That why this is incorrect circuit.

Use mine, either one works great:

The power good signal is held high till PSU internal detection of their outputs stabilizes then power signal is pulled low through a low value resistor to ground and held there as long as computer is on.

So:

Diagram correctly done is like this:

positive 5V through a 470 ohm resistor to limit current to power good signal when held low and is pulled high when switch is open. And a switch between Power good to ground through 10 ohms resistor. This switch grounds the power good signal when switch is on. That it.

Resistors are there to prevent excessive current and good idea to have cheap resistor burn out if there's a accident than the hard to replace chipset.

Another way:

Power good grounded via 220 ohms resistor to ground. Momentary push button between power good and 5V supply via a 100 ohms resistor. Turn on the PC and push the button once to reset the computer. Also serves as reset button too.

Cheers,

Can u show what happen on the post card when u switch the button?

pentiumspeed wrote on 2021-07-25, 20:41:

Incorrect diagram. […]
Show full quote

Incorrect diagram.

The power good signal is held high till PSU internal detection of their outputs stabilizes then power signal is pulled low through a low value resistor to ground and held there as long as computer is on.

So:

Diagram correctly done is like this:

positive 5V through a 470 ohm resistor to limit current to power good signal when held low and is pulled high when switch is open. And a switch between Power good to ground through 10 ohms resistor. This switch grounds the power good signal when switch is on. That it.

Resistors are there to prevent excessive current and good idea to have cheap resistor burn out if there's a accident than the hard to replace chipset.

Another way:

Power good grounded via 220 ohms resistor to ground. Momentary push button between power good and 5V supply via a 100 ohms resistor. Turn on the PC and push the button once to reset the computer. Also serves as reset button too.

Cheers,

Hello, okay so what you say is for short:
For the first method :

+5v to Power good (through 470 Ohm resistor)
Power good to Ground (through 10 Ohm resistor)
Switch "ON" (pin) to ground (through 10 ohm resistor)
Switch "ground"(middle pin) to Power good

Like this?

Did i understand everything? (Sorry it's hard to understand just by reading the whole thing)

And for the second one :
POwer good to ground (through 220 ohm resistor)
I didn't get the next one 😖

Also why all of this different type of resistors? I only have 100 ohms ones 😒
Please could you explain really simply so i can understand because it's really hard for me to get everything, i would like to try but i'm kind of lost i have to say.

Here you can use the blank diagram if you want to get to the point, it would be easier for us to understand 😀

dataino.it wrote on 2021-07-25, 20:54:

Can u show what happen on the post card when u switch the button?

Here you have it 😀

the reset led is all the time on ? (the bottom right led)

dataino.it wrote on 2021-07-25, 21:11:

the reset led is all the time on ? (the bottom right led)

Yes, it stays all the time like you see on the photo.

That is why. The one you made is not correct way to do the power good signal.

Cheers,

pentiumspeed wrote on 2021-07-25, 21:52:

That is why. The one you made is not correct way to do the power good signal.

Cheers,

Okay, could you tell me what should i modify on my current circuit please? (I'm a total beginner)

Look at previous comments of mine.

And think about it.

Cheers,

Verax wrote on 2021-07-25, 20:45:

Hey again, so i tested with the previous diagram : […]
Show full quote

Hey again, so i tested with the previous diagram :

And so nothing happends, no new noises, still nothing on the screen, nothing new.

https://www.youtube.com/watch?v=M8XXzh3Ls38

This circuit is what I suggested, and it should work. Even if we assume the first pin is PWR_GOOD, and now is permanently high (via resistor), the manual reset using the switch should properly boot the machine. I know for a fact this works with typical PC motherboards.

I'm not sure if I understood correctly but did you say the reset LED on POST card didn't change state (on/off) when you flip the switch? That's not good, I would expect ISA but reset signal to follow the motherboard reset signal. Can you try that without the resistor between +5V and pin 21? Pay attention to the reset LED on POST card when you flip the switch. Flip the switch a few times and every time wait a few seconds to be sure, in case there is some sort of delay / glitch filter circuit on the mobo.

EDIT: I've read pentiumspeed suggestion and it would seem this mobo might have the PWR_GOOD signal reversed vs typical PCs. In that case, after you try without the resistor, put it back in but connect it to GND instead of 5V. Rest is the same. Now try the switch a few times again.

Deunan wrote on 2021-07-26, 11:00:

This circuit is what I suggested, and it should work. Even if we assume the first pin is PWR_GOOD, and now is permanently high ( […]
Show full quote

Verax wrote on 2021-07-25, 20:45:

Hey again, so i tested with the previous diagram : […]
Show full quote

Hey again, so i tested with the previous diagram :

And so nothing happends, no new noises, still nothing on the screen, nothing new.

https://www.youtube.com/watch?v=M8XXzh3Ls38

This circuit is what I suggested, and it should work. Even if we assume the first pin is PWR_GOOD, and now is permanently high (via resistor), the manual reset using the switch should properly boot the machine. I know for a fact this works with typical PC motherboards.

