VOGONS


First post, by pentiumspeed

User metadata
Rank Oldbie
Rank
Oldbie

This is unique power supply design.

Not the UC3844 IC the heart of the power supply, it is the unique thing about the power switch is two wires is isolated and connected to one winding in series of a small isolated transformer, the hot side winding is constantly oscillating with simple few TO 92 transistors detector circuit and when switch is closed, this shorts the winding out changing the impedance of transformer and turns UC3844 on.

That said, other than that, normal power supply design with few extras.

The problem is every time I have it under load even with no load even disconnected from motherboard, the PSU's fan is barely twitching every just bit over every second. The output voltages reading is low, around 1V to 2V on average. Checked the capacitors with ESR, they passes. Checked diodes and transistors, they checks out. Outputs checks out for shorts.

There's no blown nor any exploded things.

Still troubleshooting. The PSU is in pieces, main transformer, all the heatsinks with their components removed for better access and ease of troubleshooting.

The power supply also is unique: 3 of the wires connected to motherboard are in series with each onboard LED (three LEDs) for status, all LED is tied together to one supply rail, no else is connected to power supply's. There is 2 wires that is remote sensing of ground and 3.3V. Other than that, normal 5V, -5V 12V -12V power supply rectifier and regulators circuits plus power good and shut down wire.

Cheers,

Great Northern aka Canada.

Reply 1 of 5, by chrismeyer6

User metadata
Rank Oldbie
Rank
Oldbie

Have you tested the the voltage with the fan disconnected? I've seen that behavior when I had a fan that was shorted internally. Can you post some pictures of the internals?

Reply 2 of 5, by pentiumspeed

User metadata
Rank Oldbie
Rank
Oldbie

More discoveries:

I thought this power supply followed the ATX design on generating 3.3V using unrectified pulses via inductor coil actually is not, in fact that is active DC-DC converter using CS3842A IC driving two mosfet or transistors and diode supplied by 5V supply and has remote ground and remote 3.3V sensing wires right at the motherboard's. This why it has 2 currents one with 3.3v in use and other without because Deskpro XL uses different processor cards that is either 486 or Pentium.

If you strip away this dc-dc converter, oscillating transformer isolated power switch detector (similar idea to metal detector except secondary side is direct wired in series with switch), to turn UC3844 SMPS IC on and the 3 LED status, it is a low end 1 transistor SMPS based on UC3844 and monitored by a LM339, around rated 250W to account for losses.

Other than that, it not that complicated.

Cheers,

Great Northern aka Canada.

Reply 3 of 5, by pentiumspeed

User metadata
Rank Oldbie
Rank
Oldbie

Update:

Gotten around ordering parts to help with troubleshooting this Deskpro XL PSU.

Capacitors not yet as I am not sure what modern equivalents are available, suggestions on what is current capacitor models? Some of them are leaking electrolyte goo even they tested good.

Thanks and cheers,

Great Northern aka Canada.

Reply 4 of 5, by HomeLate

User metadata
Rank Newbie
Rank
Newbie

I'm troubleshooting the same power supply. Mine works, but the system turns on when I plug in the mains. I have to unplug each time I want to power it down.

I'll let you know my findings when I was able to fix it.

Reply 5 of 5, by pentiumspeed

User metadata
Rank Oldbie
Rank
Oldbie

PS: turning on PSU changes the frequency of the oscillator driving small transformer by shorting across secondary side via the switch. Once the oscillator circuit senses this short, drives one of transistors on.

Mystery about that PSU's 3 LEDs finally solved via this part two youtube: This is for the front panel LEDs via plastic light pipes starting at 2:00 time mark.

https://www.youtube.com/watch?v=DL_cAj2fRFc

Cheers,

Great Northern aka Canada.