DAC build TDA1541A/SAA7220P/B *will take som time*

I'll give you some advice which you can ignore if you want.
Use only schottky diodes. I use 11DQ06
These diodes are really cheap and much much clearer sounding than Glass passivated ultra fast diodes. They handle 1 Amp continuously and 25 Amp surge current. They are hardy enough to use on this DAC, and I use them feeding 22,000uF capacitors and no problems with them burning out.

If you use schottky diodes you wont need to use those 100nF caps across the diodes.

For the filter capacitors near the op-amps you should use Wima FKP2
They are very nice film and foil capacitors which result in a clean clear sound without smearing.
 
Last edited:
I'll give you some advice which you can ignore if you want.
Use only schottky diodes. I use 11DQ06
These diodes are really cheap and much much clearer sounding than Glass passivated ultra fast diodes. They handle 1 Amp continuously and 25 Amp surge current. They are hardy enough to use on this DAC, and I use them feeding 22,000uF capacitors and no problems with them burning out.

Thanks, of course I'll look into them 🙂
 
I'll give you some advice which you can ignore if you want.
Use only schottky diodes. I use 11DQ06
These diodes are really cheap and much much clearer sounding than Glass passivated ultra fast diodes. They handle 1 Amp continuously and 25 Amp surge current. They are hardy enough to use on this DAC, and I use them feeding 22,000uF capacitors and no problems with them burning out.

If you use schottky diodes you wont need to use those 100nF caps across the diodes.

For the filter capacitors near the op-amps you should use Wima FKP2
They are very nice film and foil capacitors which result in a clean clear sound without smearing.

The caps you are thinking of is 2,2nF and 1nF?
If so, I'll use FKP2 for 2,2nF and RIFA PHE450 for the 1nF.
 
I'll give you some advice which you can ignore if you want.
Use only schottky diodes. I use 11DQ06
These diodes are really cheap and much much clearer sounding than Glass passivated ultra fast diodes. They handle 1 Amp continuously and 25 Amp surge current. They are hardy enough to use on this DAC, and I use them feeding 22,000uF capacitors and no problems with them burning out.

If you use schottky diodes you wont need to use those 100nF caps across the diodes.

For the filter capacitors near the op-amps you should use Wima FKP2
They are very nice film and foil capacitors which result in a clean clear sound without smearing.

The 100nF WIMA MKP are allready in place across diodes. Couldn't hurt to leave them in. Is there such an audible difference between BYV27-200(which I have good experiences with and the schottky's?)
 
I have tried BYV diodes before, and I find the schottky to sound cleaner and better.

Yes I was talking about the 1nf and 2.2nf.

I find the lower current schottky diodes to sound better than the high current ones, when used in a low current circuit such as a DAC, or pre amp. Which is a bonus, as they are cheaper.
 
Actually the ones I use in my valve output stage are BYV26C which are Fast soft-recovery
controlled avalanche rectifiers, and I use these because they can handle high voltage.
I tried the Fairchild stealth diodes in the same location, but I didn't like their sound.
I did also try both stealth and BYV, diode in my DAC power supply but found Schottky to sound better.
 
What op-amps will you use? In this kit I used Texas Instruments variety of NE5534 which has a faster slew rate than other brands of NE5534, and consequently sound better.
But I'm not really an op-amp fan, but sometimes you need to use them, such as in this kit.
 
What op-amps will you use? In this kit I used Texas Instruments variety of NE5534 which has a faster slew rate than other brands of NE5534, and consequently sound better.

If you do a little math, the maximum possible slew rate when there's 2.2nF of feedback capacitance is around 2V/uS. That's for a 4mA step change in current. You might notice there's an improvement in sound quality by decreasing the value of that cap, but not so low a value that the opamp's slew rate is exceeded.

Since a unity gain compensated NE5534 has 6V/uS that suggests a value of 680pF.
 
If you do a little math, the maximum possible slew rate when there's 2.2nF of feedback capacitance is around 2V/uS. That's for a 4mA step change in current. You might notice there's an improvement in sound quality by decreasing the value of that cap, but not so low a value that the opamp's slew rate is exceeded.

Since a unity gain compensated NE5534 has 6V/uS that suggests a value of 680pF.

Mathematically you are correct. Sound-wise, which as we both know is subjective, I found the TI NE5534 to sound better than the Philips equivalent. The slew rate was one obvious datasheet difference, however I assume that there are other differences in the construction of the device which results in this sonic difference. Wasn't it you advocating ultra fast video op-amps for IV conversion?

I tried lower and higher values of that 2.2nf cap and actually found around 3.3nf to sound better to my ears, but mine is a non oversampling version. (incidentally I don't use this DAC anymore)
 
These came today. Os-con 22uF/20V.
22uF_20V_oscon_1.jpg


And after a little while...
PCB_populating_11.jpg


PCB_populating_12.jpg
 
Sound-wise, which as we both know is subjective, I found the TI NE5534 to sound better than the Philips equivalent. The slew rate was one obvious datasheet difference, however I assume that there are other differences in the construction of the device which results in this sonic difference.

I had a quick peek at the datasheets of those two manus, and both say 13V/uS and 6V/uS (uncomp and comp, respectively). But yes Doug Self found differences in distortion (admittedly on NE5532, not -4) - with TI the worst of the bunch - so there must be significant differences.

Wasn't it you advocating ultra fast video op-amps for IV conversion?

Yes - I got better results with LM6172 sonically than with NE5532, but this was with an AD1955, not TDA1541A. Then reducing the feedback cap made the difference greater.

I tried lower and higher values of that 2.2nf cap and actually found around 3.3nf to sound better to my ears, but mine is a non oversampling version. (incidentally I don't use this DAC anymore)

I don't play with mine any more either.
 
SAA7220 is a very early digital filter, it is not regarded very well for its sound. Apparently it generates lots of noise, and on a number of levels measures badly. Sorry.

As I understood the noise is introduced when the SAA7220 shares PS with the other parts.

I don't really care since I started building this DAC because I loved the sound from my very mildly, so far, Marantz CD-60 which has the same IC's.
 
Couldn't find any suitable 120nF radial MKP's so I ordered 2pcs Jantzen cross cap axial MKP's, will have to squeeze them in as TDA1541 decoupling.
Also ordered 4pcs 8,2nF WIMA FKP2, 4pcs 10nF WIMA FKP2, 10pcs 100nF WIMA MKP4, one RIFA PHE450 47nF, 4pcs 2,2nF WIMA FKP2, 4pcs 33nF WIMA FKP2, 2pcs 220nF RIFA PHE426, 2pcs 1uF RIFA PHE426, 2pcs 1nF RIFA PHE450.

All set for filmcaps I think.

Also ordered 10pcs Rubycon ZA 10uF/25V

Still undecided about 100uF(according to BOM) caps around regs.

Elna Cerafine 220uF/25V?
Elna Silmic II 220uF/25V?
Nippon Chemi-con Solid polymer 100uF/25V?

Other? Suggestions?