psu ac240es-01 from dell Optiplex

[MegaZAC]

PSU does not start. There was attempt to fix it that found some faulty component, probably a MOSFET, that was removed, and one smaller component next to it also was removed. Both are unknown. Also, there is no diagram available, PSU doesn't start at all, no light turns on, no fan movements, nothing shows any activity. Not sure where to move forward.

[MegaZAC]

The removed components on the board are marked as HS7 (MOSFET) and L11. The MOSFET one is linked to the nearby capacitors 10V 2200uf. L11 is linked to the MOSFET with one end, and with the purple wire of the ATX connector, which is STANDBY. Obviously, no Standby, no start. But what do I need to put there to make it work?

The PSU is marked as outputting +5.0VFP—4.0A, +5.0V—11A, +3.3V—3.5A, +12V—17A, 80 Plus Gold. So standby shouldn't be something powerful, however, should be smart and low-heating due to the Gold standard requirements.

Btw, I would consider to increase output from 12V to something like 19V with proportional decrease of ampers. But that's a side hustle, the main task is to bring it to life first.

[harp]

Do not connect psu in power.
You have to clean this mess on pcb, before anything...
HS7 looks like double diode on heatsink. L11 is coil.
Check for shorts on output connector, and then on pcb especialy zener diodes, switching diodes and mosfets.

[MegaZAC]

You are right. The more I checked the worse it looked. Zener diod ZD5 on the pic is open in both directions in diod check mode, but in resistance check mode it's open only in one direction. 24 pin ATX mb connector is shorted nearly completely. Ground is shorted with nearly everything, 5V is shorted with 3.3V and both with ground. The only thing that is not shorted with them is 12V, but is also shorted with the ground.



The HS7 has side legs shorted with the ground, the center leg isn't.
Two 3-leg components D26 and D27 from the pic have a radiator and both have a center and one side leg shorted with ground.
The most left 3-leg component from the pic (on the previos pic marked as M12) has only the center leg with the ground.

Is it completely dead? Tell me that I'm doing something wrong and there is a chance to revive it. May I extract at least 12V? I traced it from the 24-pin connector to the mosfet FDD8896 and APW7073 PWM. Not sure if it goes further as 12V on one of them or as shorted to ground in them.
From the other side there is a group of 3-leg components on the biggest radiator from all. Can't see what's written there, but should be a sort of diods too. Not all of them are open in all directions.

[harp]

Oh... from where to I start...

You must learn the difference when express yourself that something is shorted, connected or have low resistance reading.

So, I have a questions
Have you these two faulty components that was removed arround you?
Is this psu yours, or if you know that is working before attempt to fix?
What was simptomps of failure and how it is presented in long time or instantly?

If you do not idea how to perform a fix, it is ok, you may learn that, but in meantime sharp pictures (and not angled) will be wellcame when time come

If you have some others known good atx psu, you can compare readings on output connector that ensured you nothing is concerned like shorted.

[MegaZAC]

Sorry for confusion. My understanding is that shorted is when it's connected against design, while normal connection is according to design.
All voltages in the 24-pin connector have 1.5kOm resistance with the ground and it keeps quickly growing. PSON with ground has resistance starting from 10kOm.

The simptoms are the same - it is dead, doesn't start, no signs of life. To me it was given already dead, but with signs that it has been working. I don't have the removed components, but I have one that probably (hopefully) was the coil from L11. The HS7 is missing completely and it had a visible damage on it. But if this is the only reason for failure and we know what exactly shouild be there, I can get it.

I am afraid that without desoldering half of the components top photos will not be clear, as everything it very close with no free space.

[harp]

So one time repeat, all measurement at this time till clear noted are done under disconnected from mains power.

I dont know yet have you or not have shorted outputs, maybe is best to check again and write here. So, multimeter in resistance mode, connect black and red leads to check if zeroed, and put black probe to black wire (secondary, cold ground) and red probe to:
-orange (3.3v)
-red (5v)
-yellow (12v)
-blue (-12v)
-purple (5vsb) must be repair first

In order to psu start you must have 5vsb, because it power chips and signal to green wire.
It will be great if you can retreive missing parts especially small heatsink.
For testing purposes you can put another fast switching diode (like sb560) and connect traces over missing coil.
I guess (regardless to blue dots over the board) it is not only reason for failure, but without repair 5vsb you cant continue.

In meantime you can check continuity on primary (hot) side, put dmm in diode mode and check continuity between 1 and 1, also 2 and 2, and 3 to 3 from picture. There must be reading of 0.000 voltage drop or near.
Check bridge rectifier then, on all diodes must be reading about 0.5xx - 0.6xx, but never low reading like 0.012 or so...

I guess that you are familiar with doing basic measurement, and recognizeing basic components.
Pictures are ok.

Noted thisfor future
Check mains capacitor connections joint point with arrow, is this spark, gap, dust, shaddow... must be shiny and nice.
There is no load resistor on output so some load may be mandatory
Always use bulb current limiter, toaster or similar when testing with mains
What equipment you have, for desoldering and measuring?

[MegaZAC]

check continuity between 1 and 1, also 2 and 2, and 3 to 3 from picture. There must be reading of 0.000 voltage drop or near.

1 and 2 are fine, 3 shows 2.6V and fluctuates to 2.7V.

I have FSP300-62ld with much easier design for donating components. It has standby built on 5H0165R (according to Google). What can I take from it? I have MBR3045CT 0946A, 1K44AC FQPF 5N5OC, GBU606, and something else, not visible without desoldering.

multimeter in resistance mode, connect black and red leads to check if zeroed, and put black probe to black wire (secondary, cold ground) and red probe to:

That's exactly what I did and got this

MegaZAC[/B]] All voltages in the 24-pin connector have 1.5kOm resistance with the ground and it keeps quickly growing. PSON with ground has resistance starting from 10kOm.

