VOGONS

Search

Common searches

ATX PSU - wiring power LED

Topic actions

First post, by megatron-uk

Posted on 
User metadata
Rank Oldbie
Rank
Oldbie

I am in the process of integrating a Pico PSU with my tiny 486 SBC build.

Everything is working except the operation of a power LED.
I have a switch connected to PS_ON and COM (specifically ATX pins 14 and 15), and this works to turn the system on and off. When the switch is closed the SBC boots, its buzzer works, the system boots, the hdd led flashes... everything works normally.

But...

I have also wired a LED between +3.3v DC and COM (on ATX pins 2 and 3) as a power-on indicator. This is with an appropriate resistor in line with the specs of the led (it was either 47 or 56Ohm, I forget which). The odd thing is that when PS_ON is shorted, and the system is on, the LED is unlit. With PS_ON open, the LED is lit. This is... not exactly what I thought should happen.

I was under the assumption that +5VSB (ATX pin 9) was the only other pin that had voltage while the system was off... so why is the LED connected between 2 and 3 live with the power off?

My collection database and technical wiki:
https://www.target-earth.net

Reply 2 of 6, by megatron-uk

Posted on 
User metadata
Rank Oldbie
Rank
Oldbie

Hmmm...

Helps if you count the pins on both side of the ATX connector in the same order....

I cannot confirm or deny that I may have transposed pin 2 (+3.3) and pin 9 (+5vsb), since this would be an obvious reason for off=lit and on=unlit....

No. I cannot confirm this at all.

My collection database and technical wiki:
https://www.target-earth.net

Reply 3 of 6, by kingcake

Posted on 
User metadata
Rank Oldbie
Rank
Oldbie
megatron-uk wrote on 2024-03-24, 18:34:
Hmmm... […]
Show full quote

Hmmm...

Helps if you count the pins on both side of the ATX connector in the same order....

I cannot confirm or deny that I may have transposed pin 2 (+3.3) and pin 9 (+5vsb), since this would be an obvious reason for off=lit and on=unlit....

No. I cannot confirm this at all.

🤣 that'll do it

Reply 4 of 6, by momaka

Posted on 
User metadata
Rank Member
Rank
Member

It happens. 😁

BTW, you might want to swap that 47 (or 56?) Ohm resistor with something of higher value. Even with a white or blue or violet LED (typical voltage drop of ~ 3V), such low-value resistor could still allow up to 10 mA of current, depending on how low the LED starts to conduct. With that, the LED might not last too long... and could also be a bit too bright. If it's an "old-world" red or green LED from the early 90's though, then NVM. While both red and green LEDs have a much lower voltage drop, those old LEDs are quite inefficient and require a little more current to be... barely lit, 🤣.

So I guess YMMV.
For modern/current-made LEDs, I find currents of 1-2 mA to be more that enough to make the LED already almost too bright. So I typically use 330+ Ohm resistor (typically 1k for running off of 5V or higher lines.)

Reply 5 of 6, by megatron-uk

Posted on 
User metadata
Rank Oldbie
Rank
Oldbie

Yep, it is indeed an old (probably from a pack some 15+ years old) 5mm green led. I think I recall the specs being something like 2.9v forward voltage and ~20ma. I have matching amber and red LEDs for turbo and HDD activity and they all appear be approximately the same brightness.

Yes, they are all old 😀

My collection database and technical wiki:
https://www.target-earth.net

Reply 6 of 6, by momaka

Posted on 
User metadata
Rank Member
Rank
Member

15 years aren't that old for LEDs. I remember buying quite a few (mostly "standard" 5 mm and 3 mm ones) back in the early 2000's from Radio Shack (in the US) and they all still light up quite well.

By old LED's, I meant mid-90's and older... so 30+ years old.

Go to top of page Go to top of page
Back to General Old Hardware
Close window
Close window