hi mike,
pls check the old message. I said the R1 will only use in the balance dac.
USD10 was for the basic 1541a only. BTW, if U did not like it, pls post me back & I will money back to U.
R1 grade I cannot maintain this price. the comingpric was USD17.5 per chips to me & the quality not too many, only 36 pcs only. Can match same date code only 20pcs.
so if berhard said R1 not good that will happy because I many many 1541a on hand.
Mike, I only can match same date code to U, but is all the chips all spec per chips were match together, I did not know. But this is all real chips, I will not sell remark chips. 1541AS2 was a good example, all the diyers bought & directly felt the performance was improve. why I prepare the full balance 1541a DAC, because I had AS 2 on hand, otherwise I will not do this.
thx
thomas
pls check the old message. I said the R1 will only use in the balance dac.
USD10 was for the basic 1541a only. BTW, if U did not like it, pls post me back & I will money back to U.
R1 grade I cannot maintain this price. the comingpric was USD17.5 per chips to me & the quality not too many, only 36 pcs only. Can match same date code only 20pcs.
so if berhard said R1 not good that will happy because I many many 1541a on hand.
Mike, I only can match same date code to U, but is all the chips all spec per chips were match together, I did not know. But this is all real chips, I will not sell remark chips. 1541AS2 was a good example, all the diyers bought & directly felt the performance was improve. why I prepare the full balance 1541a DAC, because I had AS 2 on hand, otherwise I will not do this.
thx
thomas
Hi MGH,
Thanks for your reply [post#792]
The octal D-I DAC setup is designed and optimized to drive 8 TDA1541A modules only.
The timing-chain is designed to handle 8 TDA1541A modules only.
Th current main power supply can't handle the double amount of TDA1541A's. The maximum available capacity of the encapsulated torroid transformers used is only 25VA.
DEM clock synchronizing would become difficult
The necessary modifications probably wouldn't fit in the current housing.
So it's not that easy.
About the printed circuit boards, I will discuss this with my printed circuit board manufacturer. But I would choose thick gold plated traces, transparent laquer and if possible a teflon based circuit board. The circuit boards are double sided, exept for the keyboard and LED indicator PCB's.
Thanks for your reply [post#792]
The octal D-I DAC setup is designed and optimized to drive 8 TDA1541A modules only.
The timing-chain is designed to handle 8 TDA1541A modules only.
Th current main power supply can't handle the double amount of TDA1541A's. The maximum available capacity of the encapsulated torroid transformers used is only 25VA.
DEM clock synchronizing would become difficult
The necessary modifications probably wouldn't fit in the current housing.
So it's not that easy.
About the printed circuit boards, I will discuss this with my printed circuit board manufacturer. But I would choose thick gold plated traces, transparent laquer and if possible a teflon based circuit board. The circuit boards are double sided, exept for the keyboard and LED indicator PCB's.
TDA1541A-$$
Hi all,
There seems to be some discussion about TDA1541A quality, again.
The same production date doesn't mean the chip's have the same specs, chances are they are not manufactured from the same wafer, and even a single wafer can have impurities that can affect individual chip specifications. Chip's are already classified, according to the measured parameters, regardless of production date. So if you order a R1, you get a TDA1541A that meets R1 classification, regardless of production date.
The marginal differences in TDA1541A quality are blown-up out of proportion. According to this the TDA1541A-R1 is so bad, it should be avoided. The double crown is the only one nearly good enough. But what are we talking about?
TDA1541A (standard): Edl < 1LSB ($10)
TDA1541A-R1 (relaxed): Edl < 2LSB ($17.5)
TDA1541A-S1/S2 (single/double crown): Edl < 0.5...0.75LSB ($90)
Yes I checked this many times now, R1 has Edl < 2LSB, R stands for "relaxed" specifications. Bernhard's measurement's seem to confirm this (thanks Bernhard). I have a relative that works at Philips semiconductor division in Eindhoven, I will ask him to verify this as well, as he knows one of the engineers that was involved in the TDA1541A design. I also added the selection specifications noted in the philips datasheet. It is dated february 1991, if the data contained in this datasheet was incorrect, they would have corrected it by now I guess.
