Building the ultimate NOS DAC using TDA1541A

Horizontal: 50 us/div
Vertical: 10 mV/div (with x100 preamplifier)

Looks like glitch width is ~ < 5uS?

This may be bit switch related.

The glitch is most visible at bit 12, 13, 14, 15 transition.

So the MSB? If so this makes sense.

The bit-switches on the MSB's lack a specific design used on the lower bits.l used to reduce glitches.

Thor
 
The page that was posted by fabrice is from an article by a philips engineer, I read it and I remember that he connected the TDA1541A output directly to the vertical input of the CRT, bypassing the vertical amplifier. So the sensitivity shown at the scope screen is probably misleading, not the real amplitude of the glitch. However, on the article, it does seem to be a much larger glitch in amplitude than what @lcsaszar was able to see on his scope.

Either way this is very low energy, because the area is small. In my opinion this glitch (which is the "major carry", the worst case) is of little importance.
 
I played a bit with LTSpice...
 

Attachments

  • sinc_compensation.jpg
    sinc_compensation.jpg
    123.8 KB · Views: 105
Ok, looking around, I decided to make a few small changes to the design, here the single TDA1541 sensible "Core".

View attachment 1370578

Key changes:

BC327 added to buffer the clock inputs of the 74F74, to avoid loading the clock. BC327 should be fine as follower at 50MHz. Internally each input is a Pin Diode (think 1N4148) and a 10k Pullup. So 4 clock inputs become a material load, we just don't want those currents to circulate across our input wiring. This way it's the base of a follower as load, with still decent beta at many MHz. Emitter load is ~ 2.5kOhm, base load less.

The attenuators and slew rate limiters are now "flying". This makes them a constant load regardless if Q or /Q are high (the other output is then low). This makes the 74F74 reclockers and signal conditioning mostly a constant load, if a chunky one at nearly 50mA. But they can now share the TDA1541 +5V and DGND with little issue.

So this overall simplifies things. DEM Clock is "Balanced Philips/Grundig" again with a flying balanced attenuator.

It may seem all a bit OTT and gilding the lily. Perhaps. I'm being quite minimalist by my standards.

I'm actually tempted to bring the MCK on board too, as the last element outside our control:

View attachment 1370584

In a lot of ways what I propose is not miles off from Johns work around SD Trans with TDA1541 DAC era.

Thor
I'd like to building this clock circuit with a 4.2336 MHz crystal. First time i am building one. Do i neede to adjust something?
 
I can probably handle a cnc batch of "golden heart" type shielding to hand out. So it's shielded and decoupled physically from the board itself, on it's standalone board.

That's a good idea. Iancanada made cutting slots inside the PCB to make the crystal pcb floating inside. But anyway the board shoulld be current feeded and damped, so a coax very thin cable with very soft nylon body makes sense.

Like @Zoran said you want to avoid most possible you can some expternal vibrations to get there, hence the USB and chinch plugs on the cabinet and not directly soldered on the pcb as well.

Maybe to reduce the cost (cutting slot inside the pcb ?) you just can use two plastic rods for hanging the little clock pcb above the core pcb front end. and use a 1" ufl conection towards the main core board perhaps. Maybe overkill, I don't know.... Or better below the core board, suspensed to it, in order its power cable has no constraints, for instance lying on a foam sheet on the cabinet base plan. Cheap...

https://www.audiophonics.fr/fr/viss...ylon-male-femelle-m25x20-6mm-x10-p-11983.html

I always use plastic rods on soft double tape in my cabinet for the DAC pcbs : it looks less professional than metal road,,but the damping behavior is waaaaaaay better ! https://www.audiophonics.fr/fr/viss...t-pcb-adhesif-hauteur-11mm-unite-p-12997.html
 
Last edited:
I'd like to building this clock circuit with a 4.2336 MHz crystal. First time i am building one. Do i neede to adjust something?
Thnx. Alsof for the nice work around the Pihilips TDA.

Are calculations compareble to a (for example) choke loaded tube anode?

Say for a Bf862 R = 1/0.045 = 22Ohm in series with 2*pi*f* 0.00015? Phase: Tan (phi) = R /Z?
Can i simplefy by ignoring the diode and 330 ohm resistor?
 
My hypothesis is that its that chip (SAA7220) dirtying the supply that's largely responsible for the bad rap oversampling gets.

This plays a role.

However I would submit a few additional notes.

Keith Howard (IIRC) in HFB&RR & SP investigated pre/post/both ringing and found they were detected differently with certain types of signal, but not with music. If the test itself interfered to a degree was not tested or controlled for.

https://www.stereophile.com/content...biquitous-filter-filter-listening-impressions

Difference's with music were minimal and hard to isolate with music, however there are some interesting notes.

Further, DRA Lab's Douglas Rife (of MLSSA fame) published first a letter to Stereophile and later a little paper on the use of "poor digital filters" to improve DA Converters, arguably in the sense it is used here, no filter is the "poorest" digital filter and if we consider the paper to the logical endpoint....

http://www.mlssa.com/pdf/Upsampling-theory-rev-2.pdf

Pioneer had a long history of using a special digital filter dubbed "legato link" which remains in use in Pioneer's upper range DJ Gear (out of all places!!!) and likely contributes to the strong preference of many DJ's for Pioneer gear, over (say) the Range Serrato Boxes. I actually had occasion to modify Serrato boxes with Extra DAC PCB's that operate a "legato link like" very "poor digital filter" (and arguably were otherwise superior quality) which broadly made the Serrato Box comparable in sound quality to the TOL Pioneer DJ Player. Legato link explained by pioneer:

1730029732862.png


Anecdotally, prior to making my own digital gear, I had a strong preference Pioneer Legato Link equipped CD & DVD Players.

Theta also relied on such filters.

So I think there is a bit more here than just PSU noise.

I could also point to a number of devices of my own design, offering selectable digital filters, including non and listener reactions to the options, but that would be likely redundant.

Thor

PS, I omit to mention that British TLA thing, because A) most of the recordings claiming the benefit from "deblurring" on the recording side were not correctly deblurred and B) the claimed "unfolding" is basically the same as the "poor digital filters" discussed here and the software component running in the DAC get's bypassed by selecting "non oversampling", at that point all it does is turn the pink light on and off.
 
  • Like
Reactions: J_Perkins
That's a good idea. Iancanada made cutting slots inside the PCB to make the crystal pcb floating inside. But anyway the board shoulld be current feeded and damped, so a coax very thin cable with very soft nylon body makes sense.

On the (likely permanently cancelled) "AMR Diablo" DAC, the clock and clock distribution was completely "carved out" from the main PCB and suspended on the SMA cables and special rubber O-Rings. Its somewhat visible in the picture below:

1730031167538.png


The extra mass-loading from the "CNC carved from solid copper and then gold plated" shielding lowered resonance frequency into Hz numbers and it was a well damped resonance IIRC.

Thor
 
  • Like
Reactions: J_Perkins
Are calculations compareble to a (for example) choke loaded tube anode?

Pass. There are texts on this I remember reading, but not the details that I read.

My experience is that circuits we absolutely do not want to oscillate do so easily, readily and without trouble, while circuits designed as oscillators often refuse to oscillate at all costs...

I might just start by scaling the inductors by the frequency and leave the rest.

Thor