Hi dear all,
this recent days I was testing for some parts. The performance is quite interesting.
I hope to collect some advise from more diyer.
pls advise!!
I checked the data sheet of the TDA1541a that can running over higher than 192Khz but I was afraid that the chips will very hot & unstable with more digital noise.
So I try to running the tda1541a in 96Khz.
The method was my DAC ( Twin TDA1541a) had I2S input.
I used CS8414 digital received & added one CS8420 change the data steam to 24Bit 96Khz ( not Use digital filter as DF1706 or 1706 so tda1541a still running at non-oversampling). The HF of the sound field is quite good.
So I test another method which used CS8414>> DF1704
( upsampling to 24Bit 96Khz)
All two condition pass to TDA1541a data steam is 96khz. so it means TDA1451a is running at 16 bit 96Khz.
The result is quite different. The CS8414 + 8420 is more musical than CS8414 with DF1704.
I will later test only added the re-clock to the original kit to compare the different.
BTW..... any diyer can advise me!!!!!!!!
CD driver running with my non-oversampling is
Cayin CDT-15A. This cd driver can provide a original 44.1/48K or 96Khz data steam to DA converter
The first result was getting by this set.
I had TDA1541a data sheet from philips on hand. anyone like to reference pls email to me.
thomas@diyaudiocraft.com
hope can get more advise.
thx
thomas
this recent days I was testing for some parts. The performance is quite interesting.
I hope to collect some advise from more diyer.
pls advise!!
I checked the data sheet of the TDA1541a that can running over higher than 192Khz but I was afraid that the chips will very hot & unstable with more digital noise.
So I try to running the tda1541a in 96Khz.
The method was my DAC ( Twin TDA1541a) had I2S input.
I used CS8414 digital received & added one CS8420 change the data steam to 24Bit 96Khz ( not Use digital filter as DF1706 or 1706 so tda1541a still running at non-oversampling). The HF of the sound field is quite good.
So I test another method which used CS8414>> DF1704
( upsampling to 24Bit 96Khz)
All two condition pass to TDA1541a data steam is 96khz. so it means TDA1451a is running at 16 bit 96Khz.
The result is quite different. The CS8414 + 8420 is more musical than CS8414 with DF1704.
I will later test only added the re-clock to the original kit to compare the different.
BTW..... any diyer can advise me!!!!!!!!
CD driver running with my non-oversampling is
Cayin CDT-15A. This cd driver can provide a original 44.1/48K or 96Khz data steam to DA converter
The first result was getting by this set.
I had TDA1541a data sheet from philips on hand. anyone like to reference pls email to me.
thomas@diyaudiocraft.com
hope can get more advise.
thx
thomas
Konnichiwa,
Don't be.
Based on some old documents from Philips I have around, the TDA1541 was intended for up to 8 Times oversampling at 48KHz, in other words, 384KHz sample rate. Anything up to that will be fine, in any CD Player it was routinely run at 44.1KHz X 4 or 176.4KHz.
What you are doing is in effect bringing back the Digital filter through the backdoor. This time it merely is off a different structure in in most ways more "primitive" than that used in a DAC, as it is not intended as oversampling filter. It's a matter of taste.
You may wish to consider the AS1896 ASRC which would allow No-Oversampling, Syncronous Oversampling to 2/4 times and asyncronous Oversampling to 1/2/4 times a seperate clock, posibly 48KHz rather than 44.1KHz or perhaps both....
Similarly, the CS8420 can be receiver and be used for either syncronous or asyncronous oversampling at a Factor of 2 with either 44.1KHz or 48KHz as baserate.
Sayonara
tube-lover said:
Don't be.
Based on some old documents from Philips I have around, the TDA1541 was intended for up to 8 Times oversampling at 48KHz, in other words, 384KHz sample rate. Anything up to that will be fine, in any CD Player it was routinely run at 44.1KHz X 4 or 176.4KHz.
tube-lover said:
What you are doing is in effect bringing back the Digital filter through the backdoor. This time it merely is off a different structure in in most ways more "primitive" than that used in a DAC, as it is not intended as oversampling filter. It's a matter of taste.
You may wish to consider the AS1896 ASRC which would allow No-Oversampling, Syncronous Oversampling to 2/4 times and asyncronous Oversampling to 1/2/4 times a seperate clock, posibly 48KHz rather than 44.1KHz or perhaps both....
Similarly, the CS8420 can be receiver and be used for either syncronous or asyncronous oversampling at a Factor of 2 with either 44.1KHz or 48KHz as baserate.
Sayonara
Dear T,
If I only combine use the CS8414 digital receiver & pass the data steam to CS8420 ( It means I did not use the Digital filter because 1541a only running in I2S).
Am I must mix with the oscillator than pass a 96Khz data steam to 1541a.
Is this a good method???
I think the working mode of the tda1541a did not use the DF.That meaning it will be same as running in non-oversampling mode!!