So basically we are making all this fuzz, when the plain TDA1541A (cheapest one) already has a Edl of < 1LSB. As can be seen on the photograph of the analog mainboard [post#715], there is no R1 or S1 marking on the TDA1541A's. So if I am correct, all listening tests and measurements are based on the plain TDA1541A version. And it's sound quality is good enough for me.
Hi all,
There seems to be some discussion about TDA1541A quality, again.
The same production date doesn't mean the chip's have the same specs, chances are they are not manufactured from the same wafer, and even a single wafer can have impurities that can affect individual chip specifications. Chip's are already classified, according to the measured parameters, regardless of production date. So if you order a R1, you get a TDA1541A that meets R1 classification, regardless of production date.
The marginal differences in TDA1541A quality are blown-up out of proportion. According to this the TDA1541A-R1 is so bad, it should be avoided. The double crown is the only one nearly good enough. But what are we talking about?
TDA1541A (standard): Edl < 1LSB ($10)
TDA1541A-R1 (relaxed): Edl < 2LSB ($17.5)
TDA1541A-S1/S2 (single/double crown): Edl < 0.5...0.75LSB ($90)
Yes I checked this many times now, R1 has Edl < 2LSB, R stands for "relaxed" specifications. Bernhard's measurement's seem to confirm this (thanks Bernhard). I have a relative that works at Philips semiconductor division in Eindhoven, I will ask him to verify this as well, as he knows one of the engineers that was involved in the TDA1541A design. I also added the selection specifications noted in the philips datasheet. It is dated february 1991, if the data contained in this datasheet was incorrect, they would have corrected it by now I guess.
So basically we are making all this fuzz, when the plain TDA1541A (cheapest one) already has a Edl of < 1LSB. As can be seen on the photograph of the analog mainboard [post#715], there is no R1 or S1 marking on the TDA1541A's. So if I am correct, all listening tests and measurements are based on the plain TDA1541A version. And it's sound quality is good enough for me.
Attachments
hi john,
very thanks for your information.
Let me share little opionions of the different grade of the 1541a.
Actually total I had over 1300pcs on hand. Most was 1541a, only little was R1. R1 was another se;;er sell to me from taiwan which I hope to prepare for the balance DAC before I get the double crown. After I get the double crown, I use Westlake BBSM-6, TAD 2251 & yamaha NS-2000 to test the different between them . They test both in three area, three sets of different set up f equipments. The last one was last night. One of my friend back from Shangshai & he was use KT-1610 ( SE 60watts) to push JBL4344.
Very interesting was the bes one was S2.
S1 & R1 was very close, In all setup. R1,S1 & S2 was more silent than standard. Since I had only 36 pcs R1 on hand. I was so happy tha i will not lost too much money. So I think my remain stock standard 1541a will enough for your DAC lovers to enjoy your products.
hope the standard 'a' version will good in my balance too because S2 was shortage & too expensive.
as I said, I will only supply standard 'A' version chips to support john's project!!!! R1 berhard said not good, I will take away & forget it!!
very thanks for john's share information& berhard's test.
thx
thomas
very thanks for your information.
Let me share little opionions of the different grade of the 1541a.
Actually total I had over 1300pcs on hand. Most was 1541a, only little was R1. R1 was another se;;er sell to me from taiwan which I hope to prepare for the balance DAC before I get the double crown. After I get the double crown, I use Westlake BBSM-6, TAD 2251 & yamaha NS-2000 to test the different between them . They test both in three area, three sets of different set up f equipments. The last one was last night. One of my friend back from Shangshai & he was use KT-1610 ( SE 60watts) to push JBL4344.
Very interesting was the bes one was S2.
S1 & R1 was very close, In all setup. R1,S1 & S2 was more silent than standard. Since I had only 36 pcs R1 on hand. I was so happy tha i will not lost too much money. So I think my remain stock standard 1541a will enough for your DAC lovers to enjoy your products.
hope the standard 'a' version will good in my balance too because S2 was shortage & too expensive.
as I said, I will only supply standard 'A' version chips to support john's project!!!! R1 berhard said not good, I will take away & forget it!!
very thanks for john's share information& berhard's test.
thx
thomas
Thanks EC for your take on the different versions of TDA1541A. You're always the voice of reason. Also thanks for considering the teflon based PCB board for your finalized DAC.
