
TDA has it's own 5.6 and 11.289 and I really want to do something with pins 16 and 17
I've seen and been told of a few ways to deal with it but just wanted your opinion as you've obviously done DEM re clocking based on your comments.
Thanks
Andrew
I suggest you at least try Non-oversampling just to try it out and see if you like it. On the analogmetric board, I just cut a few tracks and used the socket that was designed for the asyncronous reclocking circuit to install the 74HC04 and run the resistors to the DAC. I take the -15V directly off the pin of the TDA1541 -No problems with this at all.
On my other DAC I just used a 7805 chip feeding the 74HC04, point to point wiring, nothing particularly neat or nice looking, but it works like a charm.
any more questions, let me know.
Attachments
Last edited:
Im sorry that I ask so many questions, i never build a DAC before and it pretty confusing. I ordered the board from analogmetric like a week ago and I got a TDA1541 R1 from a broken Philips CD Player and I bought the CS8412 a couple days ago i'm just waiting for the stuff to arrive.
Though you may choose to ignore me, my advice is to build the kit in standard form, and make sure it works, before doing mods. At least then you will know if your mods are making a nice improvement or not. 🙂
Hi,
This is designed for TDA1545 (by me) and it does not work for TDA1541. There is simple modification that will make it work either.
There are several versions of analog stages for TDA1541 that I have posted over time. More info here:
http://www.diyaudio.com/forums/digital-source/54817-thorstens-valve-output-stage-tda1541dac.html#post614527
Past that, the Analogmetric DAC (and in fact most of the cheap as chip's china kits that copied Thomas [Tubelovers] designs are so riddled with errors, it is probably better to make a DAC from schematics on Veroboard, paying attention to layout and grounding (self adheasive copper foil is an excellent tool here).
Ciao T
I head that grounded grid is good for i/v conversion.
how would you make this work for the tda1541
This is designed for TDA1545 (by me) and it does not work for TDA1541. There is simple modification that will make it work either.
There are several versions of analog stages for TDA1541 that I have posted over time. More info here:
http://www.diyaudio.com/forums/digital-source/54817-thorstens-valve-output-stage-tda1541dac.html#post614527
Past that, the Analogmetric DAC (and in fact most of the cheap as chip's china kits that copied Thomas [Tubelovers] designs are so riddled with errors, it is probably better to make a DAC from schematics on Veroboard, paying attention to layout and grounding (self adheasive copper foil is an excellent tool here).
Ciao T
Hi,
DEM is short for dynamic element matching.
It is a technique used in the TDA1541 to improve the converter accuracy without having to individually laser trim the chips (as is done with other multibit DAC's). It uses more than one element per Bit and switches between theme, averaging out any errors.
Stock the TDA1541 DEM circuit runs of a separate local clock, which normally does not synchronise with the digital signal. The result is that the remaining error is randomised between samples, however as there is no guarantee that there is a certain number of DEM cycles in a sample a different kind of error is introduced. DEM reclocking cancels that error.
If you remove the SAA7220 Filter you can fit a small PCb in it's place and fit the DEM reclocking parts there, You can use the 4040 to give you the neccesary clock divider from BCK.
Here you can my diyhifisupply Satch DAC, where I performed this (and many other) modifications:
Ciao T
What is DEM Re-Clocking exactly and does it improve the sound quality?
DEM is short for dynamic element matching.
It is a technique used in the TDA1541 to improve the converter accuracy without having to individually laser trim the chips (as is done with other multibit DAC's). It uses more than one element per Bit and switches between theme, averaging out any errors.
Stock the TDA1541 DEM circuit runs of a separate local clock, which normally does not synchronise with the digital signal. The result is that the remaining error is randomised between samples, however as there is no guarantee that there is a certain number of DEM cycles in a sample a different kind of error is introduced. DEM reclocking cancels that error.
Can I add DEM Re-Clocking to the AnalogMetric Board by copying the setup in the first picture I uploaded.
If you remove the SAA7220 Filter you can fit a small PCb in it's place and fit the DEM reclocking parts there, You can use the 4040 to give you the neccesary clock divider from BCK.
Here you can my diyhifisupply Satch DAC, where I performed this (and many other) modifications:
An externally hosted image should be here but it was not working when we last tested it.
Ciao T
Im sorry that I ask so many questions, i never build a DAC before and it pretty confusing. I ordered the board from analogmetric like a week ago and I got a TDA1541 R1 from a broken Philips CD Player and I bought the CS8412 a couple days ago i'm just waiting for the stuff to arrive.
I have just bought a Nos/Os dual tda1541a (parallel) kit from rainbow_hui on ebay.
It uses a WM8805 & 26c32 front end, + SAA7220 which can be bypassed with a couple of jumpers and has onboard psus. (discrete)
I was going to build my own from scratch but with this kit I can make my own mods and use different components. eg Wima MKS4 220n, & 470n for the DAC decoupling.
I will omit the I/V and output stage and reconfigure it for differential and connect to my Sowter I/V transformers possibly by an LC Tee filter.
- Status
- Not open for further replies.
- Home
- Source & Line
- Digital Source
- Need help with Analogmetric TDA1541 DAC