VOGONS


What retro activity did you get up to today?

Topic actions

Reply 31220 of 31232, by giantenemycat

User metadata
Rank Member
Rank
Member
giantenemycat wrote on 2026-04-20, 01:14:
MattRocks wrote on 2026-04-19, 06:30:
My understanding is that we don't actually forget anything. Instead, what happens is certain memories become hard to access due […]
Show full quote
giantenemycat wrote on 2026-04-19, 04:04:
I previously thought this was the only photo that "shows" my old family PC, which I've been on a years long quest to identify. […]
Show full quote

I previously thought this was the only photo that "shows" my old family PC, which I've been on a years long quest to identify.

The attachment 4KUHHD.jpg is no longer available

But I had a feeling there was one big piece of evidence left. Went back through the photos today, and only noticed this in an index because I had a flashlight on it. Can't find the actual photo - looked through every single pack of photos for the negative at least, and it WAS in the very last pack with a completely unrelated set of photos.

I will get the negative digitised but not too much hope tbh. I can make out the index photo slightly better in person but it's a fumbled photo with poor lighting. Hopefully when blown up it'll show at least enough detail to make out if it's the model I think it is.

My understanding is that we don't actually forget anything. Instead, what happens is certain memories become hard to access due to bandwidth limitations.

And, we don't index or recall our memories in chronological order. Instead, we recall them by associations (smells, motions, emotions, temperature, heart rate, etc.)

So if you model how you completely felt when you were fixing your PC, and relive that complete set of associations, it's possible you might recall technical specifications you once knew.

But beware - when you recall memories, you also write new memories of recalling memories (recursion) and so un-recalled detail can be buried even further! Given that sensitivity, you might not benefit from a half-arsed practice run.

This is, imho, the entire thesis of therapy in a single post because if you think about it.. a person is the sum of their memories, and the way a person experiences anything is really just an effect of how they remember experiencing something else..

I don't trust my memories tbh. In the lifetime of that PC I was 6-12, and all I knew about PCs was that they had the Pentiums - Pentium this, Pentium that. In fact I remember telling people I had a Pentium II, but a printed Windows resource summary report I found a couple years ago indicates it was actually just a Pentium (MMX). So why was I so sure it was a PII?

I only think I know what PC it was because of that report, thanks to the BIOS revision and date listed. And some elements of the design do ring a bell, while some others kinda don't. So could be a red herring. I just need enough detail from that potato photo after processing the negative to tell if it's this specific AST Bravo LC/MS 51xx/52xx series chassis design. If it's not, then RIP.

Got the dodgy negative done professionally. It was indeed an AST Bravo MS/LC (probably LC) 51xx/52xx from the 1997 430TX series! Now I can just about make out the sticker above the W95 one is the plain blue "Intel Pentium" sticker, not MMX. So that narrows it down LC 5166 or 5200. No way to tell if the CPU was upgraded by my dad's friend before he sold it to him though. Probably not?

It was definitely a choice to have the left speaker underneath, next to the printer. Much like the bright green painted walls were also...interesting. And of course I've got my stack of Sold Out Software games at the bottom.

Anyone think they can figure out what monitor that is?

Reply 31221 of 31232, by MattRocks

User metadata
Rank Member
Rank
Member

Overall it looks like an early 90s CTX, but I couldn't find a CRT that matches the darker patches bottom right - could they be brightness and contrast controls like on the attached AST monitor?

Do you have three colour branding in the top left or is that an unrelated sticker?

Milestones [ MOS 7501 → 68030 → x86(P5/MMX) → x86(K6-2) → x86(K7*) → PPC(G3*) → x86-64(K8) → x86-64(Xeon) → x86-64(i5) → x86-64(i7) ] * original lost

Reply 31222 of 31232, by Shagittarius

User metadata
Rank Oldbie
Rank
Oldbie

The label looks sorta like viewsonic, but it's not a model I'm familiar with , what year is the suspected monitor and the picture...

Reply 31223 of 31232, by giantenemycat

User metadata
Rank Member
Rank
Member
MattRocks wrote on Today, 00:41:

Overall it looks like an early 90s CTX, but I couldn't find a CRT that matches the darker patches bottom right - could they be brightness and contrast controls like on the attached AST monitor?

Do you have three colour branding in the top left or is that an unrelated sticker?

