Dell Vostro 3350 does not (completely) shut down by Windows

[panoss]

When I am in windows (Win 7), I go to Start->Shut Down.
And the computer beggins the shut down procedure.
The screen after a few seconds freezes and the the Led on the power button does not go off.

The only way then to shut it down is to keep the button pressed for 4 Seconds.

I read that it's a software issue so I did a brand new installation of windows, but it' the same.
I also booted it with a usb puppy linux and the behavior is the same...

What do you think it could be?

This is the schematic.

[LateLesley]

It could be that your machine does not support it. If you look at page 19 of the schematic, it talks about some machines not supporting deep S4/S5. So it may be that the machine is fine, and just doesn't support S5 state, which is the soft off function you are looking for. We get that used to these functions, we forget that they weren't always around. 

https://docs.microsoft.com/en-us/windows/desktop/power/system-power-states

[mikerj]

I have the same problem with the same machine (now tucked away in my loft gathering dust).  It's a common problem, but I've never found a solution.  It's definitely not due to a lack of a power down state, these machines worked fine from new.

[LateLesley]

Well, The only parts involved are the PCH, it sends out the S3/S4 sleep signals, it looks like S5 is only connected to a test point.

On page 46 of the schematic, The PM_SLP_S4# signal goes to the 1.5V regulator enable pin, it does this via a link and a capacitor - PR4607, PC4606 (might be worth checking if it shorted) and then as PWR_1D5V_EN to the enable pin of the regulator, PU4601 (TPS51216RUKR-GP).

Page 27 has the other chip in the mix - U2701A, an NPCE795PA0DX-GP-U Which seems to receive the PM_SLP_S4# signal, and interestingly, outputs an S5_ENABLE on pin 14. It seems to operate a FET, lower down on the page, Q2706 which is part of the keyboard switch circuit. A lot of jelly-bean stuff there which might be worth testing.

Upshot is, it could be jelly bean parts failed, The Multi IO KBC on page 27, the PCH on page 19, or it's software related (which could be nothing to do with windows, but something in the BIOS programming. - Actually, you could do worse than try an old BIOS and see if it works.)

Anyway, I wish you luck in finding the fault.

[panoss]

Ok, I 'll check with some other BIOS version (older or newer) and if nothing succedds I 'll remove the board to check the mentioned parts.
I see these pins related with sleep:
- a pin named 'SLP_S5'
- a pin named 'SLP_S4'
- a pin named 'SLP_S3'
- a pin named 'SLP_A'

What I forgot to mention: Restart works fine!!! Isn't this strange? Shuts down just fine and restarts!
 

[panoss]

Not having it's PSU I 'm trying to adjust an existing one in order to use it with this laptop.
But I have it's cable, which has 3 wires:



Maybe the PSU outputs 2 voltages?

[LateLesley]

It's possible, but it could also be one of those "smart" chargers that talk to the laptop. Some Dells have chargers with 3 connections, and the laptop complains if it doesn't "see" a genuine Dell Charger. So it could be a power, Gnd, and data connector.

[amyk]

I am in agreement with others here to check the voltages while it's in the "zombie" state; the last few pages of the schematic have power sequencing diagrams, which you can use to verify that the signals are as expected.

Restart works because it doesn't require actually shutting off the power rails; it's just asserting the reset pins of the relevant components.

That 3-wire power cord can be explained by page 37 of your schematic---and this article:

https://www.laptop-junction.com/toast/content/inside-dell-ac-power-adapter-mystery-revealed

[panoss]

https://www.laptop-junction.com/toast/content/inside-dell-ac-power-adapter-mystery-revealed

According to the above link an EPROM is connected to the ID pin of the PSU.
So there is no way to make it work without a DELL psu. Oh s**t!!

[soldar]

The PSU and laptop communicate only at connection moment using SMBus. They just exchange greetings and the PSU tells the laptop what amps it can supply.

If this SMBus signal is absent the laptop will still charge but at the slowest rate because it is not aware of the better capability of the PSU.

https://kakopa.webcindario.com/Dell_PS/index.html

[wraper]

Have you installed chipset driver and especially Intel Management Engine Interface driver?  Also, I recommend doing BIOS update if you haven't done already.
It's a very common problem with win 10 on older machines because it automatically updates IME driver to one that does not work properly. I guess that the same thing on win 7 likely is related to this driver as well. Also you could try Intel ME firmware update (not BIOS).

[panoss]

If this SMBus signal is absent the laptop will still charge but at the slowest rate because it is not aware of the better capability of the PSU.

From what I read if there is no communication between the laptop and the charger then it will not charge at all.

When communication with the UniqueWare™ Add-Only Memory fails, the laptop shuts battery charging down.
No battery charging

From here.

[panoss]

Have you installed chipset driver and especially Intel Management Engine Interface driver?  Also, I recommend doing BIOS update if you haven't done already.
It's a very common problem with win 10 on older machines because it automatically updates IME driver to one that does not work properly. I guess that the same thing on win 7 likely is related to this driver as well. Also you could try Intel ME firmware update (not BIOS).

