Fin said:
What about "leveling" of I2S lines to TDA:?http://www.diyaudio.com/forums/showthread.php?postid=374865#post374865
tubee said:
Hi Guido
Nice you have the PMD working. Wich glue logic schematic do you use, the suggested one (I2S) from pmd datasheet?
What did you pay for a single PMD100 if i may ask?
On the inputside it is the suggested i2s schematic with the ff.
On the outputside there is nothing like i2s. It's left and right separated. The 1541 is in simultanious mode. All that is needed really is to create LE with two ff's.
I have had it working in hardware mode, but now i'm using software mode and am experimenting a bit with ditherlevels and 4x vs 8x os.
I'm using DG for DEM reclocking
I paid 20 USD, second hand on us ebay
guido said:
On the inputside it is the suggested i2s schematic with the ff.
On the outputside there is nothing like i2s. It's left and right separated. The 1541 is in simultanious mode. All that is needed really is to create LE with two ff's.
I have had it working in hardware mode, but now i'm using software mode and am experimenting a bit with ditherlevels and 4x vs 8x os.
I'm using DG for DEM reclocking
I paid 20 USD, second hand on us ebay
Hi Guido..
Dem with DG stands for? Or is it BG?
I removed the 176KHz dem with HC163 dividers in CD304mk2 and replaced the 470pF again. Why? Had some strange noises in treble now and then. With dem cap the sound is some warmer in lowmid (more 2nd harmonic??) but cleaner in treble. The 304 surprises me still on the sound, but had to do a lot on it.
So you use a TDA1541 on the PMD, interesting. You ordered some PCM63's too, isn't it?
I had yesterday a fast glance in PMD's datasheet, output looks more suitable for PCM dacs, or dual 1541, maybe differential.
$20 is a reasonable price for experiments, i want one PMD too to experiment with. New ones can be bought for a 5 times more, don't know if it's reasonable anymore.
I came to the conclusion nonos is not the way to go for me, with a TDA at least. Maybe some or other PCM will do better nonos, but till now i like OS better, longterm listening sessions reveiled that. But will hook a tube stage on my PCM56 soon, thus keep trying.
tubee said:
Hi,
No it's DG for the DG pin on the PMD: the deglitch signal. No need for a divider on the clock etc. DG through a ff to line up with LE. Then via cap and resistor to the dem pins, using both inverting and non-invertering output of the ff.
If you mean black gate with BG, i don't do voodoo
No strange noises here. The DG freq is too low for a full dem cycle, but at 8x the dem freq would be too high anyway to run the dem logic in the tda.
Yes, I got four pcm63 dacs also. Enough to keep me busy for a while. And another pmd for that project.
Can't follow your comment on the output, have another look
guido said:
Hi,
No it's DG for the DG pin on the PMD: the deglitch signal. No need for a divider on the clock etc. DG through a ff to line up with LE. Then via cap and resistor to the dem pins, using both inverting and non-invertering output of the ff.
If you mean black gate with BG, i don't do voodoo
No strange noises here. The DG freq is too low for a full dem cycle, but at 8x the dem freq would be too high anyway to run the dem logic in the tda.
Can't follow your comment on the output, have another look
The noises come from a not perfect filtered dem signal, ECdesigns filters hf out before it enters the 1541.
Now i understand, its a deglitcher. A second time i did'nt understand this abreviation. (Hi Bernard?)
I tried some "highly rated" Black Gate N types, they ended up in a big "leftover bin" with other electronic non usable or spare parts....
Some standard BG's found their place over Phono's cathode Resistors. Quite happy with them.
The comment on the output i did after a fast look in pmd datasheet, so i thought the pmd would be best with a PCM like format
Are you serious?
Hi Musicomputer,
Thanks for the link. I did do a search, and managed to understand it could be related to FIR coefficients, besides other things, but didn't find that explanation.
So, the C variant is not applicable, but the B is an improvement on the A. Cool.
Well, first of all, I shall have to build a very basic 7220 board that I can place in the path of my I2s before it reaches the TDA1541A.
Looking at the datasheet, I should be able to just send in I2S (including EFAB), pull up/down the other inputs so they are deemed unused, apply a clock and power and it should just 'work'. I haven't missed anything have I (like I need to initialise or configure something )?