It does look similar to some AST monitors with its big square power button on the ridge and LED to the left of it, and they seem to be "dials" like the ones in your post but smaller. Although AST always put their logo on the bottom left - there's definitely nothing there on this one. The top left I believe is the brand name not a sticker, looks to me like it starts with a stylised M. That does very vaguely ring a bell.

Shagittarius wrote on Today, 01:52:

The label looks sorta like viewsonic, but it's not a model I'm familiar with , what year is the suspected monitor and the picture...

My dad's friend either sold or bought on his behalf both the PC and monitor a fairly tight budget (~£350-ish). Based on my findings that would have been between August '98 - February '99. So the monitor could have been one of his old ones from earlier in the 90s for all I know.

I wonder if someone could determine the size of the screen? The only reference I have is the PC being 17.5" wide.

Reply 31224 of 31232, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
giantenemycat wrote on Yesterday, 23:36:
Got the dodgy negative done professionally. It was indeed an AST Bravo MS/LC (probably LC) 51xx/52xx from the 1997 430TX series! […]
Show full quote
giantenemycat wrote on 2026-04-20, 01:14:
MattRocks wrote on 2026-04-19, 06:30:
My understanding is that we don't actually forget anything. Instead, what happens is certain memories become hard to access due […]
Show full quote

My understanding is that we don't actually forget anything. Instead, what happens is certain memories become hard to access due to bandwidth limitations.

And, we don't index or recall our memories in chronological order. Instead, we recall them by associations (smells, motions, emotions, temperature, heart rate, etc.)

So if you model how you completely felt when you were fixing your PC, and relive that complete set of associations, it's possible you might recall technical specifications you once knew.

But beware - when you recall memories, you also write new memories of recalling memories (recursion) and so un-recalled detail can be buried even further! Given that sensitivity, you might not benefit from a half-arsed practice run.

This is, imho, the entire thesis of therapy in a single post because if you think about it.. a person is the sum of their memories, and the way a person experiences anything is really just an effect of how they remember experiencing something else..

I don't trust my memories tbh. In the lifetime of that PC I was 6-12, and all I knew about PCs was that they had the Pentiums - Pentium this, Pentium that. In fact I remember telling people I had a Pentium II, but a printed Windows resource summary report I found a couple years ago indicates it was actually just a Pentium (MMX). So why was I so sure it was a PII?

I only think I know what PC it was because of that report, thanks to the BIOS revision and date listed. And some elements of the design do ring a bell, while some others kinda don't. So could be a red herring. I just need enough detail from that potato photo after processing the negative to tell if it's this specific AST Bravo LC/MS 51xx/52xx series chassis design. If it's not, then RIP.

Got the dodgy negative done professionally. It was indeed an AST Bravo MS/LC (probably LC) 51xx/52xx from the 1997 430TX series! Now I can just about make out the sticker above the W95 one is the plain blue "Intel Pentium" sticker, not MMX. So that narrows it down LC 5166 or 5200. No way to tell if the CPU was upgraded by my dad's friend before he sold it to him though. Probably not?

It was definitely a choice to have the left speaker underneath, next to the printer. Much like the bright green painted walls were also...interesting. And of course I've got my stack of Sold Out Software games at the bottom.

Anyone think they can figure out what monitor that is?

Nice job having that negative developed! That is awesome to get so much more information out of such an old photo.

The first thing that came to mind when I saw the logo on the monitor was HP's old logo:
https://stellularpictures.com/product/hewlett … -19-beige-1998/
But I can't find any monitors that look like the one in your picture.

This one is closer but still not quite right.
https://www.ebay.com/itm/397780728390

It does kind of look like the Viewsonic birds (with something else to the right of them), but I don't believe they put that badge on any monitors this old, and I think the oldest version of the badge was larger and stood out a lot more than this one.

With the exposed analog dials I would say that this CRT is pre-1995... possible very early 90s.

Now for some blitting from the back buffer.

Reply 31225 of 31232, by zwrr

User metadata
Rank Member
Rank
Member

I made some modifications to my Creative Sound Blaster PRO 2 sound card.

Removed all the Wincap capacitors from the Sound Blaster PRO 2 and replaced them with Nichicon FW and FG series capacitors. I also connected a WIMA MKS2 100nF capacitor in parallel with a 470uF capacitor near the output jack.

Removed the Samsung KA3403, STMicroelectronics HCF4053, and STMicroelectronics HCF4066 chips and replaced them with Texas Instruments MC3403, NXP HEF4053, and NXP HEF4066 chips.