No I haven't installed the Intel Management Engine Interface driver.
The BIOS is updated by the previous owner.
When I resolve the PSU issue I will try the older BIOS.

Let me remind you that I also tested it with a live USB linux (puppy linux) and the behavior was the same.

[wraper]

Have you installed chipset driver and especially Intel Management Engine Interface driver?  Also, I recommend doing BIOS update if you haven't done already.
It's a very common problem with win 10 on older machines because it automatically updates IME driver to one that does not work properly. I guess that the same thing on win 7 likely is related to this driver as well. Also you could try Intel ME firmware update (not BIOS).

No I haven't installed the Intel Management Engine Interface driver.
The BIOS is updated by the previous owner.
When I resolve the PSU issue I will try the older BIOS.

Certainly install IME driver. There is no need to downgrading BIOS and probably it won't even allow doing it.

Let me remind you that I also tested it with a live USB linux (puppy linux) and the behavior was the same.

It probably means this model does not work with it properly to begin with.

[soldar]

From what I read if there is no communication between the laptop and the charger then it will not charge at all.

Not in my experience

https://kakopa.webcindario.com/Dell_PS/index.html

Searching online I find that many people have problems with the 65 W power supply which are resolved by using the 90 W PSU.  It seems the way it is supposed to work is that the computer knows what PSU is connected and will limit its power usage.  If the more powerful PSU is used then it will run at full blast and charge the battery at the same time but if the less powerful PSU is used then it will not charge the battery while the computer is in use and will wait until the computer is turned off.

    The computer and the PSU communicate through the third conductor and if you connect a generic PSU with no third cable then when the computer starts up it will display an error message saying "The AC power adapter type cannot be determined. This will prevent optimal system performance". It seems the computer assumes a very low capacity PSU and will not charge the battery and will throtle back the speed of the processor to conserve energy.

[panoss]

So you 're saying I should try with 19V only?
When the laptop is off?


I read here, in the comments section:

the centre pin is the ID signal, the inner surface of the coax connector is the positive output of the supply and the outer surface is ground.

Is it correct? The inner surface of the coax connector is the positive output? (19V)?

[Bud]

I think it is correct. And the thin center pin is ID of data, not carrying power.

[panoss]

I tried with 19V only (and gnd of course) with the laptop off and it 's not charging.
I powered it up and I get the message "The AC power adapter type cannot be determined" etc.
And in the windows tray a message "plugged in but not charging".
So I conclude that without the data pin it does not get charged at all.

[soldar]

You can try with a good charger but you might find out it does not charge either.

One of my Dell laptops was not charging and the charger was good so I thought it was the battery so I replaced it but it still does not charge so I have to use it plugged  in.

[panoss]

Ok, I found a good charger at low price, it says 'Dell' on it.
I don't know if it really is Dell but charges just fine!

I installed Win 8.1 but the behavior is the same. Only difference is that now the screen goes black.

Next I will check what wraper said about the Intel Management Engine Interface driver.

I checked here and I found no ' Intel Management Engine Interface driver' for Vostro 3350...

[soldar]

Ok, I found a good charger at low price, it says 'Dell' on it.
I don't know if it really is Dell but charges just fine!

Good. I am glad to hear that. What power is it? Is it the 65 or the 90 W? (Or other)

[wraper]

I checked here and I found no ' Intel Management Engine Interface driver' for Vostro 3350...

[amyk]

I highly doubt this is a software problem, since Linux isn't working either and the method to shutdown via APM/ACPI has been standard for decades. You could boot a pure DOS and try the many APM/ACPI shutdown programs available --- if they have the same result, then it's definitely not software.

[soldar]

I remember back in the paleolithic ages of Win98 I had a couple computers that would just not shut down and it turned out it was a device, modem I think, that refused to play nice and Win98 just waited forever. Get rid of the device and the problem went away.

I have not seen this problem since but I mention it just in case it could lead somewhere.

[panoss]

Good. I am glad to hear that. What power is it? Is it the 65 or the 90 W? (Or other)

It 's the 65W.

[soldar]

It 's the 65W.

Does it charge the battery while the computer is in use?

[panoss]

Yes it does.

[soldar]

That's good. Many people complain their Dell laptop came with a 65W PSU that would only charge the battery if the computer was not running and they resolved the problem by getting the 90W PSU. I guess your model does not have that problem.

[panoss]

I disassembled it completely.
But it's very difficult to spot the parts on the motherboard, I'm looking for hours now.
And in the schematic there is no pcb with the parts on it.
Is there a Phoneboard file for this board?

[panoss]

I removed the BIOS 's battery and did a BIOS reset (short circuited BIOS 's plus and minus for 10 seconds).
I reassembled the laptop and...everything 's fine now!
It shuts down perfectly!
I have no idea if the BIOS 's reset was the reason for this or something else.

[LateLesley]

The BIOS reset probably has had something to do with it. It'll have cleared the BIOS settings, and reset everything to default settings. I'm glad it's fixed it.