VOGONS


First post, by tayyare

User metadata
Rank Oldbie
Rank
Oldbie

I have a 486 rig that I'm working on for some time. I had this odd problem: Turbo LED is very very dim. I changed the leds, etc, but nothing worked. Then just by chance, I see this:

WhatsApp Image 2019-03-04 at 20.41.23.jpeg
Filename
WhatsApp Image 2019-03-04 at 20.41.23.jpeg
File size
157.04 KiB
Views
817 views
File license
Fair use/fair dealing exception

A resistor soldered underside the board on to the LED pins. (sorry about the bad photo)

Why on earth it was needed? May I remove it?

GA-6VTXE PIII 1.4+512MB
Geforce4 Ti 4200 64MB
Diamond Monster 3D 12MB SLI
SB AWE64 PNP+32MB
120GB IDE Samsung/80GB IDE Seagate/146GB SCSI Compaq/73GB SCSI IBM
Adaptec AHA29160
3com 3C905B-TX
Gotek+CF Reader
MSDOS 6.22+Win 3.11/95 OSR2.1/98SE/ME/2000

Reply 1 of 11, by retardware

User metadata
Rank Oldbie
Rank
Oldbie

One of the previous owners must have wanted a dimmer LED 😀
yes you can snip or solder off that, as mobos have already a proper LED series resistor if the connector post is labeled as LED post.

Reply 2 of 11, by Merovign

User metadata
Rank Member
Rank
Member

Some types of older LEDs tended to dim quite a bit with age. I'd test a replacement LED before snipping the resistor.

It might be a consumer bodge but it might also be a factory bodge, based on the apparent age.

*Too* *many* *things*!

Reply 3 of 11, by tayyare

User metadata
Rank Oldbie
Rank
Oldbie
Merovign wrote:

Some types of older LEDs tended to dim quite a bit with age. I'd test a replacement LED before snipping the resistor.

It might be a consumer bodge but it might also be a factory bodge, based on the apparent age.

Before becaming aware of this modification, I tried 3 different leds, all the same. I also exchanged it with HDD LED which was working properly when connected to HDD pins, but gets dim as soon as I connect it to turbo LED pins.

Removing it will be the logical thing to do I guess.

Thanks for all the replies 😊

GA-6VTXE PIII 1.4+512MB
Geforce4 Ti 4200 64MB
Diamond Monster 3D 12MB SLI
SB AWE64 PNP+32MB
120GB IDE Samsung/80GB IDE Seagate/146GB SCSI Compaq/73GB SCSI IBM
Adaptec AHA29160
3com 3C905B-TX
Gotek+CF Reader
MSDOS 6.22+Win 3.11/95 OSR2.1/98SE/ME/2000

Reply 4 of 11, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

I'm skeptical. You need a complete circuit, which usually looks something like +5V -> Resistor -> LED -> Gnd. (Or the resistor could be on the Gnd side of the LED -- doesn't matter which.)

If that was an after-market bodge, then you might have +5V -> Resistor -> LED -> Resistor -> Gnd, in which case it would be safe to remove that extra resistor. Except... what got removed to redirect either the high or low side of the LED through a resistor? Someone would have had to sever the Gnd or +5V connection and put that resistor in between, so just snipping the resistor is likely to leave you with an open circuit.

If you bridge those pins without a resistor then fine .... unless they removed the factory resistor and put a higher-value one in its place. If THAT'S the case, you need to remove that one and put a lower value in. Otherwise there's no current limit to the LED and it'll burn itself out (or damage whatever is controlling it.)

Reply 5 of 11, by Super_Relay

User metadata
Rank Newbie
Rank
Newbie
SirNickity wrote:

I'm skeptical. You need a complete circuit, which usually looks something like +5V -> Resistor -> LED -> Gnd. (Or the resistor could be on the Gnd side of the LED -- doesn't matter which.)

If that was an after-market bodge, then you might have +5V -> Resistor -> LED -> Resistor -> Gnd, in which case it would be safe to remove that extra resistor. Except... what got removed to redirect either the high or low side of the LED through a resistor? Someone would have had to sever the Gnd or +5V connection and put that resistor in between, so just snipping the resistor is likely to leave you with an open circuit.

