Something like this...

https://www.youtube.com/watch?v=YaRDbw38x7Q

Depending on the ON time i.e how long FET's are switched ON against the time they are OFF... will determine on the voltage levels.

The PWM determines the time ON versus the time OFF, and drives this part of the circuit.

mjnman wrote:

Before with tester I identified the ground on PSU output cable, the cable to connect at the motherboard. I connected only the ha […]
Show full quote

Before with tester I identified the ground on PSU output cable, the cable to connect at the motherboard.
I connected only the hard disk drive and floppy drive and powered on the PSU. With tester I tried to check the output voltage ad I found strange values beacause the voltage is always very variables during the readings:
From top to bottom the colors in the photo below:
1° BROWN: From +4.3V to +4.7V
2° BROWN: From +4.3V to +4.7V
1° BLACK: Ground
2° BLACK: Ground
RED: From +5V to +8,9V
BLUE: From -5V to -8,9V
WHITE: From +0,15 to +1,15V

Okay, so your PSU pinout is most likely this:
BROWN: +5V
BROWN: +5V
BLACK: GND
BLACK: GND
RED: +12V
BLUE: -12V
WHITE: PWR_GOOD (power good signal)

The power good signal should rise to 5V and stay there to indicate the PSU has reached correct operating state. In your case it's fluctuating and that's not good but in general the voltages you listed are in the right range. So it could be the load is still too small and the PSU is unable to work properly. These older PSU (well up to ATX era actually) have only one feedback and regulate one single voltage, the rest is "scaled" to that. Usually you regulate the highest voltage (so 12V) to have less error on the other ones but in this case it might be the 5V is regulated because it's the main power draw (and for logic it's important to get clean power).

But anyway - first check if your meter is working correctly. You don't want to be looking for problems in PSU if the weird values are due to your meter battery being low for example.
Next - I assume you've replace the PSU transistor with the same part? If so, and again assuming the PSU unstable voltages are due to not enough load, the question is why it doesn't work with the mobo. Well, maybe the mobo is the problem now. The clicking sound you get is the PSU trying to start but probably tripping on shorted output.

Disconnect the PSU and test the motherboard now. Measure resistance between GND pins and the rest - +5, +12 and -12. You should see at least a couple tens ohms, except maybe the +5V might be down to less than 10 ohms. But not very close to zero - anything below 1 ohm is a short.

I apologize for the ignorance, I tried to connect the PSU at motherboard and powered on. With tester I found the correct values and the voltage is fixed ...below:

1° BROWN: From +5.04V
2° BROWN: From +5.04V
1° BLACK: Ground
2° BLACK: Ground
RED: From +11.75V
BLUE: From -11.74V
WHITE: From -5.05V

At this point I think that is the end because I don't have the necessary skill to repair the mother board 🙁 😢 😢 😢
When powered on the keyboard led is lit on and hear a little bip but withour video signal. The ctrl-alt-canc can be recognised because I can hear the beep and three light on mother board blink every time that push it.
At this point I think that the mainboard is broken.
Thank you for your assistance

it's incredible...maybe a miracle. I unsoldered the new CMOS battery (Installed new one some months ago) and now it's started and boot os from hard disk!!!
Thank you guys...I'm very happy now 😁 😁 😁
tomorrow I will try with 8 bit isa sound card, new os install a so on.
Thanks everyone

Good stuff... glad to hear you got it working!

In the end, did replacing the T3, BU903 make a difference? or would the fuse have blown without changing this?

Also, what type of CMOS battery does the system have... barrel or coin cell? Do you have a picture of the board even?

Good end to the story... really happy for you.

That's good to hear! The Olivetti PCS is a nice compact machine, deserves to be saved.

Your problem with CMOS is strange indeed.
My system had a 3.6v barrel battery.
Replaced with a 3v coin cell, with battery holder, placed far away from the motherboard. I've also added one diode for protection, i don't want the battery to be under excessive voltage when the computer's on.
Despite the 3v and diode, the coin cell is good enough to pass the initial RTC check.Still holds after a couple of years.

At the end the main problem was the transistor T3, BU903 which break down the fuse everytime.
In the first time anyhow I replaced two gold capacitors from 0.47 microfarad and orange 100 microfarad.
When I receied this Olivetti I replaced the CMOS and for many months it's worked I don't know because this battery now don't work.
If the battery is connected I cannot see any output from video .
It's a 60mAh 3.6V battery. I will try to replace with another one.

What is the battery voltage now? Did you check?
Also, what is it connected to? One terminal should be ground (that goes to battery minus) and the other should go to +5V via a diode and a resistor to limit the charging current. If that resistor is somehow shorted now, the battery might be overloading the PSU - although I would think the PSU would "win" that battle and just boil/burst the battery...

I checked the battery now and it's 3.65v. I verified as is without a connected pin.
I remeber that it's connected with correct way (+ and -). I removed the old battery, many months ago, and replaced with this. It's worked.
Do you think I need to add a diode? I don't undertand the reason because the old original battery was without diode.

The voltage is correct so it's not shorted internally (and as I've said for this to be a problem the mobo would also need to have that resistor shorted). So no need to add a diode, and after all it worked for some time without any problems. I'm just trying to figure out why the PSU was not starting. Maybe the battery is a coincidence? Did you move or reconnect anything at the time you cut the battery connection?

