[new project] The BIY ("Buy It Yourself") High End DAC - Page 2

The datasheets say very little. The WM8804 is by far the best availible today, trust me on this.

Here is an example.

That regulator is a text book copy of a super regulator, originally built by Walt Jung. You can find his original articles here:

http://waltjung.org/PDFs/Regs_for_High_Perf_Audio_1.pdf
http://waltjung.org/PDFs/Regs_for_Hi..._Audio_2_A.pdf
http://waltjung.org/PDFs/Regs_for_Hi..._Audio_2_B.pdf
http://waltjung.org/PDFs/Regs_for_Hi..._Audio_2_C.pdf
http://waltjung.org/PDFs/Regs_for_High_Perf_Audio_3.pdf

On what basis should I trust you ? Have you any jitter measurements lying around to back up your point ? "By far" is quite a statement...

There have been some measurements by spencer at diyaudio on the DIR9001 and he shows it to have "by far" the lowest jitter . I think his measurements were 50ps for the DIR9001.

On the basis that I probably know more than you?

Joking aside, I have aquired jitter measurements on the WM8804 from the good people at Wolfson. I do not believe they would appreciate if I share these however, you will have to email them and ask for yourself.

I will try to explain what makes the WM8804 different from almost every other SPDIF reciever out there.

The DIR9001 uses PLL to lock on to the incoming signal, a fairly standard PLL at that. This is the same methods as used by most other SPDIF recievers, including the CS8416. I will not go into how a PLL works, you have to read up on this yourself. The DIR9001 have an intrinsic jitter of 50 ps. That is, when it is fed a perfectly clean and jitter free signal, the output will have some 50 ps of jitter added by the PLL.

In reality however the input signal is never clean and jitter free. Quite the opposite in fact, several nanoseconds of jitter are quite common. The PLL will try to keep lock on the signal, constantly making small adjustments to the VCO. Regardless of how good the PLL is, a certain amount of jitter from the input signal will make it's way to the output. The lower the amount, the better the jitter rejection of the reciever. For example, under normal circumstances the DIR9001 might have some 200 ps of jitter.

The WM8804 is different. Rather than using an normal PLL, it uses a fractional-N PLL locked on to an external crystal. The speed of this PLL is adjusted by a second, digital control loop which then recovers the data. This offers far superior jitter rejection. So good in fact that unless the PLL looses lock, the output jitter will be almost indistinguishable from the intrinsic jitter of the PLL.

Now I've got a basis to trust you I confess I'm as defiant on the web as I'm trusting in real life.

I'll fire an email to wolfson. It's a bit strange that they didn't publish those graphs if they're so good. The lonely "50ps intrinsic jitter" lost in their datasheet doesn't tell much.

Is there a way to easily get wm8804 ? None of the big distributors carry it.

I see that you live in Belgium. Email Wolfson in the UK and ask for samples, they are very helpful.

Any update on this project? Kind of interested myself, but not until I finished my Opus build. Twisted Pear has a new TXD I/V stage that would be nice for this build I think.

Did the website work in english when you used it? I really like the case, but the site is terrible!

Try this link instead:
modu

Hifi2000 cases are nice, I've bought 11 of them so far

You're drawing too much power.