I'm not sure if I understood correctly but did you say the reset LED on POST card didn't change state (on/off) when you flip the switch? That's not good, I would expect ISA but reset signal to follow the motherboard reset signal. Can you try that without the resistor between +5V and pin 21? Pay attention to the reset LED on POST card when you flip the switch. Flip the switch a few times and every time wait a few seconds to be sure, in case there is some sort of delay / glitch filter circuit on the mobo.

EDIT: I've read pentiumspeed suggestion and it would seem this mobo might have the PWR_GOOD signal reversed vs typical PCs. In that case, after you try without the resistor, put it back in but connect it to GND instead of 5V. Rest is the same. Now try the switch a few times again.

Okay well i trust what you say, unfortunately it didn't work for me. 😒

To be clear, the red reset LED on POST card is always on when the PC is turned on with PSU. And it doesn't change anything when using the manual switch i just installed.

I can try that yes, so no resistor between +5V and pin 21 (Power Good)? I'll take a photo or a video of the post card at that moment.

Okay so playing with the switch for some times and looking the reset post led got it.

What do you mean "Reversed Power good"?
OKay so second test resistor to ground instead of +5V.

I'm going to do all of that and i'll be back after ! 😀

Okay i did the two tests :

- First one with resistor POwer good to Ground = reset led still on
https://youtu.be/fyTb0Y52560

- Second one whitout the resistor = reset led still one
https://youtu.be/svIgTBon_iU

Verax wrote on 2021-07-26, 12:43:

Okay i did the two tests :

- First one without resistor POwer good to Ground = reset led still on
- Second one whitout the resistor = reset led still one

I call it reversed PWR_GOOD because in typical power supply the signal is low, then goes high when voltage is stable (actually the voltage is usually not tested, it's a time delay of about 100ms). But it would seem on the Olivetti machine it's the opposite, the bad is high state and good is low state. This is why I just suggested to try with resistor to GND instead of 5V, and also without resistor at all to see what happens.

I assume the second try you meant to write "with the resistor", not "without"? What value resistor did you use for this test, 100 ohm as suggested? Maybe this mobo has unusually strong pull-up resistors and this value is not good. But that can be tested - please try again both position of the switch (leave the other resistor as-is for now, preferably to GND) and measure what voltage you get at pin 22 each time. For the motherboard to detect the signal as low the voltage has to be belo 0.8V, and for high it needs to be higher than about 2.5V (preferably higher than 3.5V).

In ideal conditions you should see 0V and 5V, this is what the switch is connected to. But there will be some drop on the resistor and maybe the drop is too big. It's a simple test so worth investigating.

EDIT: Oh I see on YT that you are using the modified circuit with only one side of the switch connected and another 10 ohm resistor? This will not work. Please try with the circuit you made earlier.

Okay, to make sure we are not misunderstanding each other, here is what I mean.

First, make a cut where the X is. Test like that with each switch position.
Then make a connection where the dot is next to arrow head. Test again both switch position. Measure voltage in point 22 in each position as well.

I hope that will work now - at least for the LED.

Deunan wrote on 2021-07-26, 12:58:

I call it reversed PWR_GOOD because in typical power supply the signal is low, then goes high when voltage is stable (actually […]
Show full quote

Verax wrote on 2021-07-26, 12:43:

Okay i did the two tests :

- First one without resistor POwer good to Ground = reset led still on
- Second one whitout the resistor = reset led still one

I call it reversed PWR_GOOD because in typical power supply the signal is low, then goes high when voltage is stable (actually the voltage is usually not tested, it's a time delay of about 100ms). But it would seem on the Olivetti machine it's the opposite, the bad is high state and good is low state. This is why I just suggested to try with resistor to GND instead of 5V, and also without resistor at all to see what happens.

I assume the second try you meant to write "with the resistor", not "without"? What value resistor did you use for this test, 100 ohm as suggested? Maybe this mobo has unusually strong pull-up resistors and this value is not good. But that can be tested - please try again both position of the switch (leave the other resistor as-is for now, preferably to GND) and measure what voltage you get at pin 22 each time. For the motherboard to detect the signal as low the voltage has to be belo 0.8V, and for high it needs to be higher than about 2.5V (preferably higher than 3.5V).

In ideal conditions you should see 0V and 5V, this is what the switch is connected to. But there will be some drop on the resistor and maybe the drop is too big. It's a simple test so worth investigating.