After the modifications, I felt the noise level was reduced, making it much quieter, and the bass and treble were also enhanced, but I'm not sure if this is an actual improvement or just my placebo effect.

The attachment SBP-4.jpg is no longer available

SBC1: Cyrix 5x86-120, HS-5x86HVGA, 16MB EDO, GD54M30, SB Pro II, HardMPU-wt
SBC2: VIA C3-800, PCISA-C800, 128MB SDRAM, Savage4 Pro, SB AWE64 Gold
SBC3: Tualatin-S 1.4G, PCI-6872, 256MB SDRAM, FX5200 PCI, Voodoo2 SLI, SB Live

Reply 31226 of 31232, by tehsiggi

User metadata
Rank Oldbie
Rank
Oldbie
zwrr wrote on Today, 06:18:
I made some modifications to my Creative Sound Blaster PRO 2 sound card. […]
Show full quote

I made some modifications to my Creative Sound Blaster PRO 2 sound card.

Removed all the Wincap capacitors from the Sound Blaster PRO 2 and replaced them with Nichicon FW and FG series capacitors. I also connected a WIMA MKS2 100nF capacitor in parallel with a 470uF capacitor near the output jack.

Removed the Samsung KA3403, STMicroelectronics HCF4053, and STMicroelectronics HCF4066 chips and replaced them with Texas Instruments MC3403, NXP HEF4053, and NXP HEF4066 chips.

After the modifications, I felt the noise level was reduced, making it much quieter, and the bass and treble were also enhanced, but I'm not sure if this is an actual improvement or just my placebo effect.

The attachment SBP-4.jpg is no longer available

What do you hope to get from having the 100nF parallel to the 470µF?

AGP Card Real Power Consumption
AGP Power monitor - diagnostic hardware tool
Graphics card repair collection

Reply 31227 of 31232, by zwrr

User metadata
Rank Member
Rank
Member
tehsiggi wrote on Today, 07:17:
zwrr wrote on Today, 06:18:
I made some modifications to my Creative Sound Blaster PRO 2 sound card. […]
Show full quote

I made some modifications to my Creative Sound Blaster PRO 2 sound card.

Removed all the Wincap capacitors from the Sound Blaster PRO 2 and replaced them with Nichicon FW and FG series capacitors. I also connected a WIMA MKS2 100nF capacitor in parallel with a 470uF capacitor near the output jack.

Removed the Samsung KA3403, STMicroelectronics HCF4053, and STMicroelectronics HCF4066 chips and replaced them with Texas Instruments MC3403, NXP HEF4053, and NXP HEF4066 chips.

After the modifications, I felt the noise level was reduced, making it much quieter, and the bass and treble were also enhanced, but I'm not sure if this is an actual improvement or just my placebo effect.

The attachment SBP-4.jpg is no longer available

What do you hope to get from having the 100nF parallel to the 470µF?

The AI ​​told me this. 😁

Connecting a 100nF (0.1μF) capacitor in parallel with a 470μF continuous capacitor at the sound card output (usually a coupling or analog circuit) is a classic circuit design technique. Its main functions are high-frequency decoupling (analog), improving high-frequency characteristics, and boosting high frequencies.

Specific functions are as follows:

Bypassing high-frequency noise and improving frequency characteristics (suppression): Large capacitor (470μF): Primarily used to filter low-frequency noise and provide energy reserves. Its large capacitance results in high parasitic inductance and impedance gain for high-frequency signals.

Small capacitor (100nF ceramic/film capacitor): Possesses excellent high-frequency characteristics, effectively "bypassing" high frequencies (such as RF interference and switching power supply noise) to ground, resulting in a more thorough signal reception.

Parallel effect: The small capacitor compensates for the large capacitor's insufficient high-sensitivity performance, resulting in lower impedance across a wide frequency range for better noise filtering.

Improving sound quality (decoupling): In high-fidelity audio output, high-frequency noise can cause the sound to be "harsh" or the background to be not quiet enough. A 100nF capacitor can eliminate noise caused by the failure of large capacitors at high frequencies, resulting in a cleaner sound background and more delicate treble.

