Power Mac G4 Mdd Acbel power supply (not Samsung)

[CapLeaker]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

there you go! The 12V rail is shorted. Trace it back to the diodes mounted on the heat sink. These have 3 legs. Test the middle leg (negative) to the 2 outer legs (positive) with your multi meter in diode mode. If you find a diode shorted, reverse the leads, put the positive probe on the middle pin and probe the 2 outer pins with the negative probe. If you find something questionable, ALWAYS take it out and test the component out of circuit.

Remember, to pull up the spec sheet of the diode on the internet. You have to know the voltage drop. Schottky diodes have a very low voltage drop @ around 0.2V compared to a regular diode (0.5 to 0.7V)

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

Perfect
Can you tell me based on the photo I sent you which are the diodes to try? on the heat sink there are 5 components on one side and 4 on the other. A curiosity, I did the same test on a working ATX common and when off it gives me the short on both red, yellow and orange, why? yet it is functional.
I'll update you later thanks

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

I disassembled the heatsink and the diodes do not give any short but also no value, I noticed that the red is still short, despite there are no diodes on the printed circuit

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

(maybe it will be stupid) but I also noticed that putting the test leads on the red wire and on the output of the transformers gives me the short on all the feet

[CapLeaker]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

Look at the MBR3045 look at the symbol below it. There are 2 diodes in one package. The center pin is the cathode. I think there are 4 MBR3045 and 2 other ones in a smaller package. Make sure none of them is shorted with your DMM in diode mode.

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

Look at the MBR3045 look at the symbol below it. There are 2 diodes in one package. The center pin is the cathode. I think there are 4 MBR3045 and 2 other ones in a smaller package. Make sure none of them is shorted with your DMM in diode mode.

Hello
I tried but I have no value on any pin, maybe I have to remove it from the heatsink?

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

Nothing, it does not give me any value, I also tried with another dmm

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

I did it, the diodes are ok they don't give me any shorts, and the mosfet is ok too, now what?

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

If it can be useful I have a working acbel and it also gives me a short circuit on the red wire, so I think it's normal

[CapLeaker]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

If it can be useful I have a working acbel and it also gives me a short circuit on the red wire, so I think it's normal

It must be O.K. then. Back to basics: So when you turn the PSU on, do the 3.3V, 5V and 12V rails briefly turn on? What voltage do you measure? You can use the min max function on the DMM, if it has it.

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

ok tomorrow I will do this test, please try to be clearer, do I have to measure the output voltages on the psu working? I don't know what this function is on the DMM

[CapLeaker]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

On your very first post, you stated that the LED turns on for a brief second, then fades away. I wonder if all the voltage rails turn briefly on, maybe one rail does not. No you don't have to measure anything right now on the working power supply.

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

ok I try to do this test even if it will not be easy, because it gives me that hint of ignition twice after it dies, to make it redo the operation I have to turn off the power leave it off for a while and then do it again, I can tell you, that when the power supply unit also starts, the processor and power supply fan also give a hint

[CapLeaker]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

it's kind of hard to work off of pictures and not having a schematic for it. There is this green little daughter board on the secondary low voltage side. Make sure it is soldered properly / inspect it for cracked solder joints.

BTW, do you have a variable constant voltage, constant current power supply?

[CapLeaker]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

ok I try to do this test even if it will not be easy, because it gives me that hint of ignition twice after it dies, to make it redo the operation I have to turn off the power leave it off for a while and then do it again, I can tell you, that when the power supply unit also starts, the processor and power supply fan also give a hint

so maybe it is shutting down from it thinking it is a over / under voltage condition?

[R_J]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

(maybe it will be stupid) but I also noticed that putting the test leads on the red wire and on the output of the transformers gives me the short on all the feet

How are you checking for the short? what is the meter range set to and what is the actual resistance?

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

the daughter card should take care of the fans I don't think that's the problem, what type of power supply do you say ac or dc? and how many volts?

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

How are you checking for the short? what is the meter range set to and what is the actual resistance?

now I don't remember I'm not in the laboratory tomorrow I'll let you know

[R_J]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

The reason I ask is that some people use a meter that beeps and assume there is a short.

[falcontech]

Re: Power Mac G4 Mdd Acbel power supply (not Samsung)

however tomorrow I want to try to move the heatsink block with all the diodes in the working one and if it were to give a negative result I want to try to move the transformers I doubt that they are the ones, could it also be the integrated lm ****?