If you bridge those pins without a resistor then fine .... unless they removed the factory resistor and put a higher-value one in its place. If THAT'S the case, you need to remove that one and put a lower value in. Otherwise there's no current limit to the LED and it'll burn itself out (or damage whatever is controlling it.)

If the resistor is soldered across the connector so it is in parallel with the LED. removing it wont break the circuit but shorting it out would draw a significant amount of current through the built in resistor. A lower value would actually make the LED dimmer because the current will have an even easier path to flow around the LED.

Reply 6 of 11, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

Oh ... yeah, I didn't even think about someone adding a resistor in parallel to the LED. That's a good point. That would explain a lot, and would make it really easy to remove and return the circuit to stock.

Reply 7 of 11, by Anonymous Coward

User metadata
Rank l33t
Rank
l33t

This is pretty weird. I wonder what the hell the previous owner was doing with that header. Maybe it was connected to a super bright LED? Maybe he was using the header to drive something other than an LED...who knows.

"Will the highways on the internets become more few?" -Gee Dubya
V'Ger XT|Upgraded AT|Ultimate 386|Super VL/EISA 486|SMP VL/EISA Pentium

Reply 8 of 11, by tayyare

User metadata
Rank Oldbie
Rank
Oldbie
Anonymous Coward wrote:

This is pretty weird. I wonder what the hell the previous owner was doing with that header. Maybe it was connected to a super bright LED? Maybe he was using the header to drive something other than an LED...who knows.

Anonymous Coward wrote:

This is pretty weird. I wonder what the hell the previous owner was doing with that header. Maybe it was connected to a super bright LED? Maybe he was using the header to drive something other than an LED...who knows.

I removed the resistor and LED is working as it should be now.

This board came from a very old industrial testing device, used to measure forces on springs. Since I'm the only expert in pre W95 stuff, I helped IT, as usual, to make the device working with a newer computer and this baby become obsolete and surpluss. There was nothing extra ordinary connected to the LED header, It was just a run of the mill no-name 486 PC in a beige mini tower, except being EISA and having quality parts like a Tseng Labs VGA card.

While it was working in its original case, turbo led was not connected and turbo button headers was shorted with a jumper. I really have no idea what they were thinking.

IT guy knows that I'm into that kind of thing and he allows me to go thru "the garbage" before he sends them for recycling, hence now I have it. 🤣

Re: Bought these (retro) hardware today

GA-6VTXE PIII 1.4+512MB
Geforce4 Ti 4200 64MB
Diamond Monster 3D 12MB SLI
SB AWE64 PNP+32MB
120GB IDE Samsung/80GB IDE Seagate/146GB SCSI Compaq/73GB SCSI IBM
Adaptec AHA29160
3com 3C905B-TX
Gotek+CF Reader
MSDOS 6.22+Win 3.11/95 OSR2.1/98SE/ME/2000

Reply 9 of 11, by yawetaG

User metadata
Rank Oldbie
Rank
Oldbie
tayyare wrote:

While it was working in its original case, turbo led was not connected and turbo button headers was shorted with a jumper. I really have no idea what they were thinking.

Maybe the original industrial appliance software could only run at a certain CPU/bus speed. Does the turbo button speed it up to a higher speed or slow it down to IBM AT or XT speed, or can the behaviour of the turbo button be defined in the BIOS?

Reply 10 of 11, by tayyare

User metadata
Rank Oldbie
Rank
Oldbie
yawetaG wrote:
tayyare wrote:

While it was working in its original case, turbo led was not connected and turbo button headers was shorted with a jumper. I really have no idea what they were thinking.

Maybe the original industrial appliance software could only run at a certain CPU/bus speed. Does the turbo button speed it up to a higher speed or slow it down to IBM AT or XT speed, or can the behaviour of the turbo button be defined in the BIOS?

I don't have much idea about that, the board had a depleted Dallas chip on it, not able to hold any BIOS setup parameters.

GA-6VTXE PIII 1.4+512MB
Geforce4 Ti 4200 64MB
Diamond Monster 3D 12MB SLI
SB AWE64 PNP+32MB
120GB IDE Samsung/80GB IDE Seagate/146GB SCSI Compaq/73GB SCSI IBM
Adaptec AHA29160
3com 3C905B-TX
Gotek+CF Reader
MSDOS 6.22+Win 3.11/95 OSR2.1/98SE/ME/2000