Another PCM2707 problem...

Status
Not open for further replies.
Akira,

That's good news - glad it's working - my DAC came presoldered on board so unless ESD damage before soldering? I will replace it - just waiting for 2706 samples to arrive.

Another simple DAC possibility is the TDA1387 - an I2S version of TDA1545 (although easier to implement than 1545 chip) which has been loosely compared to TDA1541 as in the same sonic character - analogue like. Slightly lower THD than 1543 -88dB Vs -75dB

Only two issues with this chip -

It is probably only available in smd - although your resoldering of PCM chip leads me to believe that is not a problem for you.

Output is lower than 1543 and will require active stage for I/V (this is also recommended for 1543 for best results)

John
 
Hi John
Even if it's already soldered to board, still is possible to ESD damage occurs while soldering the other components, isn't it?
You are lucky, TI dont send samples to this part of the world :bawling: .

Interesting this TDA1387.
I think it will not be difficult to obtain this DAC in Brazil. 😎


akira
 
Akira,
Any update on your DAC?

I killed another chip through reversing D+ & D- Will have to replace chip.

Saw this post on DIYHiFi
I think on the PCM2706 the problem is with the dac. My best results with that part have been in the 50's for jitter, using batteries and seperate supplies for VDD, VCCP and power to the 3ps OSC (no power to the dac section). I think if you could totally disable the dac this maybe a pretty good part.

Has anyone else tried this?

John
 
Hi John!

I'm (still) working on a design based in PCM2707 -> I2S -> 4 x TDA1543. A much-delayed project... but I think I will finish it in a month or so. 😀

preview

...and power to the 3ps OSC (no power to the dac section).
I don't understand this part. What's "3ps OSC"?

And how's the AW-3D's TDA1543, did you got it working?

Regards
Akira
 
Hi Akira,
don't understand this part. What's "3ps OSC"?

I am assuming that this is the oscillator (crystal) with 3ps jitter spec

I sent my AWD3 board back to AW for repair - which they kindly offered to do free of charge. It took quite a time & got it back a couple of days ago. Believe it or not while wiring it up for TDA1543 output I stupidly switched the D+ & D- USB input signals & blew the chip.

At least this time I know what blew the chip & will set about cutting it off the board & resoldering new chip in place


Brent - looked at your site - nice looking board & good reviews - what is the approx price of the stuffing parts for the board?

John
 
Very impressive board, Brent!

Unfortunately board swap will not be possible... because I plan to etch it myself. :apathic:
Here where I live, the cost to have a few boards manufactured is too high for a poor hobbyist like me to afford. :dead:


Regards
Akira
 
DAC cost

Hi jkeny,

The cost for parts for my DAC is around $350 US. Which sounds high but it is really two separate circuits on one board and there are some expensive parts like Black Gate caps and OPA627 opamps.

Parts for just the DAC half of the board (there is a score line between the boards to separate them if one wishes or order half a board) is around $200.

The reason I include a Preamp/buffer is that a passive I/V stage sounds very lifeless without it. Having a volume control onboard makes it easy to connect directly to a power amp.

~~~~~~~~~~~~~~

On the subject of "dead PCM2707" chips... I am starting to think that some computers are more prone to killing PCM2706/7 chips. For example my computer is incapable of killing a PCM2706/7 chip but I recently sent a DAC to a fellow and while it was playing the power cord to the DAC fell out and the PCM2706/7 is now dead. A $20 USB hub may be one way to solve this problem, if it is a real problem.
 
Re: DAC cost

Brent Welke said:
On the subject of "dead PCM2707" chips... I am starting to think that some computers are more prone to killing PCM2706/7 chips. For example my computer is incapable of killing a PCM2706/7 chip but I recently sent a DAC to a fellow and while it was playing the power cord to the DAC fell out and the PCM2706/7 is now dead. A $20 USB hub may be one way to solve this problem, if it is a real problem.

Or a $0.25 AND gate.

Why on earth would a hub make a difference, and isn't it just as likely, even if your thesis is correct that certain USB supplies are more likely to cause damage, that the hub would be more dangerous than the computer itself? And how would you ever figure this out?
 
Brent,
Looking at your schematic you have done what I quoted above :
I think on the PCM2706 the problem is with the dac. My best results with that part have been in the 50's for jitter, using batteries and seperate supplies for VDD, VCCP and power to the 3ps OSC (no power to the dac section). I think if you could totally disable the dac this maybe a pretty good part.

Is leaving out power to internal DAC of PCM chip noticeably better than all not in self powered mode i.e. is this worth the effort or is running an external 5V (not from USB power) through the Vbus & letting the chip regulate internally to 3.3V OK?

I intend to try some configurations to find the best but as I have a tendency to blow chips i want to limit the options to only the worthwhile ones.

BTW, has anybody tried other active I/V output stages - I'm thinking of trying OPA660 stage, discrete I/V stage & John Swenson's stage based on Gary Pimms CCS. Anybody got experience with these?

John
 
USB hub

Hi dsavitsk,

If my thesis is correct and I find an inexpesive USB hub that is incapable of damaging the PCM2706/7 then I would include it in my DAC kits.

I don't want to mess with current board (next version it would be included) but now that your question has me thinking more about it, I should at least see how simple it would be to attach your fix to my board and test it out.

I am a lazy man, would you mind pointing me to your prior posts regarding what exactly you recommend? For example what (DigiKey) part and so on.

Thanks,
Brent
 
jkeny,

I have not tried what you suggest for 3.3 source or I/V.

When I tried the clock oscillator (recommended by John Swenson) it made such a big difference over a crystal that I had to keep it. The clock requires a 3.3 V supply so a 3.3 V supply is required in my opinion. Two if you want to have some isolation between the PCM2706/7 and the clock.

Cheers,
Brent
 
Hi all,

Does anybody have one or two of the low jitter 12mHz Crystek oscillators to mail to Ireland - will pay for parts & postage or trade if preferred.

Can't seem to find any European distributor who will accept order less than 100 & shipping from Mouser is $31.

Here's hoping

John
 
Status
Not open for further replies.