MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

[Shotgun25]

Hi,

MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview is needed to check what voltage I should have at the bios chip and the PWR SW pin(s) since this board doesn't turn on even when all peripherals and RAM are removed and the PWR SW is shorted with a screwdriver.

I get 2.42V on pin 8 and I thought it should be 3-3.3V so wanted to confirm.

Any schematic or boardview would be appreciated, or even guidance of common parts to check on MSI boards of this era.

This is an AMD base board for reference.

Regards,

MATT (Shotgun25)

[piernov]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

BIOS will be read only after CPU VCore comes up and CPU reset signal is de-asserted, so it doesn't really matter for now. Check directly the voltage on the power button pins.
2.42V on a 3.3VSB line usually means that either the 5VSB line is low or the 3.3VSB line regulated from the 5VSB line is low.
For the 5VSB line, it can be an ATX PSU issue, almost always a bad cap on the 5VSB output that's near the heatsink or for some reason the motherboard is pulling low the 5VSB line (well, never happens except if the 3VSB regulator is shorted).
For the 3VSB line, either the regulator is bad (usually a AMS1117/LM1117 or similar) or something is pulling it low such as a bad Ethernet controller (*very* common) or a bad southbridge.
Inject 3 volts in the 3VSB line, monitor the current drawn and check the board for parts that get hot.

[Shotgun25]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

Hi Piernov,

Sorry for the delayed response, I wasn't subscribed to my own post for some reason.

I have 5.10v at PWR SW+ pin, and I have tried a known good working PSU from my machine.

I have had success booting the machine by using an external bench PSU to provide 3v to an area between the PCI slots where that voltage is measured on a 3pin n channel mosfet.

I can get it to boot if I supply that voltage and short the PWR SW pins. If I remove mains power from the PSU, or I shut down and turn off the isolation switch on the PSU, then it is intermittent whether it starts (as though some capacitor retains enough charge to keep voltage high enough somewhere).

I have replaced that mosfet but it is still the same situation.
Mosfet (on left of attachment yellow line) is APM2054, the chip to the right is uP0104S.

Additionally the voltage doesn't change even with a new CMOS battery, so it is definitely PSU supplied voltage, so as you mention "the 3.3VSB line regulated from the 5VSB line is low".

Any ideas where to check next?

I think the uP0104S chip where the low voltage goes to and diode mode traces to might be where the power is generated and not well enough?
What do you think?

[piernov]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

No schematics or boardview and I don't have the board so it's hard to make a guess…

The diode probably switches between coin cell battery and +3VSB for the RTC well.
uP0104S seems to be what creates the voltage indeed. MS-7699 (schematics available) has this part too for this purpose.
Check for a short to ground with your multimeter, then check with your bench power supply if the current draw is not excessive. If it's all good, uP0104S is probably the culprit and should be replaced.

[Shotgun25]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

I found a shorted 2R2 resistor near the atx connector, but the one I replaced it with was smaller and it smoked straight away.
I have tested with two axial resistors which gives 2.2ohm and it along with a short on the drain+source of the P06P03A allows start up with the power button everytime...but the resistors are scalding to the touch.

I think I got a reading between 0.1A and 0.4A on the 1.5v input to that resistor and someone advises me the resistor is likely only for current sensing so shouldn't have a reading that high. I'm going to check that again tomorrow.

The gate of the P06P03A is the same voltage that is seen down near the BIOS battery so I think that area was a receiver of bad power so not the right area to focus.

[Shotgun25]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

Thanks for persisting with this and for your time.

Here are the standby voltages and forced on operation voltages on the uP0104s chip.
STBY PWRON
Pin 1 0.2v 0.24
Pin 2 2.41v 3.21v
Pin 3 2.41v 3.24v
Pin 4 3v 5v
Pin 5 0v 0v
Pin 6 1.1v 1.1v
Pin 7 0.79v 0.79v
Pin 8 0v 0v

Where do you propose I supply power to to check current draw?
Also where are you expecting to see a short to ground? a cap? a specific leg of one of those two ICs?

When supplying the tab of the 3-pin APM2054 IC with 3v it would initially draw around 0.3A, then drop to around 0.1A approximately I guess due to charging a nearby capacitor.

[piernov]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

2.2ohm resistor is probably not for current *sensing* but rather for current *limiting* (or protection). Can be put before the VCC pin of an IC for example. Also it wasn't shorted but open.

The uP0104s is in fact probably not the 5VSB to 3VSB chip. Its pin 3 (VIN) is connected to the diode and is receiving 3.3V. Its pin 6 (VOUT) is sending out 1.1V.
It's most likely unrelated to the problem at hand.

If where you supply the current is indeed the 3VSB line, 0.1A is ok, the 3VSB line should be designed for at least 0.5A. Check with the board unplugged (and your PSU supplying the 3VSB line) to make sure it's not the existing rail supplying most of the current.

It still sounds like there is an issue with the 5VSB to 3VSB converter, but apparently it's not the uP0104S you measured. Try to find another uP0104S or a similar chip on the board, or a 1117 one, that gets around 5V on a pin and output the (bad) 2.4V you're measuring on another when the board is off.

[Shotgun25]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

Thanks for the guidance piernov, I just bridged pins of the mosfet at the bottom right in the attached picture and I replaced the failed 2.2ohm resistor with 2x1.x ohm 1/2 watt resistors in series. The resistors get scalding hot but haven't burned after a few hours of being powered on with the machine doing some youtube and some windows updates.

The resistor was for the 1.5v section for the RAM by the sounds of it. I'm wondering if it is provided as a part of a protection/limiting circuit for overclocking purposes?

Anyway I plan to give this back to my brother at some point this week and we will see how long it lasts.

You don't see those resistors causing a fire or killing the RAM do you?

[Shotgun25]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

Hi Piernov,

Do you have an idea where I could buy that SMD resistor, but more importantly could you hazard a rough guess at the specification of that SMD resistor beyond the 2.2ohms?

The two resistors are still working ok, but I haven't returned my brother's machine to him yet as he wanted some software stuff doing too and haven't had time.

Thanks,

Matt (Shotgun25)

[piernov]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

Measure its size with calipers, to know the SMD package size. Refer to the following table https://en.wikipedia.org/wiki/Surfac...ology#Packages
Basically the most important characteristics are the value (2.2 ohms) and the size (for example 0805 or something like that). Then it may have a tolerance (for example 5% meaning the value of the resistor is within 5% of 2.2 ohms), sometimes it matters on resistor divider than must have a very precise output. There's also power rating (which usually depends on the package size) and temperature rating but this shouldn't matter.

[aquafinatop]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

I can help you but I still don't understand what you mean "I get 2.42V on pin 8" pin 8 of what? take a picture for me

[Shotgun25]

Re: MSI A68HM Grenade (MS-7891 ver:1.1) schematic/boardview needed

I fixed the 2.2ohm resistor with two film resistors in parallel.
Also I was referring to Pin 8 of the BIOS chip.
I fixed this last year and gave it back to my brother.
He hasn't complained of fire so far