Am I wrong??
thanks for your helpful.
thx
thomas
If I only combine use the CS8414 digital receiver & pass the data steam to CS8420 ( It means I did not use the Digital filter because 1541a only running in I2S).
Am I must mix with the oscillator than pass a 96Khz data steam to 1541a.
Is this a good method???
I think the working mode of the tda1541a did not use the DF.That meaning it will be same as running in non-oversampling mode!!
Am I wrong??
thanks for your helpful.
thx
thomas
There are certain conditions that apply. The TDA1541/TDA1541A will run with Fs at anything up to 192K in I2S input mode. For up to 384K the data has to be both simultaneous and in offset binary form.
dear rfbrw,
U right, I also check the data sheet already. But I was afraid that 1541a running over 96Khz the chips will too hot, more digital noise & unstable. So I thought running in 96Khz will be better!
BTW, I would like to try use one CS8414 than pass 2 CS8420 with 192Khz (16 bit) to try in I2S mode. But I thought I will not success. Try this experiment was interesting!
Any suggestion for this??
TDA1541a DAC hadn't running 96Khz or higher mode before. I would like to try.
thx
thomas
U right, I also check the data sheet already. But I was afraid that 1541a running over 96Khz the chips will too hot, more digital noise & unstable. So I thought running in 96Khz will be better!
BTW, I would like to try use one CS8414 than pass 2 CS8420 with 192Khz (16 bit) to try in I2S mode. But I thought I will not success. Try this experiment was interesting!
Any suggestion for this??
TDA1541a DAC hadn't running 96Khz or higher mode before. I would like to try.
thx
thomas
tube-lover said:
I don't see any problem running the TDA1541A at 96k, that would be the maximum for the I2S input mode. At 192K the maximum value for BCK would be exceeded.
rfbrw,
thanks for your comment.
I would like to run 1541a in 96khz & test for several weeks for its stability. Any more noise & actual sound quality.
I will test this mode in both standard 1541a & 1541a S1 same time.
This will easy to upgrade for normal non-oversampling DAC with added the CS8420 & some parts.
I think 192K will not stable although run with S1.
BTW, this will be interesting, I can hear & try to compare 16,20,24 bits 96 Khz sound quality. Too high ~KHz is not my cup of tea.
thx
thomas
thanks for your comment.
I would like to run 1541a in 96khz & test for several weeks for its stability. Any more noise & actual sound quality.
I will test this mode in both standard 1541a & 1541a S1 same time.
This will easy to upgrade for normal non-oversampling DAC with added the CS8420 & some parts.
I think 192K will not stable although run with S1.
BTW, this will be interesting, I can hear & try to compare 16,20,24 bits 96 Khz sound quality. Too high ~KHz is not my cup of tea.
thx
thomas
Hi Thomas,
Kuei Yang Wang is right. Once you alter the sampling rate using CS8420 (or other SRC device), the CS8420 is also acting as a FIR filter by itself. So, there is still time smearing. AD1896 and SRC4192 datasheets cover more about their architecture and filter performance.
The advantage of SRC is the asynchronous reclock that remove low frequency jitter.
Kuei Yang Wang is right. Once you alter the sampling rate using CS8420 (or other SRC device), the CS8420 is also acting as a FIR filter by itself. So, there is still time smearing. AD1896 and SRC4192 datasheets cover more about their architecture and filter performance.
The advantage of SRC is the asynchronous reclock that remove low frequency jitter.
hi banana,
so U will advise me to use re-clock on the 1541a or use cs8420 run with 96Khz for 1541A!!
pls advise!
thx
thomas
so U will advise me to use re-clock on the 1541a or use cs8420 run with 96Khz for 1541A!!
pls advise!
thx
thomas
Hello Thomas,
Ofcourse Kuei Yang Wang and Banana is right! Any sample rate conversion uses digital filters (FIRs - because they have linear phase characteristic) - look at page 15 of CS8420 datasheet.... You can use 8420 in SRC bypass mode but there is no sense to use it in your circuit, it has sense only when 8420 works as spdif receiver. Diffirence beetween sound in CS8414 + 8420 and CS8414 + DF1704 is the result of diffirent FIRs in 8420 and DF1704 (diffrent filters coeficients, number of taps). Anyway both combinations are surly NOT non-os.
@Banana - ASRCs remove high frequency jitter due to use analog PLL or digital servo loop (both uses LPF in the feedback...)
Sorry for my english...
Regards,
Marek
Ofcourse Kuei Yang Wang and Banana is right! Any sample rate conversion uses digital filters (FIRs - because they have linear phase characteristic) - look at page 15 of CS8420 datasheet.... You can use 8420 in SRC bypass mode but there is no sense to use it in your circuit, it has sense only when 8420 works as spdif receiver. Diffirence beetween sound in CS8414 + 8420 and CS8414 + DF1704 is the result of diffirent FIRs in 8420 and DF1704 (diffrent filters coeficients, number of taps). Anyway both combinations are surly NOT non-os.