Thanks Thomas for your reply. I don't mind the standard version of the TDA1541A at all since I got them for a reasonable price from you.
Thanks Thomas for your reply. I don't mind the standard version of the TDA1541A at all since I got them for a reasonable price from you.
octal D-I DAC modules
Hi adhoc,
Thanks for your reply [post#795]
> Basically, as noted earlier, I plan to offer printed circuit boards and housing parts.
> The X8 behind the DAC TDA1541A module indicates, eight single TDA1541A modules are necessary in total, currently I use two TDA1541A DAC PCB's that contain four single TDA1541A modules each.
> Next I will give a overview of modules used, * (completed) + (partially completed) - (not completed) ? (experimental).
MAIN POWER SUPPLY:
* 1 X main power supply module (for octal D-I DAC core and auxilary circuits)
OCTAL D-I DAC CORE (the actual D/A converter):
* 1 X analog mainboard (holds all following modules)
* 8 X single TDA1541A module
* 2 X I/V diff amp module
* 1 X dual format timing-chain module
* 1 X UHS buffer module (for buffering/inverting BCK)
TUBE AMPLIFIER:
* 2 X tube amplifier module
* 1 X tube amplifier power supply
INPUT INTERFACE:
+ 1 X differential I2S input selection module (2 inputs + USB)
+ 1 X USB interface module (with reclocking and isolation)
? tuned circuit jitter reduction module
? I2S replacement (WS/DATA decoder)
OUTPUT SELECTOR:
- 1 X analog output module (semiconductor, tube, mixed mode switch)
CONTROL:
+ 2 X keyboard module (illumination, I2S1, I2S2, USB, tube mode, OP-amp mode, MIX mode, power)
+ 3 X LED indicator module (illumination, I2S1, I2S2, USB - logo - tube mode, OP-amp mode, MIX mode, power)
+ 1 X system controller (controls keyboard, display, input switching, output switching and audio network interface)
Hi adhoc,
Thanks for your reply [post#795]
> Basically, as noted earlier, I plan to offer printed circuit boards and housing parts.
> The X8 behind the DAC TDA1541A module indicates, eight single TDA1541A modules are necessary in total, currently I use two TDA1541A DAC PCB's that contain four single TDA1541A modules each.
> Next I will give a overview of modules used, * (completed) + (partially completed) - (not completed) ? (experimental).
MAIN POWER SUPPLY:
* 1 X main power supply module (for octal D-I DAC core and auxilary circuits)
OCTAL D-I DAC CORE (the actual D/A converter):
* 1 X analog mainboard (holds all following modules)
* 8 X single TDA1541A module
* 2 X I/V diff amp module
* 1 X dual format timing-chain module
* 1 X UHS buffer module (for buffering/inverting BCK)
TUBE AMPLIFIER:
* 2 X tube amplifier module
* 1 X tube amplifier power supply
INPUT INTERFACE:
+ 1 X differential I2S input selection module (2 inputs + USB)
+ 1 X USB interface module (with reclocking and isolation)
? tuned circuit jitter reduction module
? I2S replacement (WS/DATA decoder)
OUTPUT SELECTOR:
- 1 X analog output module (semiconductor, tube, mixed mode switch)
CONTROL:
+ 2 X keyboard module (illumination, I2S1, I2S2, USB, tube mode, OP-amp mode, MIX mode, power)
+ 3 X LED indicator module (illumination, I2S1, I2S2, USB - logo - tube mode, OP-amp mode, MIX mode, power)
+ 1 X system controller (controls keyboard, display, input switching, output switching and audio network interface)
Hi Thomas,
Thanks for your reply [post#806]
I appreciate your effort to supply (TDA1541A) chip's for this and other projects, it's good to have a reliable source for this. As you can see in the specs below there is only 1dB difference at -60dB between the plain TDA1541A and the R1 version, that's all.
