Post by DanPost by PaulPost by DanI have tried directly using the HDMI cable directly to the
motherboard. Same result
I have tried pressing F8, no result
I think card is a PCIe Rev2 video card
BIOS screen does not always come up as once power button is pressed,
the screen is empty as I have switched it on before power up
Re: "https://www.smps.us/atx-connector-20-24pin.jpeg"
I will ask him to try to try: does it
boot when you short PS-ON to ground? It's the green wire in the
20-wire
power connector/header on the mobo to which you connect the PSU
If you have "fans spinning", then there is no need to access PS-ON# signal
on the main power connector.
Probing "PowerGood" (an indication from the power supply, that
power is available) would not be complete in any case. There
are additional instances of PowerGood on the motherboard, which
must be ANDed together to tell the motherboard that "all is well".
Since you have "one SPKR beep", the BIOS is running, power must have
been good. You have indirect proof that Power was Good.
The BIOS making a short (controlled-length) beep tells you a lot.
But, the symptoms are still mysterious, as the evidence is,
some part of the video output is not working. It could be a
loose connector (not fully seated). Remember I told you they
did a poor job on allowing the connector to seat properly.
That drives me nuts here (I occasionally see a black screen
after running for a while, and it is because I bumped the HDMI
connector).
Post by DanRe: "Tried resetting BIOS settings by shorting the 2-pin mobo header?"
I wil ask to try this.
Remember not to do that for too long. That draws 3 milliamps from the
battery. There is a 1000 ohm resistor from the top of the battery, to
the PCH. And when you CLRTC, that connects the far end of the 1000 ohm
resistor to ground. The battery has limited capacity, and that is how
they decided to select 3 milliamps as the number. You SHOULD NOT
short the two pin thing, while the PC has +5VSB operating, as
that has higher current sourcing capability. Power should be
completely off to the PC, if you intend playing with that.
There are different ways to design CLRTC. Intel even provided a
reset signal, as an alternative to the industry standard method.
But nobody uses it. The instructions on CLRTC in the manual
have many times been wrong. I have received motherboard boxes,
with a separate sheet of typing paper, with corrected CLRTC
instructions which replaces the page in the manual (which is wrong).
Post by Dan"was an adequate amount of thermal paste applied to the CPU?"
It was the stock Intel cooler suppiled with the CPU. When it does get
to the bIOS, CPU temps are 29-30 Celcius.
I checked th clips underneath the motherboars, all the "legs" of stock
coller weere the same. The press down areas of the cooler legs were
all not moving andfirmly affixed to the motherboard.
I used smaller screw-heads than the area of the motherboard where the
screws are meant to be touching
No bad or burnt smell we could smell
Tried initially built in graphics gpu and then Asus - both same result
"GPU inside the 12400 will give you a working screen" sadly not and
only in 1 of 50 boots
https://www.asus.com/motherboards-components/motherboards/prime/prime-b760-plus-d4/helpdesk_qvl_cpu?model2Name=PRIME-B760-PLUS-D4
Core i5 i5-12400 2.5GHz 65W 18MB 6 H-0 all <=== All BIOS versions support this chip
(14xxx chip needs BIOS version check)
The metal standoffs on the motherboard, touch the bottom of the motherboard.
You check the bottom of the motherboard, before assembly, for "warning stickers".
The plated holes top and bottom, provided for mounting, are at
electrical GND potential on purpose. The motherboard standoffs are
meant to touch the metal ring. The screw heads are designed to
contact the plated area as well. This is for emissions, although
I have my doubts that makes any difference, and it is other aspects
of motherboard design that reduce emissions (buried clock signals
being an example of a way to make a quieter board). Many of the
clocks have "spread spectrum" as a means of cheating the lab
test equipment from an honest appraisal of the electrical noise :-)
This broadens a noise peak, and reduces the amplitude.
Your setup has lots of positive indicators (it is not broken).
Very little of it, is signaling something is wrong.
Without a video card in it, you could likely run the motherboard
on top of a cardboard telephone book or other insulator. Nothing
is needed to keep the faceplates upright, when there is no video card
plugged in and you are using HDMI/DP in the I/O plate area.