You can actually test this battery again. You will need a piece of thin wire - you can use 1 strand from a thicker copper cable. Connect the battery terminal with this, I think you have enough of the leads left to just wrap it around, no need to solder. Then try powering the PC up. Since the wire is thin it will act as a fuse, so if it smokes/glows/breaks then you have some sort of battery charging issue.

I did mention the diode only because i replaced mine with a cr2032, had to. Sorry for the misunderstanding.

your battery seems to be a good replacement of the original one, 3.6v so it should work fine as is. But it's also rechargeable, i don't remember if the original one was nimh or lithium. Do you?

Maybe Cmos memory just needed a reset and removing the battery did it? I don't know...

I replaced the CMOS battery with another one, same model. Now the computer can poweron correctly and I reinstalled the OS, drivers and some games.
I don't know the reason about the old CMOS battery because the voltage is always 3.6v. I'll keep going with usage and tests.
Thank you

Just started on this journey. I've got my PSU stable (thanks MJN for the voltage levels, it *really* helped working backwards to find the issues). I've got the system now to the point where it'll boot MS-DOS. So, feeling chuffed. I have a rather odd question, does anyone have a scan of the user manuals? I can't get into the BIOS and haven't found a utility that seems to work to set up the system.

Boot Screen: http://ns4.reboot.net.au/hardware/pics/misc/O … _MS-DOS_3.3.jpg

Some pics (yes I know I need to clean it function over form.. ): http://ns4.reboot.net.au/hardware/pics/misc/Olivetti-PCS86/

Really appreciate any insight. Thanks for your time 😀

Regards,
Al.

Hey, Hi!
Surfing the web I found this thread...
I have a faulty PSU for my Olivetti PCS 86 too.
I found the issue of mine was the blue cable to the motherboard, the voltage varied from -4V to -8V, instead to remain stable to -12V.
BTW, the blue line is just a service line that is used to adjust the contrast of the video signal.

I am looking to find a brand new miniatx PSU that fits the case to substitute the Olivetti's PSU, that it's not repairable.

@mjnman feel free to connect with me, I am Italian too. We can find a common solution.

Cheers.

Hi, i believe the problem are in the electrolytic caps.
Try to replace all electrolytic caps and recheck the voltages.
P.S. absolutely replace the two 0.47uF rifa caps!!!!

Thanks @mjnman for the suggestion.

Just a question, have you a "official"/"stable" mapping of the voltages and max currents for the PSU connector?
For example:

1° BROWN: +5V (± 0.7V tolerant) <- OK
2° BROWN: +5V (± 0.7V tolerant) <- OK
1° BLACK: Ground <- OK
2° BLACK: Ground <- OK
RED: From +5V to +8,9V <- ?! not sure of this for example. Should be not a +12V?
BLUE: From -5V to -8,9V <- not sure even, this should be -12V I think
WHITE: Power good. +1V (± 0.15V tolerant) <- sure?

Thanks in advance,
Francesco

I wrote my voltages some post ago. The computer start correctly with my voltage...I believe are ok.

Below the values:
1° BROWN: From +5.04V
2° BROWN: From +5.04V
1° BLACK: Ground
2° BLACK: Ground
RED: From +11.75V
BLUE: From -11.74V
WHITE: From -5.05V

mjnman wrote on 2022-11-01, 16:01:

I wrote my voltages some post ago. The computer start correctly with my voltage...I believe are ok. […]
Show full quote

I wrote my voltages some post ago. The computer start correctly with my voltage...I believe are ok.

Below the values:
1° BROWN: From +5.04V
2° BROWN: From +5.04V
1° BLACK: Ground
2° BLACK: Ground
RED: From +11.75V
BLUE: From -11.74V
WHITE: From -5.05V

So sorry, I miss that post. It was really strange a "power good" line to +1V 😁
Thanks again @mjnman

After the repair of the power supply, my next mission is to bring new power to the two gotek(s) and to the 8MB compact flash I put as hard drive. Just experiments...

Yes, the pinout is, memory to front:
+5 / +5 / 0 / 0 / -5 / +12 / -12 / PG

My PCS 286 used the same pinout (including colors and missing -5 V) but I had to replace the power supply with a generic one, as not only the X capacitors exploded after 10 seconds of use but all of the rectifier, transistor, and inverter chip were bad -_-, so I was about to connect the cables with matching colors but luckily I wondered why a 1987 computer needed more current on +12 than +5, so I found out the real colors by measuring continuity between the power socket (it's a "MODU 1" by the way) and the ISA slot!

Then of course I had to flip the fan and remember to not put the finger in the empty switch hole...

Hi guys,

sorry for digging up the old thread, but i guess my problem is quite similar...

I'm trying to resurrect an Olivetti PCS 286 machine at the moment and the power supply stands in my way. The fuse was blown, i already changed the two Rifa/ Kemet 0.47 Uf caps and put in a new fuse. At least there's no explosion anymore. When turning on, i hear a click (but no fan) and can measure the following voltages (while connected to mobo with hdd) .

1° BROWN: +5.06V
2° BROWN: +5.06V
1° BLACK: GND
2° BLACK: GND
RED: -0.19V (shoud be +12V)
BLUE: -12.10V
WHITE: +5.06V (shoud be -5V)

RED and WHITE seem to be wrong, but i don't know if that's a problem created by capacitors or any other component. I just measured the bridge rectifier, which seems to be ok.

Maybe anyone has a hunch?