Experience with this DIY DAC ?

[Kevin Wood]

Hi Chris,

Not a hijack. Perfectly on topic as far as I can see.

The current requirements are pretty modest although I haven't measured them myself. Most of the power is probably consumed by the irritating blue LEDs on each rail. I will remove them next time I've got it apart to save my retinas.

I used a couple of little toroids from Rapid electronics. Part no. 88-2502 (9+9v 15va) for the 5v supplies and 88-2506 (15+15v 15va) for the +/- 12v supplies.

You can connect these directly to the terminals on the board.

Kevin

[hafp]

Have a schematic of a similar DAC I got a few months ago on eBay.
It may not be exactly the same as the model you guys recently purchased, but is does use the CS4397/8416 chips and 2 opamps.
Would be interested in any successful mods you come up with on bypassing the second opamp and coupling caps.
PM me with your email if any of you want a PDF of the schematic.
I can't post is as the file is too large.
Cheers.

[johnm]

Just been looking at page 11 of the Cirrus datasheet - interesting that they appear to suggest using the NE5532 op-amp too, and in their top of the line DAC... I think I'll pop my 5532s in again and have another listen, though it sounds so good with the LM4562s I'm probably wasting my time.

Would it be worth while to replace the PSU diodes, and if so what would be an 'ideal' replacement?

- J

P.s. I too have the datasheet, but wasn't sure about the legalities of posting it here as its for a current product?

[Kevin Wood]

I have just found This rather useful application note (referred to by the CS4398 eval board application note) which details the design of low pass filters with differential input.

The filter stage posted on lampizator.eu didn't look right to me and has just occurred to me that the reason there is a DC offset at the output of the first Op-Amp stage after removal of the coupling capacitors is that the value of R14 is incorrect. To achieve common mode rejection down to DC it needs to be the same value as R8 (36K). The 14.7K resistor is then no longer required.

Looking at the application note they have made input impedance of the filter stage much lower (close to the 1K minimum AC load resistance specified for the CS4398) so maybe there is a noise advantage to be gained by recalculating all the component values around the filter stage accordingly. It would probably also depend on the final choice of Op Amp, its' bias current and output drive characteristics, of course.

Kevin

[johnm]

Interesting stuff - perhaps if we all agreed in an op-amp we could then proceed from there? I'm very happy with the LM4562s - what's everyone else using?

So if i changed R14 I wouldn't need 14K7, and could then run the 'pure' feed from the LM4562 to the output phonos?

[fredlock]

Currently I am using the OPPO970 as the 'transport' to feed into this DAC

Hi ah-luo-ko-ko,

Did you connect your DAC to the SPDIF out of your OPPO or did you tap it some where inside the circuit board? I'm just curious since I got the same Oppo player. Would you be able to email me the mods on the Oppo? I would really appreciated.

Fred

[Kevin Wood]

So if i changed R14 I wouldn't need 14K7, and could then run the 'pure' feed from the LM4562 to the output phonos?

Yep. (You'd also need to change R14's counterpart in the other channel, of course). You would also lose the output muting, unless you left the relay contacts in the output path. I would leave a low value resistor in series with each output in case you get stability issues or the output is accidentally shorted.

The fact that the eval board uses a 5532 may or may not be significant. It could have been done to show the quality achieveable with very low cost output parts. I note they use LTC1128s in the balanced output stage. I can't believe it would sound worse with a LM4562.

Kevin

[johnm]

Thanks for that Kevin - I think I'll try this route then.

Is output muting really necessary? I have been using an Audiosector/Peter Daniel NOS DAC for some time and that doesn't use one, and - to date - no problems yet using a PC as a source, or a CD player. Unless the TDA1543 has muting built into it of course?

[johnm]

OK, so for those wishing to implement the 'Op-Amp' mod - using an LM4562, these are the steps involved (please jump in anyone if there's any obvious mistakes!) :

1) Remove both NE5532s. Insert one (only) LM4562 in the first op-amp socket - the one nearest the center of the board.

2) Remove the following 4 capacitors: L+ / L- / R+ / R-. Replace with wire links.

3) Replace R14, and R17 (both 22k) with 36k.

4) Run a wire directly from the output pins (1 and 7) of the LM4562s to the phono sockets (via the output rely if you prefer this added safety measure).

- John

[chriscox]

Originally posted by Kevin Wood
Hi Chris,

Not a hijack. Perfectly on topic as far as I can see.

The current requirements are pretty modest although I haven't measured them myself. Most of the power is probably consumed by the irritating blue LEDs on each rail. I will remove them next time I've got it apart to save my retinas.

I used a couple of little toroids from Rapid electronics. Part no. 88-2502 (9+9v 15va) for the 5v supplies and 88-2506 (15+15v 15va) for the +/- 12v supplies.

You can connect these directly to the terminals on the board.

Kevin

Thanks Kevin.