These should be the correct specs for the TDA1541A versions:
TDA1541A/S2 : Typ. THD+noise (0dB) -97dB, THD+noise (-60dB) -47dB, Channel separation 98dB, S/N ratio 112dB
TDA1541A/S1 : Typ. THD+noise (0dB) -95dB, THD+noise (-60dB) -47dB, Channel separation 98dB, S/N ratio 112dB
TDA1541A : Typ. THD+noise (0dB) -95dB, THD+noise (-60dB) -43dB, Channel separation 98dB, S/N ratio 112dB
TDA1541A/R1 : Typ. THD+noise (0dB) -95dB, THD+noise (-60dB) -42dB, Channel separation 98dB, S/N ratio 112dB
> I used the plain A version in my twin DAC (balanced/DATA inverted on one TDA1541A) with the 8th order Butterworth filter, it works fine, so your balanced DAC should work fine too using the TDA1541A. However when using only one or two DAC chip's, their specifications have more impact on sound quality of course.
> The main difference between TDA1541A versions is low level performance. So this does improve with both S1 and S2 versions, this also explains the better sound quality you heared, using these chip's.
I added the TDA1541A specs from philips that conflict with the TDA1541A datasheet
Thanks for your reply [post#806]
I appreciate your effort to supply (TDA1541A) chip's for this and other projects, it's good to have a reliable source for this. As you can see in the specs below there is only 1dB difference at -60dB between the plain TDA1541A and the R1 version, that's all.
These should be the correct specs for the TDA1541A versions:
TDA1541A/S2 : Typ. THD+noise (0dB) -97dB, THD+noise (-60dB) -47dB, Channel separation 98dB, S/N ratio 112dB
TDA1541A/S1 : Typ. THD+noise (0dB) -95dB, THD+noise (-60dB) -47dB, Channel separation 98dB, S/N ratio 112dB
TDA1541A : Typ. THD+noise (0dB) -95dB, THD+noise (-60dB) -43dB, Channel separation 98dB, S/N ratio 112dB
TDA1541A/R1 : Typ. THD+noise (0dB) -95dB, THD+noise (-60dB) -42dB, Channel separation 98dB, S/N ratio 112dB
> I used the plain A version in my twin DAC (balanced/DATA inverted on one TDA1541A) with the 8th order Butterworth filter, it works fine, so your balanced DAC should work fine too using the TDA1541A. However when using only one or two DAC chip's, their specifications have more impact on sound quality of course.
> The main difference between TDA1541A versions is low level performance. So this does improve with both S1 and S2 versions, this also explains the better sound quality you heared, using these chip's.
I added the TDA1541A specs from philips that conflict with the TDA1541A datasheet
Attachments
-ecdesigns- said:
It is obvious that there is an error in that specs table as one can see from left to right S2 S1 plain selection grades go down and R1 is the last so it must be the worst. -60dB specs are mixed up between R1 and plain here, that could be the reason that people think R1 is a + grade.
Oh, i see now i have to read a lot to catch up after my holiday.....
Solution: keep iMac!
I wonder if Linux-ubuntu 6.06 is bit-perfect.
And yes i was astonished of processor-size in the old iMac i have, 1.5*1.5 cm with 266Mhz and passive cooling! 386 clones with Intel chips is about 50 times greater and has to be cooled extensively.
Btw, do you play pcm like music files, or something like compressed Mp3's John? I like your automatic detecting I2S solution!
After reading i discovered: lossless compression. Glad its no mp3 rubbisch
Solution: keep iMac!
I wonder if Linux-ubuntu 6.06 is bit-perfect.
And yes i was astonished of processor-size in the old iMac i have, 1.5*1.5 cm with 266Mhz and passive cooling! 386 clones with Intel chips is about 50 times greater and has to be cooled extensively.
Btw, do you play pcm like music files, or something like compressed Mp3's John? I like your automatic detecting I2S solution!
After reading i discovered: lossless compression. Glad its no mp3 rubbisch
Hello EC,
Thank you for your reply (post #809).
Given that the vast majority of the DAC is finalised, perhaps it would be a good idea to release individual part lists for the various modules? It would be an excellent way for you to get more people involved in this project.
Or perhaps an all-in-one part list for whatever is currently completed would be great. Well - it would certainly be great for eager beavers like me who intend to start acquiring parts now.
Thank you for your reply (post #809).
Given that the vast majority of the DAC is finalised, perhaps it would be a good idea to release individual part lists for the various modules? It would be an excellent way for you to get more people involved in this project.