Reducing Parasitic Effects: The 470μF capacitor's resonant series inductance (ESL) increases, and its capacitive characteristics deteriorate when operating under high-speed signals. Connecting a 100nF capacitor in parallel, with its excellent high-frequency characteristics, effectively reduces the total impedance at high frequencies, preventing self-oscillation or signal distortion.

Summary: This design utilizes a large capacitor to handle low-frequency energy and a small capacitor to handle high-frequency noise. The 470μF capacitor provides "coarse filtering," while the 100nF capacitor provides "fine filtering," working together to ensure a periodic output sound signal unaffected by high-frequency noise.

SBC1: Cyrix 5x86-120, HS-5x86HVGA, 16MB EDO, GD54M30, SB Pro II, HardMPU-wt
SBC2: VIA C3-800, PCISA-C800, 128MB SDRAM, Savage4 Pro, SB AWE64 Gold
SBC3: Tualatin-S 1.4G, PCI-6872, 256MB SDRAM, FX5200 PCI, Voodoo2 SLI, SB Live

Reply 31228 of 31232, by tehsiggi

User metadata
Rank Oldbie
Rank
Oldbie
zwrr wrote on Today, 07:29:

Bypassing high-frequency noise and improving frequency characteristics (suppression): Large capacitor (470μF): Primarily used to filter low-frequency noise and provide energy reserves. Its large capacitance results in high parasitic inductance and impedance gain for high-frequency signals.

This is for buffering supply voltages and filtering them, not really for audio signal applications.

zwrr wrote on Today, 07:29:

Parallel effect: The small capacitor compensates for the large capacitor's insufficient high-sensitivity performance, resulting in lower impedance across a wide frequency range for better noise filtering.

Same, not really applicable on line level signals.

zwrr wrote on Today, 07:29:

Improving sound quality (decoupling): In high-fidelity audio output, high-frequency noise can cause the sound to be "harsh" or the background to be not quiet enough. A 100nF capacitor can eliminate noise caused by the failure of large capacitors at high frequencies, resulting in a cleaner sound background and more delicate treble.

This still talks about noise coming from the power supply of the amplifier components, not the signal itself.

zwrr wrote on Today, 07:29:

Reducing Parasitic Effects: The 470μF capacitor's resonant series inductance (ESL) increases, and its capacitive characteristics deteriorate when operating under high-speed signals. Connecting a 100nF capacitor in parallel, with its excellent high-frequency characteristics, effectively reduces the total impedance at high frequencies, preventing self-oscillation or signal distortion.

Summary: This design utilizes a large capacitor to handle low-frequency energy and a small capacitor to handle high-frequency noise. The 470μF capacitor provides "coarse filtering," while the 100nF capacitor provides "fine filtering," working together to ensure a periodic output sound signal unaffected by high-frequency noise.

Again, very vague and not related to your use-case.

The sound-quality of course depends on the capacitors within your signal path, however you already have a couple of electrolytics in there, so there is really not much gain from the 100nF parallel to the 470µF on the output. The 470µF are even only that large because your card does not have a real line-out, but rather a speaker out. The 470µF being important as the speakers would have a low resistance / impedance, thus forming a high pass with the capacitor. If the capacitance is too low, the lower frequency cutoff will be significant.
For low output loads with high resistance, such as line-inputs from active speakers, the 470µF are pretty much oversized in theory. A good 1-10µF would be plenty. I know you don't wanna re-design the card, so it's not a big deal. But you can save yourself the hassle to parallel the 100nF on the output, you'll not get anything noticeable from that.

The worst part in your signal path will probably be the speaker amp on the card anyways, as the TEA2025 is no "low-noise" or "low-distortion" part at all.

The attachment Screenshot 2026-04-24 at 10.16.45.png is no longer available

It already adds 0.2% THD+N distortion just by being in your signal path 😁

So from my years in audio development / design: Recap if you like to. If the footprint only allows you for electrolytics, replace it with proper ones and call it a day. Wherever possible, switch to film capacitors. But this is only important in the signal path. The rest can just stay electrolytics.

If 100% looks nice with the new caps 😀

AGP Card Real Power Consumption
AGP Power monitor - diagnostic hardware tool
Graphics card repair collection

Reply 31229 of 31232, by zwrr

User metadata
Rank Member
Rank
Member
tehsiggi wrote on Today, 08:27:
This is for buffering supply voltages and filtering them, not really for audio signal applications. […]
Show full quote
zwrr wrote on Today, 07:29:

Bypassing high-frequency noise and improving frequency characteristics (suppression): Large capacitor (470μF): Primarily used to filter low-frequency noise and provide energy reserves. Its large capacitance results in high parasitic inductance and impedance gain for high-frequency signals.