@Banana - ASRCs remove high frequency jitter due to use analog PLL or digital servo loop (both uses LPF in the feedback...)
Sorry for my english...
Regards,
Marek
tube-lover said:
The TDA1541A will run at up to 8Fs but not without a bit more work.
But TDA1541A is only 16bit.
dear all,
I also agreed banana & ciao comment, but the digital filter of 8420 looks only for re-sampling. The function was not same as the DF1704 or 1706. Certainly CS8420 is I2S but 1704 was SPDIF.
BTW, this is only for my test & prepare to print a samll PCB for my kit user to enjoy more.
thx
thomas
I also agreed banana & ciao comment, but the digital filter of 8420 looks only for re-sampling. The function was not same as the DF1704 or 1706. Certainly CS8420 is I2S but 1704 was SPDIF.
BTW, this is only for my test & prepare to print a samll PCB for my kit user to enjoy more.
thx
thomas
Dear marek & banana,
I had several questions & I hope to get some advises from U.
1. I was finished the re-clock of my Twin TDA1541a already.
Do U think that this is a good choice for upgrade for the NON-oversampling DAC?
2. About Mix using with chips. If I need to build a 24/96Khz DAC.
Will U agree that use CS8414+8420>>DF1704 ( 24.576Mhz)>>PCM1704 will be a good choice?
or only CS8420+DF1704+PCM1704!!!!
which one will be better!
I hope can build a DAC for Diyers which combine two kits in one chassis for interesting playing.
thx
thomas
I had several questions & I hope to get some advises from U.
1. I was finished the re-clock of my Twin TDA1541a already.
Do U think that this is a good choice for upgrade for the NON-oversampling DAC?
2. About Mix using with chips. If I need to build a 24/96Khz DAC.
Will U agree that use CS8414+8420>>DF1704 ( 24.576Mhz)>>PCM1704 will be a good choice?
or only CS8420+DF1704+PCM1704!!!!
which one will be better!
I hope can build a DAC for Diyers which combine two kits in one chassis for interesting playing.
thx
thomas
Dear marek,
About the 24/96 DAC.
I was ask for NPC already.
But SM5847 & 5849 is tooooooooooo expensive for diyers.
any advise except the DF1704 & 1706!
thx
thomas
About the 24/96 DAC.
I was ask for NPC already.
But SM5847 & 5849 is tooooooooooo expensive for diyers.
any advise except the DF1704 & 1706!
thx
thomas
Hello Thomas, are there any upgrades available for the Dual DAC yet?
Your Website is pretty quiet.
Regards
Anthony
Your Website is pretty quiet.
Regards
Anthony
hi Thomas, I guess you mean whether to use CS8420 or not.tube-lover said:
Personally, I don't like the idea of ASRC. Synchronous oversampling to an integral multiple is the better way, IMO.
Dear Banana,
I agreed that Synchronous was better than ASRC.
But How can I do that, Pls advise!!!!
Dear Anthong,
The RE-clock was funished already. I will post to forum ASAP.
The new versions of kit 24/96 & 1541a non-oversampling was coming which can combine with Op-amp ( multi use with different op-amp as AD811,844 & OPA627 which private set up by user).
Tube buffer can combine use with WE 412a, EZ80/81 & 6X4 tubes.
The digital input will use PE-65612 pulse input transformer.
Added one Schaffner filter for the power input!!
All caps was same OS-com, Blackgate , Pureism BP & Elna non-polar caps.
There still had some improve but I will keep secrets first & will inform later.
Mainly was the price approx the same.
thx
thomas
I agreed that Synchronous was better than ASRC.
But How can I do that, Pls advise!!!!
Dear Anthong,
The RE-clock was funished already. I will post to forum ASAP.
The new versions of kit 24/96 & 1541a non-oversampling was coming which can combine with Op-amp ( multi use with different op-amp as AD811,844 & OPA627 which private set up by user).
Tube buffer can combine use with WE 412a, EZ80/81 & 6X4 tubes.
The digital input will use PE-65612 pulse input transformer.
Added one Schaffner filter for the power input!!
All caps was same OS-com, Blackgate , Pureism BP & Elna non-polar caps.
There still had some improve but I will keep secrets first & will inform later.
Mainly was the price approx the same.
thx
thomas
hi Thomas,tube-lover said:
Synchronous sample rate conversion, that's what the 8X oversampling in DF1704/1706 are about.
44.1->352.8KHz
Running the TDA1541 at 8x is not straightforward, assuming it will run at all. 352K8 is well above the maximum value specified for LE. It also needs additional circuitry to reformat the data when connected to the DF1704/6 or the SM5842/3/7.
Why the dual dac, Coulomb?
I asked this question on the Thomas' other dead thread earlier today, but anyway - did you build the dual version to have fully balanced output?
Coulomb said:
I asked this question on the Thomas' other dead thread earlier today, but anyway - did you build the dual version to have fully balanced output?
- Status
- Not open for further replies.