DIY DAC info? (any kind)

[Possum]

I'm looking for some known-to-be-good designs for DIY DACs.

aos mentioned the SDS DAC from Sheldon Stokes, as did many posts in a deja.com search, but the linked webpage no longer exists (something like www.clarkson.edu/~sstokes/blah). I did get a glimpse of a finished SDS DAC, but forgot what that page was. It looks quite complicated and I expect it is very expensive to build.

Is there any such thing as a cheap (maybe at most being double the price of a CAL Gamma) DIY DAC that's better than the CAL Gamma, yet simple to build too? Maybe the simple part is being too wishful, since the IC's I've seen use 20+ pins at least, which should be hell to work with considering what tools I have.

Any suggestions?

 

[aos]

Possum, actually SDS labs has 2 DACs. I built the smaller one. It's not expensive to build at all. There are only two parts that are somewhat special to order, CS8412 receiver and CS4329 and they're not expensive. Total cost is low, I'd say $100-200 US at most. The other DAC is significantly more expensive plus the DAC chips are now obsolete. This is also the cheapest one I've seen so far

By the way, good part of this DAC is almost a carbon copy of Crystal's application notes for the chip. Digital section and power supply are Sheldon's but the analog section is really carbon copy. Rick commented that he doesn't like the sound of active filter. So I removed filter capacitors and it does sound better. It also saves you some money

. The sound also depends on the opamp.

 

[rickcr42]

Sheldon's new website , enjoy

http://www.quadesl.com/

 

[aos]

Thanks rick! I'll remove my own link to pdf now, especially since there's some new info on his site.

 

[ppl]

How do those one Bit Dac's compair to the 24 Bit ones, I have been told that the OneBit DAC's have week bass compaired to both the 16 Bit and the 24 Bit ones. TIA

 

[aos]

ppl, I haven't noticed lack of bass in any of my sigma-delta DACs. I don't have any R2R stuff yet, but that's about to change

.

 

[ppl]

Thanks aos. I have been considering that SDS Dac as a winter project. Mostly for use with an optical output to my SB live Card. Do you think this will result in Better sound than the DAC included on the card?

 

[aos]

ppl, without any question, even simple SDS DAC beats SB Live to the ground. CS4329 is a very good chip and Sheldon Stokes uses liberal amounts of voltage regulators. I added some high speed rectifier diodes and bypassed main capacitor bank with polypropylenes and removed capacitors from the output opamp filter but those are small changes. And I like this design because it's so compact, everything fits on the board, even the transformer. You can replace opamp with something better as well.

 

[ppl]

Thanks aos: I have the Optional Optical I/O card for the SB live but havent installed it yet. Also do you think it would improve the Technics DVDA40 DVD player using pannasonics one Bit dac?
A Tip regarding High speed Diodes like the MUR series and the Hexferds. While thay do improve the sound over your conventional 1n4000 series I stumbled by accident once and found Consistent improvement by using the Base collector juntion of High Current High ft TO-220 But transistors. Now you may LOL at this But as you know i am way into symmetry so i thought well heck if i use Complimentry Output transistors as Diodes base as the input Col as the Output then i would be able to Continue my pure complimentry topology all the way to the rectifiers. Well gess what A quite noticable improvement like a low level white noise was removed. This was in comparrison to those Ultra fast rect's. I tuhs went on a diode replacing bing and got consistent improvements. Now why this happened i cant say maby the complimentry thing maby the low leakage? I dono but try it you will be impressed. i used MJE-15030 & MJE15031 Transistors but i gess most any High current type will worl. The Amp rating is determond by the max. base current spec whitch is 2 amps in the MJE's case. I know I am crazy.

 

[aos]

ppl, I've seen people use transistor base-emitter junction instead of a diode as rectifier before, so I'm not going to call you crazy. Although, insisting on **complementary** transistors for that purpose is admittedly somewhat over the top

. Though I am really surprised if they outperform a dedicated, expensive ultrafast&ultrasoft rectifier diode. I personally found that the biggest difference comes from bypassing each "diode" with a 680pF or so monoblock ceramic capacitor.

I don't know how good is the Technics DVD-A player's DAC so I can't say how much (if any) better SDS DAC would be. I believe the most important thing with DACs is power supply, and the analog section.

 

[ppl]

Ya i would not have thought it that way also. as far as bypassing the diode i have done this for years, but it seems that with the ultrafast diodes that the Diode gets hot whereas it did' not before. But i understand that from a Technical aspect reducing the Diodes switching spikes should improve EMI. It dose in radio circuits. Most greatly in AM receivers. Anyhow try the Transistor method i discribed and let me Know your results. You can bend the Leeds of the TO-220 case to match a standard 1 amp diode spacing. As far as the complimentry rect's aspect. Ya i gess i am obsessive, But we dare to be different i gess.