This is for buffering supply voltages and filtering them, not really for audio signal applications.

zwrr wrote on Today, 07:29:

Parallel effect: The small capacitor compensates for the large capacitor's insufficient high-sensitivity performance, resulting in lower impedance across a wide frequency range for better noise filtering.

Same, not really applicable on line level signals.

zwrr wrote on Today, 07:29:

Improving sound quality (decoupling): In high-fidelity audio output, high-frequency noise can cause the sound to be "harsh" or the background to be not quiet enough. A 100nF capacitor can eliminate noise caused by the failure of large capacitors at high frequencies, resulting in a cleaner sound background and more delicate treble.

This still talks about noise coming from the power supply of the amplifier components, not the signal itself.

zwrr wrote on Today, 07:29:

Reducing Parasitic Effects: The 470μF capacitor's resonant series inductance (ESL) increases, and its capacitive characteristics deteriorate when operating under high-speed signals. Connecting a 100nF capacitor in parallel, with its excellent high-frequency characteristics, effectively reduces the total impedance at high frequencies, preventing self-oscillation or signal distortion.

Summary: This design utilizes a large capacitor to handle low-frequency energy and a small capacitor to handle high-frequency noise. The 470μF capacitor provides "coarse filtering," while the 100nF capacitor provides "fine filtering," working together to ensure a periodic output sound signal unaffected by high-frequency noise.

Again, very vague and not related to your use-case.

The sound-quality of course depends on the capacitors within your signal path, however you already have a couple of electrolytics in there, so there is really not much gain from the 100nF parallel to the 470µF on the output. The 470µF are even only that large because your card does not have a real line-out, but rather a speaker out. The 470µF being important as the speakers would have a low resistance / impedance, thus forming a high pass with the capacitor. If the capacitance is too low, the lower frequency cutoff will be significant.
For low output loads with high resistance, such as line-inputs from active speakers, the 470µF are pretty much oversized in theory. A good 1-10µF would be plenty. I know you don't wanna re-design the card, so it's not a big deal. But you can save yourself the hassle to parallel the 100nF on the output, you'll not get anything noticeable from that.

The worst part in your signal path will probably be the speaker amp on the card anyways, as the TEA2025 is no "low-noise" or "low-distortion" part at all.

The attachment Screenshot 2026-04-24 at 10.16.45.png is no longer available

It already adds 0.2% THD+N distortion just by being in your signal path 😁

So from my years in audio development / design: Recap if you like to. If the footprint only allows you for electrolytics, replace it with proper ones and call it a day. Wherever possible, switch to film capacitors. But this is only important in the signal path. The rest can just stay electrolytics.

If 100% looks nice with the new caps 😀

Yes, I know that TEA2025 is a weak point, but after checking some information, it seems that there is no component that can be used to directly replace it. There are some good suggestions, but they require modifications to the surrounding circuitry of TEA2025, which is somewhat beyond my capabilities.

SBC1: Cyrix 5x86-120, HS-5x86HVGA, 16MB EDO, GD54M30, SB Pro II, HardMPU-wt
SBC2: VIA C3-800, PCISA-C800, 128MB SDRAM, Savage4 Pro, SB AWE64 Gold
SBC3: Tualatin-S 1.4G, PCI-6872, 256MB SDRAM, FX5200 PCI, Voodoo2 SLI, SB Live

Reply 31230 of 31232, by tehsiggi

User metadata
Rank Oldbie
Rank
Oldbie
zwrr wrote on Today, 08:39:
tehsiggi wrote on Today, 08:27:
This is for buffering supply voltages and filtering them, not really for audio signal applications. […]
Show full quote
zwrr wrote on Today, 07:29:

Bypassing high-frequency noise and improving frequency characteristics (suppression): Large capacitor (470μF): Primarily used to filter low-frequency noise and provide energy reserves. Its large capacitance results in high parasitic inductance and impedance gain for high-frequency signals.

This is for buffering supply voltages and filtering them, not really for audio signal applications.

zwrr wrote on Today, 07:29:

Parallel effect: The small capacitor compensates for the large capacitor's insufficient high-sensitivity performance, resulting in lower impedance across a wide frequency range for better noise filtering.

