This is my wet dream 🤣

appiah4 wrote on 2023-06-14, 09:28:

This is my wet dream 🤣

Wait, you didn't even see the real hardware pictures yet 😁

Finally the prototypes.. and ridiculously huge stencil

Large stencil is good, I pretty much always order full size for the important things... the ones cut to small size are often bent out of shape so much that they're hard to get good pasting with 😒
Welcome to the world of meat based pick & placing 🤣 (beats cutting and bending pins and hand soldering bazillion through hole parts)

Tiido wrote on 2023-06-30, 10:58:

Large stencil is good, I pretty much always order full size for the important things... the ones cut to small size are often bent out of shape so much that they're hard to get good pasting with 😒
Welcome to the world of meat based pick & placing 🤣 (beats cutting and bending pins and hand soldering bazillion through hole parts)

Needs some practice though to grip it well for a good contact in the center 😀

Not perfect, but at least there are lots of unused pcbs left from another projects to make it work

very nice!
looking forward to the finished card!

keropi wrote on 2023-06-30, 13:47:

very nice!
looking forward to the finished card!

Hey, mate!

LABS wrote on 2023-06-30, 10:31:

SB8-ppcb.jpg

Finally the prototypes.. and ridiculously huge stencil

That is one big ass stencil 🤣

LABS wrote on 2023-06-30, 12:49:

Needs some practice though to grip it well for a good contact in the center 😀

Making a perimeter using same thickness PCBs is exactly what I am doing, I tape them down to the desk so they cannot move and then I tape the stencil on them once it is perfectly aligned and then I do the swipe, with another PCB that is just as wide as one side of the PCB being pasted. It helps when the entire swipe is done in one go with good pressure, then the pads end up getting pasted nicely and there's no spillover which normally happens when I have to do anothes wipe in some other part as the stencil often lifhs slighly. It isn't gonna be a problem until you get ot 0.5mm pitch parts though, in which case even small spills can make unwanted bridges but with 1.27mm this is not gonna be an issue ~

Quite marvelous I must admit 😁

Now to the hard part - testing and debugging 🤓

Tiido wrote on 2023-06-30, 17:02:

LABS wrote on 2023-06-30, 12:49:

Needs some practice though to grip it well for a good contact in the center 😀

Making a perimeter using same thickness PCBs is exactly what I am doing, I tape them down to the desk so they cannot move and then I tape the stencil on them once it is perfectly aligned and then I do the swipe, with another PCB that is just as wide as one side of the PCB being pasted. It helps when the entire swipe is done in one go with good pressure, then the pads end up getting pasted nicely and there's no spillover which normally happens when I have to do anothes wipe in some other part as the stencil often lifhs slighly. It isn't gonna be a problem until you get ot 0.5mm pitch parts though, in which case even small spills can make unwanted bridges but with 1.27mm this is not gonna be an issue ~

I used a plastic card instead and applied the paste with rapid short swipes and the stencil was moving up and down a bit and as a result there were some unpleasant bridges and poorly filled pads which I had to correct by a needle 😕
Your tips are very useful, will try next time, thanks!

Wow! This is really impressive looking. Very tidy, very simplified. Having a discrete mixer is just perfect- we can then make adjustments on the fly without needing to route the individual signals to a mixer or bouncing between app and setup program.

LABS wrote on 2023-06-30, 18:17:

Quite marvelous I must admit 😁

IMG_2917.jpg

Now to the hard part - testing and debugging 🤓

Such splendour..

LABS wrote on 2023-06-13, 21:30:

Hello,
let me introduce my humble attempt to design an even more compatible and simple sound card for DOS 😀

Wow, for the first time I see a correctly cut plane between analog and digital.
Can I whine like an old man?
- Why no via stitching?
https://resources.altium.com/ru/p/everything- … -stitching-vias
- I would recommend using combo pads.
You can solder anything on them, starting from MLСС
- I would ground the mounting bracket a little differently.
- L1 and 2 are best replaced with 1-2 ohm resistors.
Filtering is the same, but there are no uncontrolled resonances.
- V1 and 2 are better to use LM317 / 338 with a capacitor on the adj pin of 10+ uF.
With a voltage drop of 5+ volts across them, they are much better than lowdrop regulators at noise and ripple rejection.

PS. Very nice design.

Thanks for your input, shevalier! I'm actually quite happy with current design with -95dB RMS noise floor which is more than enough for sound sources of such fidelity to consider the audio output as "dead quiet". However at the early stage of designing Blasterboard I already tried almost everything you mentioned. The first prototype even had 2 stages of voltage regulation to extremely reduce the PSU noise 😎

But the real tests showed that there is no improvement compared with a single pair of 79/78L0x regs when used in such noisy environment. So I dropped that idea in favor of simpler design. What really introduced a horrible noise - was grounding the bracket because of the ground loop. Anyway the construction does not allow to ground the bracket now, as it is mounted using nylon mounts.
L1 and L2 should go away and I'll put wires here, again, in favor of simplicity - the filtering frequency is far far away from the audio spectrum and even from the TL074's bandwidth anyway, so it does not affect the noise and there are 22p caps on opamps for stability.
Interesting article on via stitches by the way 😀
Cheers, mate! 🤙

Yeah, vias in the right place can make a dramatic difference

Tiido wrote on 2023-07-01, 16:54:

Yeah, vias in the right place can make a dramatic difference

That's exactly why they are not all over the place 😀

Engineering 😌