Anyway, that'll be my project for the week - make a very crude board using the SAA7220, and just see if it works.
Cheers,
Phil
Thanks for the link. I did do a search, and managed to understand it could be related to FIR coefficients, besides other things, but didn't find that explanation.
So, the C variant is not applicable, but the B is an improvement on the A. Cool.
Well, first of all, I shall have to build a very basic 7220 board that I can place in the path of my I2s before it reaches the TDA1541A.
Looking at the datasheet, I should be able to just send in I2S (including EFAB), pull up/down the other inputs so they are deemed unused, apply a clock and power and it should just 'work'. I haven't missed anything have I (like I need to initialise or configure something )?
Anyway, that'll be my project for the week - make a very crude board using the SAA7220, and just see if it works.
Cheers,
Phil
philpoole said:Hi Musicomputer,
Thanks for the link. I did do a search, and managed to understand it could be related to FIR coefficients, besides other things, but didn't find that explanation.
So, the C variant is not applicable, but the B is an improvement on the A. Cool.
Well, first of all, I shall have to build a very basic 7220 board that I can place in the path of my I2s before it reaches the TDA1541A.
Looking at the datasheet, I should be able to just send in I2S (including EFAB), pull up/down the other inputs so they are deemed unused, apply a clock and power and it should just 'work'. I haven't missed anything have I (like I need to initialise or configure something )?
Anyway, that'll be my project for the week - make a very crude board using the SAA7220, and just see if it works.
Cheers,
Phil
Maybe it's an idea to check a schematic of a Philips, Marantz, or Magnavox CDP using a 7220/1541, and compare used connections. But you want only the FIR isn't it?
Important for the 7220 is: proper clock and proper PS. (preferable a shunt PS?) Good luck!
Hi Tubee,
Well, I made that admittedly sweeping statement after reading the data sheet.
Basically, I reckon (although I am willing to be proven wrong - probably by the circuit I'm building) SCAB, MUTE* and ATSB* can all be pulled up to 5v and then I need to feed in I2S in and feed the I2S output to my DAC, provide power and ground, and a clock signal.
This is just a quick experiment for now, and I don't have the bits at the moment for another Kwak clock, so I'm just going to use an 11Mhz crystal, 2x 10pF (I don't have 33pF) and a 10M (I don't have 1M) resistor. If that doesn't work then I'll have to get some other values (the caps I have are taken from a CD63 that ran on 16Mhz, so it might not work).
I've built it its own regulator stage and PI filter - which is a vague step in the right direction.
All I want is to prove the theory for now. If it works, then I shall rebuild with on onboard KC of sorts.
I'll hopefully try it out in the next few days (if it works! LOL) and let you know what its like.
Cheers,
Phil
Well, I made that admittedly sweeping statement after reading the data sheet.
Basically, I reckon (although I am willing to be proven wrong - probably by the circuit I'm building) SCAB, MUTE* and ATSB* can all be pulled up to 5v and then I need to feed in I2S in and feed the I2S output to my DAC, provide power and ground, and a clock signal.
This is just a quick experiment for now, and I don't have the bits at the moment for another Kwak clock, so I'm just going to use an 11Mhz crystal, 2x 10pF (I don't have 33pF) and a 10M (I don't have 1M) resistor. If that doesn't work then I'll have to get some other values (the caps I have are taken from a CD63 that ran on 16Mhz, so it might not work).
I've built it its own regulator stage and PI filter - which is a vague step in the right direction.
All I want is to prove the theory for now. If it works, then I shall rebuild with on onboard KC of sorts.
I'll hopefully try it out in the next few days (if it works! LOL) and let you know what its like.
Cheers,
Phil
Well, I have to say that I still prefer NOS.
However, this is mostly because my circuit isn't quite working yet LOL!
I can hear the music, but there is horrible interference.
I currently suspect its because of my incorrect clock component values.
So, I'll be ordering some parts off of Farnell to do it properly one day. Might be a week or so.
Watch this space for further comedy.
However, this is mostly because my circuit isn't quite working yet LOL!
I can hear the music, but there is horrible interference.
I currently suspect its because of my incorrect clock component values.
So, I'll be ordering some parts off of Farnell to do it properly one day. Might be a week or so.
Watch this space for further comedy.