Same, not really applicable on line level signals.

zwrr wrote on Today, 07:29:

Improving sound quality (decoupling): In high-fidelity audio output, high-frequency noise can cause the sound to be "harsh" or the background to be not quiet enough. A 100nF capacitor can eliminate noise caused by the failure of large capacitors at high frequencies, resulting in a cleaner sound background and more delicate treble.

This still talks about noise coming from the power supply of the amplifier components, not the signal itself.

zwrr wrote on Today, 07:29:

Reducing Parasitic Effects: The 470μF capacitor's resonant series inductance (ESL) increases, and its capacitive characteristics deteriorate when operating under high-speed signals. Connecting a 100nF capacitor in parallel, with its excellent high-frequency characteristics, effectively reduces the total impedance at high frequencies, preventing self-oscillation or signal distortion.

Summary: This design utilizes a large capacitor to handle low-frequency energy and a small capacitor to handle high-frequency noise. The 470μF capacitor provides "coarse filtering," while the 100nF capacitor provides "fine filtering," working together to ensure a periodic output sound signal unaffected by high-frequency noise.

Again, very vague and not related to your use-case.

The sound-quality of course depends on the capacitors within your signal path, however you already have a couple of electrolytics in there, so there is really not much gain from the 100nF parallel to the 470µF on the output. The 470µF are even only that large because your card does not have a real line-out, but rather a speaker out. The 470µF being important as the speakers would have a low resistance / impedance, thus forming a high pass with the capacitor. If the capacitance is too low, the lower frequency cutoff will be significant.
For low output loads with high resistance, such as line-inputs from active speakers, the 470µF are pretty much oversized in theory. A good 1-10µF would be plenty. I know you don't wanna re-design the card, so it's not a big deal. But you can save yourself the hassle to parallel the 100nF on the output, you'll not get anything noticeable from that.

The worst part in your signal path will probably be the speaker amp on the card anyways, as the TEA2025 is no "low-noise" or "low-distortion" part at all.

The attachment Screenshot 2026-04-24 at 10.16.45.png is no longer available

It already adds 0.2% THD+N distortion just by being in your signal path 😁

So from my years in audio development / design: Recap if you like to. If the footprint only allows you for electrolytics, replace it with proper ones and call it a day. Wherever possible, switch to film capacitors. But this is only important in the signal path. The rest can just stay electrolytics.

If 100% looks nice with the new caps 😀

Yes, I know that TEA2025 is a weak point, but after checking some information, it seems that there is no component that can be used to directly replace it. There are some good suggestions, but they require modifications to the surrounding circuitry of TEA2025, which is somewhat beyond my capabilities.

You could, in theory, just remove it and put in a couple of bridge wires for it instead. You lose the speaker driving capability, but I guess you'll be using it with active speakers anyway.

But seeing what they are being traded for on ebay, just leave it as is.

AGP Card Real Power Consumption
AGP Power monitor - diagnostic hardware tool
Graphics card repair collection

Reply 31231 of 31232, by zwrr

User metadata
Rank Member
Rank
Member
tehsiggi wrote on Today, 08:43:
zwrr wrote on Today, 08:39:
tehsiggi wrote on Today, 08:27:
This is for buffering supply voltages and filtering them, not really for audio signal applications. […]
Show full quote

This is for buffering supply voltages and filtering them, not really for audio signal applications.

Same, not really applicable on line level signals.

This still talks about noise coming from the power supply of the amplifier components, not the signal itself.

Again, very vague and not related to your use-case.

The sound-quality of course depends on the capacitors within your signal path, however you already have a couple of electrolytics in there, so there is really not much gain from the 100nF parallel to the 470µF on the output. The 470µF are even only that large because your card does not have a real line-out, but rather a speaker out. The 470µF being important as the speakers would have a low resistance / impedance, thus forming a high pass with the capacitor. If the capacitance is too low, the lower frequency cutoff will be significant.
For low output loads with high resistance, such as line-inputs from active speakers, the 470µF are pretty much oversized in theory. A good 1-10µF would be plenty. I know you don't wanna re-design the card, so it's not a big deal. But you can save yourself the hassle to parallel the 100nF on the output, you'll not get anything noticeable from that.

The worst part in your signal path will probably be the speaker amp on the card anyways, as the TEA2025 is no "low-noise" or "low-distortion" part at all.