On all diodes with radiators readings are about 0.4xx - 0.5xx and over 1v in one way.
There's no dust or anything wrong. It looks like the psu has worked just for a couple of months.

[harp]

You must have continuity from bridge rectifier either + and - (on both sides) to main electrolitic cap + and -, so in diode mode must show 0.000v between + and +, and - and -. Just follow the trace. On this main HV cap has to be lethal voltage when pugged in mains, so rather be sure that is everything ok and clear in this area. Also check that discharging resistors of HV cap are ok. Update all measure here for others, it is not informative just write all is ok...

MBR3045 may be on place HS7, but find some output coils who fit also on L11.
This psu have 7-DIP standby switching element, and look to see markings on top.

[MegaZAC]

I found the 7DIP switch, the 280 one https://pdf1.alldatasheet.com/datash.../TNY280PG.html Unfoirtunately, it's in a deeply inaccesable place, so to get to it I'll need to desolder middle of the board.

This is the coil that I suspect was there in L11, but I'm not confident. Would it work with MBR3045?

I am not sure which + and - you mean, and which bridge rectifier, which discharging resistors...

[MegaZAC]

The pdf above has a recommended design on page 11, and it matches my board. I marked the 7DIP and nearby transformer.

[harp]

We hope that it is no need to desolder tny280, just confirm that you have not short between pin#4 and pin#5 of it.

L11... on picture is probably not original coil due to layout on pcb. On 5vsb is much more smaler coil and in vertical position. It will work, but you have hardtime to fit it in board. Have you something else adequate on spare boards?

Bridge rectifier have plus and minus on their sides, who is connected to plus and minus of main capacitor, sometimes through pfc diode, so you have to confirm that both paths is ok.
Also, again, check bridge rectifier that not shorted, so in diode mode one leads on plus and check against each AC middle pins, and then same on minus against AC pin... you should have some 0.5xx readings on all points.

[MegaZAC]

Bridge rectifier shows 0.531v if left pin - to the right pin + on the main capacitor.
Bridge rectifier shows 0.531v if left pin - to the right pin + of the same Bridge rectifier
Bridge rectifier shows 0.528v if left pin - to the both midddle pins + of the same Bridge rectifier
both midddle pins - shows 0.533v to the right pin + on the main capacitor
Bridge rectifier shows 0.002v if right pin - to the right pin + on the main capacitor. The same if swap + and -.

Bridge rectifier shows 0.408v if left pin + to the left pin - on the main capacitor.
Bridge rectifier shows 0.925v if second from the left pin + to the left pin - on the main capacitor.
Bridge rectifier shows 0.918v if third from the left pin + to the left pin - on the main capacitor.
Bridge rectifier shows 0.888v if right pin + to the left pin - on the main capacitor.
Bridge rectifier shows open wire (not connected) if right pin - to the left pin + on the main capacitor.

tny280, just confirm that you have not short between pin#4 and pin#5 of it.

It shows 5.4 Ohm between these pins.

[harp]

Rectifier seems good, but tny280 has bad reading, so measure resistance arround it between identical color dots:

Red
Yellow
Green
Blue1
Blue2
Blue3
Green-blue1
Purple

[onix331]

Just made an account to chime in on this.
Currently also repairing the same exact model, what i found so far was a short between pin 4 and 5 of the TNY280, i desoldered the mains capacitor C4 (a CapXon 150µF/420V part) which leaked on the bottom (invisible when soldered in)
The Mosfets and diodes seem fine, though i'm gonna do more measurements before plugging it in.

I was hoping i could combine two units to make one work, but all of them have the same fault (C4 leaky, TNY280 destroyed)

[onix331]

I don't see an edit button so it's gonna have to be a double post, sorry.

I got the PSU to turn on now, the double-diode D10 (original part number 20A60CT) and a fusable resistor? R46 (the rings are GREEN PURPLE GOLD GOLD WHITE, which doesn't make any sense to me) were defective too.
I found a working resistor with that colour band, it seems to be 5.6Ω where the white ring is temperature coefficient.

To summarize:

M5 TNY280PG
C4 150µF/420V
D10 20A60CT
R46 5.6Ω

[MegaZAC]

I don't see an edit button so it's gonna have to be a double post, sorry.

I got the PSU to turn on now, the double-diode D10 (original part number 20A60CT) and a fusable resistor? R46 (the rings are GREEN PURPLE GOLD GOLD WHITE, which doesn't make any sense to me) were defective too.
I found a working resistor with that colour band, it seems to be 5.6Ω where the white ring is temperature coefficient.

To summarize:

M5 TNY280PG
C4 150µF/420V
D10 20A60CT
R46 5.6Ω

Good job!

Can you take a look what components you have in HS7 (MOSFET) and L11 (coil)?

Sorry, HS7 is the D10 that you mentioned, this question is answered. May you please post the name or photo of L11 (coil)?
Also, check if you have a dead SMD as said here https://www.youtube.com/watch?v=AxeL7if_IDQ

[onix331]

Good job!

Can you take a look what components you have in HS7 (MOSFET) and L11 (coil)?

I don't know where HS7 is supposed to be, my board only has three Heatsinks (HS1-HS3)
I've attached an image of the coil though

[MegaZAC]

Great, thank you. Also, check if you have a dead SMD as said here https://www.youtube.com/watch?v=AxeL7if_IDQ
I'm thinking, if tny280 are bad, should I order tny290pg or something else?

[onix331]

I don't know where that resistor would be, maybe it's for a different 7010 PSU