That card uses National Semi's fliter chips and those are very noisy. I bypassed the one on my BXII card and it became nearly silent as far as that hiss goes.

Tiido wrote on 2023-11-21, 12:44:

That card uses National Semi's fliter chips and those are very noisy. I bypassed the one on my BXII card and it became nearly silent as far as that hiss goes.

Sorry, I know this is an old post, but can you explain how you did this and if there are any downsides to doing so? Thank you!

You'll have to look at the datasheets to identify the input and output pins, then separate them and run a wire across in the most simplest way. Ideally you skip one of the DC blocking capacitors too, doesn't matter if the input or the output one, as long as you don't have both of them since that just will end up reducing low freq response.

As far as potential drawbacks go, low sample rates will have more aliasing noise which some think is a bad thing, some others such as myself think it is fine and makes stuff sound crisper and not so muffled even if there's a metallic or robotic quality to the sound. Sort of like the debate about filters in SB clones vs none (and I'm in the none camp 🤣).

The card is in my Toshiba T3200, sort of integrated into it so that machine's PC speaker sounds come through the audio output of the card (and also go into the internal speaker of the luggable) so I cannot easily extract it to at least show photos of what I did. It has been a while since I worked on it in any capacity...

Tiido wrote on 2024-12-22, 04:42:

You'll have to look at the datasheets to identify the input and output pins, then separate them and run a wire across in the most simplest way. Ideally you skip one of the DC blocking capacitors too, doesn't matter if the input or the output one, as long as you don't have both of them since that just will end up reducing low freq response.

As far as potential drawbacks go, low sample rates will have more aliasing noise which some think is a bad thing, some others such as myself think it is fine and makes stuff sound crisper and not so muffled even if there's a metallic or robotic quality to the sound. Sort of like the debate about filters in SB clones vs none (and I'm in the none camp 🤣).

The card is in my Toshiba T3200, sort of integrated into it so that machine's PC speaker sounds come through the audio output of the card (and also go into the internal speaker of the luggable) so I cannot easily extract it to at least show photos of what I did. It has been a while since I worked on it in any capacity...

Thanks for the input!

Just to get me started on the right path, which chip would probably be the one to bypass on a card like this?

I will look up the datasheet of course, but I don't know enough about audio circuits to be able to just look at it and know which chip is the offending filter.

(Sorry, this is a picture I took years ago... I just realized the text on the chips isn't very clear. The picture here is much much clearer.)

The filter chip is the DIP8 with vertical stripe and "MF 4CN-100" written on it. https://www.digchip.com/datasheets/parts/data … eet/321/MF4.php

Tiido wrote on 2024-12-22, 05:53:

The filter chip is the DIP8 with vertical stripe and "MF 4CN-100" written on it. https://www.digchip.com/datasheets/parts/data … eet/321/MF4.php

Great, thank you!

I would obviously check all of these traces up-close first in case there is anything that isn't visible in this picture, but is this the basic idea, with the cyan line indicating the wire I would use to bypass the filter and the input DC blocking cap?

The output of the chip has to be separated from everything since it is where all the noise comes from, even if it is receiving no input. The input should also be separated since it likely gets influenced due to way it operates.

The DC blocking cap is identified correctly, you'll want to remove it, and connect the hole that goes to input on the chip to ground so that there is nothing for the chip to work with so to speak. The other hole you leave the capacitor in with one leg in air and that one you'll use to connect wire to, which will go to wherever the output DC blocking capacitor is, but the pin that is toward rest of the circuitry and not one connecting to the chip output. I'm not totally sure which of the capacitors is the output one and what other parts there may be that should get removed in the process.

EDIT: Hold up, I got the card out my T3200 and I'll have photos etc. to show what I did to it soon ~

EDIT2 :
* I moved R42 to where C35 is, this separates the input and grounds it so the chip has nothing to work with
* I removed R38 and C29, which separates the filter output.
* I made a bridge with a resistor leg clipping to connect audio out from Aztech chip to its input, bypassing the filter chip in the process.

My two Sound Galaxy cards

Sound Galaxy BX (sadly don't have the little passive speakers that came with the card)

Sound Galaxy NX II / CPS Soundblaster 2.5 / HighScreen Blaster 3.0

Bonus - Sound Galaxy Wave Power / Escom Wave Power (gave that one away though along with the SB16):