I don't feel like giving the chinese more money, can I pull the pcb out of my cd-65?

[freax]

thanks for getting back to me, I think I'll wait for you to do that test if thats alright with you to do it, that would simplify things a great deal.

The problem with the red baron is that its like half my income for a month gone just on some PCB's, and all for a lowly old TDA1541 non-A? sheesh..

I understand what youve done there in the schematic, I would leave the psu's in place for now tho.

It is my understanding that the CS8412 pin 11 FSYNC output goes to pin 4 DATA R/SCK input on the TDA1541 according to this schematic:
http://www.fetaudio.com/wp-content/u...10/TDA1541.pdf

Shouldnt be too difficult to put a 74HC4050 in-line should it.

[qusp]

fast logic chips for audio signals is one of the hardest things to get right for PCB layout actually.

also Thorsten gives that Red Barron a worse review than the Ebay dacs in that thread I linked and he doesnt think kindly of those....

[freax]

Originally Posted by qusp

fast logic chips for audio signals is one of the hardest things to get right for PCB layout actually

ah....

[qusp]

Originally Posted by freax

ah....

because it and its power supply demands comes directly from the RF layout manual, used for digital, but with a very analogue purpose, mixed signal design at its best, nothing that comes from one school of thought only will come up with the goods.

[freax]

Originally Posted by qusp

fast logic chips for audio signals is one of the hardest things to get right for PCB layout actually.

also Thorsten gives that Red Barron a worse review than the Ebay dacs in that thread I linked and he doesnt think kindly of those....

almost missed that, yes, it looks to me like someone has spent too much time in pcb cad software, especially the first pcb design.

I was reading up on thorstens comments earlier... mmm

I bet if I put the tda down onto some perfboard and point to point wire everything analog and pcb wire everything digital and then create little copper shield boxes for the digital pcb that maybe i could make a nice sounding dac.

Something to think about anyway.

[josha]

That pin 4's a bit odd, according to the 1541A datasheet it can be left unconnected but in most designs you see see it connected to pin 2.

On the 1541 datasheet it says to connect it to the system clock.

That said, in the player I have (same PCB as yours) with the original 1541 you can see that it's connected to pin 2 as per the A. I tried it both ways on the A and couldn't tell any difference!

There's also the question of the DEM cap, on the 1541 datasheet pins 17 & 18 are not connected but on our players we have a cap accross them as per the A version!! This cap makes a big difference on the A but the non-A? who knows?

It might be a few days before I get a chance to try the non-A in this config but I'll keep you posted!

Leaving the PSU's in place is fine if you're bypassing the 7210 & 20 as they're noisy, if you leave them in place then it's worth providing a clean +5v to the 7220, just an LM317 is ok - all the better if it's got it's own traffo.

The only issue I can see with the perfboard is the decoupling caps as they're close together - they're smd on our pcb's so it's not an issue, on perfboard you could go half on top, half underneath to keep them close to the dac?

Qusp - I'll have a look at that USB decoder thanks! I need to move on from the 2706, it was lying around from an old project so when the mech in the cdp packed up it was handy.

Have a great Christmas!

[freax]

I'll have to figure something out anyway, I've got insomnia again so I won't even be ready until tomorrow at least to start experimenting.

It certianly sounds exciting I know that much, I'm leaning towards leaving it all in the phillips pcb because I don't feel like making a new I/V stage at the moment, which means either giving the SAA7220 a signal or cutting it out of the loop.

I doubt the CS8412 will go high enough to get to 174.4KHz, 24bit.

[ryanj]

Hi Freax,

If your ok with etching PCBs, then you may just as easily make your own board for the 1541.

Here is my current setup with the TDA1541A, sounding great but still just a prototype. I will change the current grounding layout of the DAC and also integrate the CEN on the same board for the next version.

Ryan

[freax]

[EDIT] I think I have WS and BCK around the wrong way! lol, I'll check it and try again.

This is what I've come up with so far, there is only 3 wires connecting to the TDA's pins, LE/WS aka SCK and BCK aka FSYNC, and the third, SDATA aka DATA L and DATA R.

I don't have much experience with this so I'm working in the blind here, the CS8412 is producing a signal as I've verified it with my TDA1543 dac.

I lifted the old bridge wires and connected my own wires with the cable coming from the TDA1543 dac.

However it doesn't work, do I need to connect a ground lead aswell? I'm about to do that, then after that I'm going to lift that trace connecting pins 2 and 4 and give it a clock signal from the CS8412

[ryanj]

Originally Posted by freax

[EDIT] I think I have WS and BCK around the wrong way! lol, I'll check it and try again.

This is what I've come up with so far, there is only 3 wires connecting to the TDA's pins, LE/WS aka SCK and BCK aka FSYNC, and the third, SDATA aka DATA L and DATA R.

I don't have much experience with this so I'm working in the blind here, the CS8412 is producing a signal as I've verified it with my TDA1543 dac.

I lifted the old bridge wires and connected my own wires with the cable coming from the TDA1543 dac.

However it doesn't work, do I need to connect a ground lead aswell? I'm about to do that, then after that I'm going to lift that trace connecting pins 2 and 4 and give it a clock signal from the CS8412

Best way with ribbon cable is - GND-BCK-GND-WS-GND-DATA-GND

This way return current will be close to signal to ensure lowest impedance (for ribbon cable). A better choice is mini coax (U.FL for example).