# ATX Lab Power Supply Help.

Status
Not open for further replies.

#### Smulle

##### Newbie
Hi,
I recently converted an ATX PC power supply to a lab supply. I connected -12V, +12V, 3.3V, 5V and Ground to binding posts.
I added 2 LEDs to the Gray and Purple wires and connected a 750ohm resistor between the LEDs and ground.
Then I put 2 10w 10ohm resistors in series between +5V and Ground.
Finally, I added a switch between The Green wire and Ground.

It all worked fine the first time, then when I flicked the switch connecting the Green wire to Ground off and on again it just caused the fan to twitch. One of the LEDs was lit, but the power supply was in standby. So I plugged it out and in again, it worked fine flicked the switch on and off again and the same problem, after plugging it out and in again the problem still continued.

So, when I plug it into the mains one of the LEDs turn on but when I turn on the switch the fan inside twitches. I have tried bypassing the switch by connecting the Green wire to Ground but I still have the same problem.

Any solutions?

#### bigdogguru

If I remember correctly, ATX power supplies typically have one or more voltage sense lines (PWR_ON or Power Good) which must be properly handled to power on the unit.

Also some units have a 3.3v voltage sense line which must be pulled up to +3.3v.

ATX Power Supply Wiki

Use a PC Power Supply as a Bench Supply

BigDog

#### Smulle

##### Newbie
If I remember correctly, ATX power supplies typically have one or more voltage sense lines (PWR_ON or Power Good) which must be properly handled to power on the unit.

Also some units have a 3.3v voltage sense line which must be pulled up to +3.3v.

BigDog

Ok, I connected the Brown 3.3V Sense (Brown) to +3.3V (Orange) but I am still having the same problem.
Also found out the hard way that the whole metal case seems to have a 3.3V current running through it.

#### D.A.(Tony)Stewart


...
Then I put 2 10w 10ohm resistors in series between +5V and Ground.
Finally, I added a switch between The Green wire and Ground.

It all worked fine the first time, then when I flicked the switch connecting the Green wire to Ground off and on again it just caused the fan to twitch. One of the LEDs was lit, but the power supply was in standby. So I plugged it out and in again, it worked fine flicked the switch on and off again and the same problem, after plugging it out and in again the problem still continued.

So, when I plug it into the mains one of the LEDs turn on but when I turn on the switch the fan inside twitches. I have tried bypassing the switch by connecting the Green wire to Ground but I still have the same problem.

Any solutions?

The PSU is shutting down instantly from OV due to some condition. The internal OVP (over voltage protection) triggered a shutdown within a 20 ms.... or a twitch ( an official lab unit of measurement....

1. Old PSU SMPS designs need 5~10% preload.
2. New ones need 0~2% preload for regulation.
3. You are using 2x10 Ohm or 5^2/(10+10)=1.25W which is probably in the 0.4% range.
4. Try more preload... i.e. 10 Ohm on 5V = 2.5W or 25% of R capacity and almost 1% of the PSU ( assuming 300W)
5. Then try two 10 OHms in parallel. or 5W total or 1.7% preload.
6. If your power miser sensibility is miffed replace the resistors with a 12V 8W PAR Lamp to run dim. (On 5V) over your workbench.

#### bigdogguru

The second link, I previously posted, offers several step-by-step examples of how to configure a ATX power supply as a bench power supply.

Including the use of idle loads as outlined by SunnySkyguy above.

Also carefully observe the pin number, rather than any coloring is advisable, when make your connection.

BigDog

#### Smulle

##### Newbie

The PSU is shutting down instantly from OV due to some condition. The internal OVP (over voltage protection) triggered a shutdown within a 20 ms.... or a twitch ( an official lab unit of measurement....

1. Old PSU SMPS designs need 5~10% preload.
2. New ones need 0~2% preload for regulation.
3. You are using 2x10 Ohm or 5^2/(10+10)=1.25W which is probably in the 0.4% range.
4. Try more preload... i.e. 10 Ohm on 5V = 2.5W or 25% of R capacity and almost 1% of the PSU ( assuming 300W)
5. Then try two 10 OHms in parallel. or 5W total or 1.7% preload.
6. If your power miser sensibility is miffed replace the resistors with a 12V 8W PAR Lamp to run dim. (On 5V) over your workbench.

I think I might have a switch mode power supply if so it works out that I need 15W?

Status
Not open for further replies.