EDIT: Oh I see on YT that you are using the modified circuit with only one side of the switch connected and another 10 ohm resistor? This will not work. Please try with the circuit you made earlier.

OKay it makes sense now thanks for the precisions. 😀

Yes of course it was an error sorry!

I only got 100 ohms resistors actually yes.

So like this one : But with resistor to ground not +5V, and i with the multimeter one side on pin 22 and the other one on ground?

I don't know what you are talking about sorry (that's why i prefer visual diagram 😖) on the videos i've tried exactly what you explain one message earlier. 😀 Like i said, all of the resistors are 100 ohms, i only got them for now.

Edit : so with previous circuit, testing pin 22 = 0.49V - same in both switch position, and reset led still on.

Verax wrote on 2021-07-26, 14:33:

So like this one : But with resistor to ground not +5V, and i with the multimeter one side on pin 22 and the other one on ground?

Yes, correct.

Verax wrote on 2021-07-26, 14:33:

Edit : so with previous circuit, testing pin 22 = 0.49V - same in both switch position, and reset led still on.

Well, that's not good. I wonder if the switch is perhaps broken? After all one side is connected to 5V so in one position the voltage should be higher than this.
Try again but now measure the voltage at the switch pin (so, on other side of resistor that goes to 22). Make sure the switch actually provides 0V (GND) and 5V in each position. If not, try a different switch, or perhaps even use wires for the test. Just make sure not to short 5V to GND by accident.
Unless by "previous circuit" you mean with just one side of the switch connected? In that case rebuild the circuit first and try again.

Deunan wrote on 2021-07-26, 19:49:

Yes, correct. […]
Show full quote

Verax wrote on 2021-07-26, 14:33:

So like this one : But with resistor to ground not +5V, and i with the multimeter one side on pin 22 and the other one on ground?

Yes, correct.

Verax wrote on 2021-07-26, 14:33:

Edit : so with previous circuit, testing pin 22 = 0.49V - same in both switch position, and reset led still on.

Well, that's not good. I wonder if the switch is perhaps broken? After all one side is connected to 5V so in one position the voltage should be higher than this.
Try again but now measure the voltage at the switch pin (so, on other side of resistor that goes to 22). Make sure the switch actually provides 0V (GND) and 5V in each position. If not, try a different switch, or perhaps even use wires for the test. Just make sure not to short 5V to GND by accident.
Unless by "previous circuit" you mean with just one side of the switch connected? In that case rebuild the circuit first and try again.

So i retry everything that everyone said previously (you can go back to see in details), checked everything, still nothing is happening.

I don't think the switch is broken, but i can test it so you said one side of the multimeters on one switch pin and the other on which one?

So how do i proceed to make a "manual wire" reset (without the switch)?

I'm not sure anymore about which diagram you are refering haha, so i remake this one actually that's my circuit :

I think maybe you are trying too many things at once, so the results are confusing - we all loose track of what is being tested.

So, the diagram you just posted is OK except you are supposed to connect the resistor to GND instead of 5V. So not to pin 29 but 25-27.
Also, you did not check the voltages with this circuit. Please check the voltages on pin 22 now.

As the pinouts between the PCS-33 & PCS-11 appear to be the same (please correct me if I'm wrong here), I'm wondering if it's not actually the power supply at fault here (assuming all your voltages are correct of course) and in fact something wrong with the Motherboard or Riser itself.

The reason I say this, as per my previous posts in this topic - I was and am still able to get my PCS-11 up and running by simply splicing together the wires, with no resistors or special circuits. I don't even have -5V or "Power Good" connected, nor is the Reset connected. This results in my PCS-11 occasionally failing to POST (being stuck at reset), with no screen output or activity aside from a spinning HDD. Toggling the power off, wait 5-10 seconds, then power back on usually kickstarts the machine to life and it POSTs. So if the PCS-11 can handle the lack of -5V, Power Good & Reset with just some minor no-POST issues, then I'm thinking you may have other issues aside from just a power supply issue.

PTherapist wrote on 2021-07-28, 16:03:

This results in my PCS-11 occasionally failing to POST (being stuck at reset), with no screen output or activity aside from a spinning HDD.

This is exactly the kind of issue I would expect with missing reset / PWR_GOOD signal (or when timing of PWR_GOOD is somehow not to mobo liking). But the problem is no two motherboards, even of the same type, are identical. With LSI chips the behaviour on power-up (and without proper reset) can be very different between each chip. In other words your mobo might mostly boot and sometimes not but that does not mean it's universally true. His might almost never boot for example.

Once the setup with resistor to GND is tested, and voltage measured to make sure the switch works as expected and there is still no no life signs from motherboard, then I would assume it's somehow faulty. So far however all the combinations that have been tried might actually prevent correct operation - and the reset LED on POST card being always on confirms it.