The attachment Screenshot 2026-04-24 at 10.16.45.png is no longer available

It already adds 0.2% THD+N distortion just by being in your signal path 😁

So from my years in audio development / design: Recap if you like to. If the footprint only allows you for electrolytics, replace it with proper ones and call it a day. Wherever possible, switch to film capacitors. But this is only important in the signal path. The rest can just stay electrolytics.

If 100% looks nice with the new caps 😀

Yes, I know that TEA2025 is a weak point, but after checking some information, it seems that there is no component that can be used to directly replace it. There are some good suggestions, but they require modifications to the surrounding circuitry of TEA2025, which is somewhat beyond my capabilities.

You could, in theory, just remove it and put in a couple of bridge wires for it instead. You lose the speaker driving capability, but I guess you'll be using it with active speakers anyway.

But seeing what they are being traded for on ebay, just leave it as is.

You're absolutely right. I searched for it for a long time and just bought it a few days ago. It was surprisingly cheap, only $60! I will keep it and use it.

SBC1: Cyrix 5x86-120, HS-5x86HVGA, 16MB EDO, GD54M30, SB Pro II, HardMPU-wt
SBC2: VIA C3-800, PCISA-C800, 128MB SDRAM, Savage4 Pro, SB AWE64 Gold
SBC3: Tualatin-S 1.4G, PCI-6872, 256MB SDRAM, FX5200 PCI, Voodoo2 SLI, SB Live

Reply 31232 of 31232, by BitWrangler

User metadata
Rank l33t++
Rank
l33t++
Ozzuneoj wrote on Today, 05:27:
Nice job having that negative developed! That is awesome to get so much more information out of such an old photo. […]
Show full quote
giantenemycat wrote on Yesterday, 23:36:
Got the dodgy negative done professionally. It was indeed an AST Bravo MS/LC (probably LC) 51xx/52xx from the 1997 430TX series! […]
Show full quote
giantenemycat wrote on 2026-04-20, 01:14:

I don't trust my memories tbh. In the lifetime of that PC I was 6-12, and all I knew about PCs was that they had the Pentiums - Pentium this, Pentium that. In fact I remember telling people I had a Pentium II, but a printed Windows resource summary report I found a couple years ago indicates it was actually just a Pentium (MMX). So why was I so sure it was a PII?

I only think I know what PC it was because of that report, thanks to the BIOS revision and date listed. And some elements of the design do ring a bell, while some others kinda don't. So could be a red herring. I just need enough detail from that potato photo after processing the negative to tell if it's this specific AST Bravo LC/MS 51xx/52xx series chassis design. If it's not, then RIP.

Got the dodgy negative done professionally. It was indeed an AST Bravo MS/LC (probably LC) 51xx/52xx from the 1997 430TX series! Now I can just about make out the sticker above the W95 one is the plain blue "Intel Pentium" sticker, not MMX. So that narrows it down LC 5166 or 5200. No way to tell if the CPU was upgraded by my dad's friend before he sold it to him though. Probably not?

It was definitely a choice to have the left speaker underneath, next to the printer. Much like the bright green painted walls were also...interesting. And of course I've got my stack of Sold Out Software games at the bottom.

Anyone think they can figure out what monitor that is?

Nice job having that negative developed! That is awesome to get so much more information out of such an old photo.

The first thing that came to mind when I saw the logo on the monitor was HP's old logo:
https://stellularpictures.com/product/hewlett … -19-beige-1998/
But I can't find any monitors that look like the one in your picture.

This one is closer but still not quite right.
https://www.ebay.com/itm/397780728390

It does kind of look like the Viewsonic birds (with something else to the right of them), but I don't believe they put that badge on any monitors this old, and I think the oldest version of the badge was larger and stood out a lot more than this one.

With the exposed analog dials I would say that this CRT is pre-1995... possible very early 90s.

Using my ultra high tech "Squint at it through a pin-hole" technique, it actually looks like the two darker patches on the lower right are not knob shadows but LEDs with a horizontal bar diffuser. Then it looks like button controls, with maybe or maybe not knobs, sitting right under the edge of the bezel recess on the left.

Also from some angles, it seems that the system unit badge and monitor badge could match, and the different angle to the lighting records them differently. However, at other levels of squint they look sufficiently different.

Though for monitors, who made it and who badged it can be two different things. The OEM might not put their badge in the same location as the rebrand either.

Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.