68kMLA Classic Interface

This is a version of the 68kMLA forums for viewing on your favorite old mac. Visitors on modern platforms may prefer the main site.

Click here to select a new forum.
Classic Power Supply with weird voltage, repair
Posted by: ironborn65 on 2020-12-07 10:03:20
Dear all,

I'm trying to fix this PSU but I got strange voltage, here is the story.

Thorough cleaning

I replaced all the capacitors mentioned by Bruce from Branchus Creations.

the two input cracked filters

I replaced the optocoupler OP1, the TDA4605, the MOSFET IRFB040 and L7912 voltage regulator

Diodes DP1 DP3 DP4

The 2 Zener diodes

I had to recreate a couple of pads that were damages by the electrolytes.

It now shows some life now:

  • green wire is -12.5 volt (OK)
  • BUT orange wire is oscillating from 7.1 to 7.4 (it's supposed to be 5 volt)
  • BUT yellow wire is 19,27 volt, it is supposed to be 12volt

The voltage are stable over time.

What can be the cause, a bad voltage regulator?

I do not know much about electronics

thanks for the help

best

PF

do not worry about the empty spaces for the ICs, they have been added 🙂

IMG_20201118_210404__01.jpg
Posted by: dochilli on 2020-12-07 23:19:47
Did you measure the voltages when the AB was connected to the LB and the CRT? Or did you measure the voltages without these connections? If I remember right, my AB showed similiar voltages when it was not connected.
Posted by: ironborn65 on 2020-12-08 09:41:50
Did you measure the voltages when the AB was connected to the LB and the CRT? Or did you measure the voltages without these connections? If I remember right, my AB showed similiar voltages when it was not connected.
hi, thanks for the reply. As a matter of fact I tested it disconnected.

I thought about it, for this reason I added a 12v fan, and the values fluctuated a bit, while the 12 was stable.

*maybe* 7,3 volt would not harm the ICs, 43% more, I dunno...

19 volt instead of 12 volt is a lot, 63% more, ok it's not driving any ICs but I'm worried about my lovely MB

I'm tempted though ...
Posted by: techknight on 2020-12-09 04:29:57
DONT connect this back to the logic board. 

with voltages that high, youll blow the snot out of it. 

There should be an adjustment that should bring the voltages back in line with a potentiometer somewhere near the opto-isolator. 

If that doesnt work, you still have a problem somewhere which will require troubleshooting. 
Posted by: ironborn65 on 2020-12-17 11:59:31
back on this PSU - it is pretty frustrating now

story short the fuse keeps popping

as @techknight was suggesting I mangled with the potentiometer, but no way

so I noticed I was using a CNY17F-3 instead of the CNY17G-3 🙁

I installed the original CNY17G-3 after testing it was ok.

I then had

+4.7/4.7 V

-11/-12 V

+10.6/+10.11 V

Still fluctuating a bit. The +10.6 was too low but at least it was safe to try with the mother board.

There was a small flickering on the screen but after a few second the fuse blown.

The optoisolator was broken so do the IRFBC40.

I replaced them both and I tried again.

I can not test the TDA4607 which is a new one.

This time I had a chess pattern on the screen and after a few seconds .. FLASH, again the fuse was gone.

I attach here the picture of the board and all the replaced components 

Can someone be of help?

IMG_20201118_210404__01.jpg
Posted by: ironborn65 on 2020-12-17 14:06:57
I did some more work on this thing

story short, the fuse does not pop, low voltage, PP1 does not work

my bad I was using a 2A fuse instead of 5A, ok, let's do it gain...

I found the IRBC40 to be broken, replaced

the optocoupler was fine, reinstalled

I replaced the TDA4506-3 just in case (I don't how to test it)

then:

10.6 V 

4.3 V

-10.5 V

no fluctuations this time

BUT

the PP1 trim does not change the voltages a single mV, might it be about the optocoupler section?

or the DP3 DP4 DP1 zeners maybe, but they have been replaced a couple of attempts ago

I measured a total of 1.3A with no load, in case it tells something

Besides changing components I do not have a clue on how to make further diagnosis

sob
Posted by: ironborn65 on 2020-12-18 02:55:58
ah, the MOSFET IRFBC40 goes up to 72 degrees centigrade without load, is it normal?
Posted by: Rjp663 on 2021-01-26 09:06:09
ironborn65,

Did you ever get your power supply issue solved?
Posted by: ironborn65 on 2021-01-27 20:16:14
Hi @Rjp663,

unfortunately not.

The voltages are all a couple of volts below the optimal value, despite my many attempts. The PP1 trim does not affect the voltage, as I reported in more detail above.

I have replaced almost everything, I'm now considering to replace it with a ATX + DC regulator and inject the currents in the AB as @mogs did.

best

PF
Posted by: GiGaBiTe on 2021-01-27 20:58:42
Have you checked any of the resistors? I've had bad resistors cause SMPS power supplies act erratically. They can go open, short or drift far off their rated value. They can also start arcing internally and cause intermittent connections. There are color code calculators online to get the resistance values and tolerances. If the resistor is measured to be outside their tolerance range, usually 5% for carbon composition resistors or less for metal film resistors. There are four, five and six band types, which you look to have the former two on your analog board.

https://www.digikey.com/en/resources/conversion-calculators/conversion-calculator-resistor-color-code

Another culprit could be failed solder joints, which plague compact Mac analog boards. I would suggest reflowing all of the solder joints with fresh solder to ensure good connections. You can look for failed joints with a sufficiently powerful magnifier, they'll generally look like a secondary ring somewhere between the pin of the component and the PCB, but they can also be microscopic and impossible to see. If I see a failed joint on something, I generally reflow everything if possible because the heat that caused one failed joint is going to give everything else trouble.
Posted by: AndiS on 2021-01-28 04:26:45
I also recently had problems with a Classic AB (See this thread)

Short version, it turned out that the PP1 potentiometer was oxidized and did no work anymore, so I still had low voltages even after I replaced most of the parts you also did. Voltage a bit too low gave me a wobbly screen. Even lower lead to the described chess pattern.

Contact spray fixed the issue, but I'm currently still waiting for replacement 220 Ohm potentiometers.
1