ShinyMetal

leo · Joined: 12 Jun 2007 Posts: 122 Location: UK

Would it be better to use lower values for both C32/34 and C33/35? C32 and C33 are 56pf, C34 and C35 is 1N.
Would 1N in these places be a bit high?
_________________
Leo

6h5c · Joined: 13 Aug 2007 Posts: 105

RichA · Joined: 11 Jun 2007 Posts: 114 Location: Stafford, UK

Full schematic here:-

http://www.megafileupload.com/en/file/72170/192K-pdf.html

I took the plunge and ordered a Zero DAC... Surprised

--Rich

6h5c · Joined: 13 Aug 2007 Posts: 105

Well, at least it's not a Twisted Pee... ;-)

I think it's a nice DAC Rich. Thanks for the link.

Anyone noticed the forum was down during the weekend?

Ray

RichA · Joined: 11 Jun 2007 Posts: 114 Location: Stafford, UK

This forum was down for me as well. Smile

--Rich

6h5c · Joined: 13 Aug 2007 Posts: 105

Well, I simulated the CS4397 DAC's output filter, and the curve looks like the dutch dunes Smile

. No wonder, with 1n and 56p where these should be the same.

I adapted the filter to make the curve as straight as possible. If you change C32 and C34 both to 150pF, and R8 and R14 to 30k, the curve is almost straight. There's virtually no boost in the high range (only about 0.2dB), and <0.1dB loss at 20kHz. It's not a perfect Butterworth, but it's very close and with the least changes to the other components.

Regards,

Ray

leo · Joined: 12 Jun 2007 Posts: 122 Location: UK

6h5c · Joined: 13 Aug 2007 Posts: 105

Hi Leo,

I also simulated a perfect Butterworth (of course... Smile

, after recalculating the entire filter, but the difference is very small. Group-delay is a bit smoother, and the amplitude curve is a bit less steep, but personally i'm not going to take the extra trouble to change all the components for that. If you'd like to, however, I can give you the values.

Strange circuit, with the LT1057. What's that 3k92 doing there? Why isn't it symmetrical?

Regards,

Ray

leo · Joined: 12 Jun 2007 Posts: 122 Location: UK

Hi Ray,

Its supposed to properly load the dac, I thought it was a little strange but decided to try it anyway seeing as I had the bits spare Laughing

http://www.diyaudio.com/forums/showthread.php?s=&threadid=77992&perpage=25&pagenumber=13

I can easily remove the parts and install the values for the Butterworth using either the original circuit plus your alterations to C32/34 and R8/14 or your perfect Butterworth

All the parts have to come off anyway so I don't mind trying either Wink

The thing I do like about this cheap dac pcb is that the pads seem quite good and don't lift as easily as a lot of boards

Cheers,
Leo
_________________
Leo

6h5c · Joined: 13 Aug 2007 Posts: 105

See post 280:

'You can probably ignore R9 in most applications - it makes sense if using ac coupling with reasonably well matched electrolytics to maximize cmrr.'

Regards,

Ray

6h5c · Joined: 13 Aug 2007 Posts: 105

Hi Leo,

Here are the values I calculated for a Butterworth curve:

- R15/16/18/19: 19.74k (was 22.1k)
- C36/37/38/39: 1n (not changed)
- R8/9/14/17: 29.62k (was 22.1k & 36k)
- R10/11/12/13: 3.95k (was 6.8k)
- C32/33/34/35: 150p (was 56p & 1n)

These values give the same gain as the original circuit.
-3dB is a little under 38kHz.

Regards,

Ray

leo · Joined: 12 Jun 2007 Posts: 122 Location: UK

Hi Ray,

Thank you, I don't have any of those values yet so will try and get some ordered, I do still have the original parts for the standard output stage saved so can fit those first and alter the few weird values for R8/14, C32/34.

I do have 30k resistors R8/14 but unfortunately don't have any 150pf, all I have is 100pf and 220pf

Cheers,
Leo
_________________
Leo

leo · Joined: 12 Jun 2007 Posts: 122 Location: UK

Until I can get the parts to change the output stage I tried some different decoupling caps

The 100uf decoupling for the CS8416 and CS4397 was replaced with 100uf Panasonic FM (C16,C17,C18,C19 on schematic)

Pin25 CMOUT fitted with a 10uf Oscon SP

C20,21 before and after 3,3v reg fitted with 10uf Tants

decoupling for +/-v of op-amp (I'm using AD826 ) 2.2uf tants

This has changed the sound to very tight and warm, funny thing is I can't decide if I like it or not Laughing

Anyone else tried the panny FM's?

Also noticed the type of cap used on the CMOUT has quite an influence so maybe the Oscon SP isn't the best to use here
_________________
Leo