Or perhaps an all-in-one part list for whatever is currently completed would be great. Well - it would certainly be great for eager beavers like me who intend to start acquiring parts now.
tubee said:
Windows can output bit-perfect as long as your soundcard supports it. I'm for example using a E-mu 0404 soundcard which outputs bitperfect. Also, for Windows, I'd suggest you use the program foobar2000 for playing back files, which is THE program for audiophiles.
-Chris
!
Hi tubee,
I hope you had a good holliday.
off topic this time, but I think it's important since you mentioned you updated to the latest Ubuntu version yesterday (Dapper Drake), be carefull not to use today's xserver update!!! I did, and it has cost me almost half a day to figure out how to fix it.
It can be fixed by downgrading to the previous xserver version until the bug has been fixed. This can be done by using the following command:
sudo apt-get install xserver-org-core=1:1.0.2-0ubuntu10
Hi tubee,
I hope you had a good holliday.
off topic this time, but I think it's important since you mentioned you updated to the latest Ubuntu version yesterday (Dapper Drake), be carefull not to use today's xserver update!!! I did, and it has cost me almost half a day to figure out how to fix it.
It can be fixed by downgrading to the previous xserver version until the bug has been fixed. This can be done by using the following command:
sudo apt-get install xserver-org-core=1:1.0.2-0ubuntu10
Ubuntu
Thanks John for tip.
I simply updated via the update-manager and choosed new 6.06 version. I did not have any problems so far, but the updating took a long time, about 4 hours. I still don't know all in & outs of it, the printer won't work either. So have a Win pc upstairs to print with.
But i like this Linux version a lot, nice ikons and the extra programs to choose(for kids too) But its a tad slower also, bigger version. On this machine (503Mhz running 450 with 128mb ram), Xubuntu would be a better choice.
Thanks John for tip.
I simply updated via the update-manager and choosed new 6.06 version. I did not have any problems so far, but the updating took a long time, about 4 hours. I still don't know all in & outs of it, the printer won't work either. So have a Win pc upstairs to print with.
But i like this Linux version a lot, nice ikons and the extra programs to choose(for kids too) But its a tad slower also, bigger version. On this machine (503Mhz running 450 with 128mb ram), Xubuntu would be a better choice.
Bit perfect audio in Windows
One way to achieve bit perfect audio in Windows is to use the free ASIO driver http://www.asio4all.com/
ASIO4ALL works great with internal sound cards and USB audio devices using the PCM2706/7 chip and others.
You also need a music player that will output to the ASIO driver. Foobar and Winamp are known to work, iTunes and Windows Media Player are two examples that currently do not support ASIO.
On a dual PIII 500 I can play two USB DACs using two Foobar players at the same time as I go about my daily business on this computer with rarely a hiccup in the sound. I can even burn a DVD at the same time.
By the way... Currently Windows 2003 Server appears to be the best (most powerfull) Windows OS to use for Audio. I have a IBM laptop that would not run without pops and clicks until I installed Win2k3 on it. I spent a couple days and reinstalled WinXP on it before I gave up and put Win2k3 on it. Also if you have a IDE Promise controller card or on the motherboard then replace the Windows driver for it with the one from Promise website to solve pops and clicks in audio and make the machine more responsive.
One way to achieve bit perfect audio in Windows is to use the free ASIO driver http://www.asio4all.com/
ASIO4ALL works great with internal sound cards and USB audio devices using the PCM2706/7 chip and others.
You also need a music player that will output to the ASIO driver. Foobar and Winamp are known to work, iTunes and Windows Media Player are two examples that currently do not support ASIO.
On a dual PIII 500 I can play two USB DACs using two Foobar players at the same time as I go about my daily business on this computer with rarely a hiccup in the sound. I can even burn a DVD at the same time.
By the way... Currently Windows 2003 Server appears to be the best (most powerfull) Windows OS to use for Audio. I have a IBM laptop that would not run without pops and clicks until I installed Win2k3 on it. I spent a couple days and reinstalled WinXP on it before I gave up and put Win2k3 on it. Also if you have a IDE Promise controller card or on the motherboard then replace the Windows driver for it with the one from Promise website to solve pops and clicks in audio and make the machine more responsive.