If you are using a 65W CPU and the Intel in-box cooler, then you
can almost pull the motherboard out of the case (without using
a removable tray casing). I have a couple builds here, where the
heatsink is too big, and must be removed to get the thing out
of the case. But a simplified build can sometimes be light enough,
you can get it in and out of the case without too much sweat.
*******
This particular Asus motherboard has a VGA connector. Since those
stand out a different amount than the HDMI and DP, you may have
better luck using the VGA to the monitor, instead. All my monitors
have VGA. VGA has saved my ass multiple times. It even works
good at 1920x1080, something I was not expecting to happen.
Paul
So, could it be possible that the motherboard is touching the case
somehow?
3 at the top and 3 at the bottom. But I had to use a chrome plated
screw from his spares.
I only initially tried the internal GPU then his old 550ti.
Same result apart the fanw were spinning normally on use of the 550ti
but - results were the same.
Nothing on the monitor.
How many storage devices are connected ?
If there is nothing to boot from, the screen should light up and tell you that.
The BIOS failing to boot, there's no beep for that, but text should be on the screen.
The only codes left to the BIOS after that, is the donkey siren noise of an
out-of-spec VCore or if you are lucky, some beeps on temp before THERMTRIP drops
the machine.
If the BIOS can run, pressing <del> early on (Asus or MSI mobo), should cause it to enter the BIOS.
And not to boot.
If you run out of ideas:
1) Turn off all power (use switch on PSU, or unplug).
2) Remove all RAM sticks, and put into antistatic bag.
3) Verify that BIOS makes three-beeps sound. This is a double
check that the BIOS is running properly. The BIOS can't do that,
unless it verifies the checksum of the main BIOS module. The BIOS boot segment
is much smaller... and dumber. The BIOS runs register level code, discovers
nothing on the SPD bus, and sounds three-beeps for missing RAM.
Flashing the BIOS on this board, only seems to be possible via EZ-Flash 3.
And the problem with that, is getting the suite to install on Windows 11.
I can't get the suite to install here on an Asus. Silent failure. Install stops.
And that's on a system which is running Windows and is fine. I'm trying to
get to the AISuite 3 hardware monitor, for temperature measurements.
The lack of features on this 760 mobo, leaves a bad taste. (The test machine, with
the 4930K in it, there is a standalone chip that flashes the BIOS, a microcontroller,
and you don't even need a CPU in the socket to use it. That's the "premium" method.
Costs them a buck to do that. The next level after that, is just a push button
on the I/O plate and all HW installed, and that's not as good. This board has... nothing
at all for manual flashing, according to docs. On the other hand, the initial BIOS
runs all 12th and 13th gen CPUs. Combining a 14th gen CPU with this board, would
be a mistake, without a shop to help you fix the BIOS.)
Note that, a suggestion to flash this BIOS, is for an updated microcode.
Microcode is updated at two points in time: In the past, the BIOS microcode
update, was purely to solve boot issues with disks, leaving it to the OS microcode
to apply a final version (for the session). Microcode is RAM inside the processor
and contents are lost at end of session. Microcode is protected by crude crypto,
and the CPU will not accept an offered microcode, unless the checksum (or something
a bit stronger), passes. Your 12th gen, might not need microcode, and perhaps
the file offered today, won't help your board in the slightest. Only 13th and 14th gen
need eTVB. Intel will not modify the microcode on chips that don't need it.
T=0 Processor runs initial microcode (eTVB out of spec)
BIOS BIOS applies microcode. Version 1661 BIOS applies latest microcode (eTVB patch).
OS Windows will pass along the latest BIOS, but this is subject to
Microsoft testing, before it will necessarily be offered. The "revision"
of the CPU listed, is a measure of the microcode. A given level of
function, is only able to update the microcode, if the new version is
higher than the old version. The CPU should (eventually) be protected
from eTVB, just via a late microcode updater.
Clearing CMOS, must be done with all power off, wait at least 60 seconds
for +5VSB to drain. Touching the two contacts together (jumper plug,
screwdriver tip), should cause VCC3 to drop for a fraction of a
second, clearing 256 byte CMOS RAM in PCH. The BIOS should do "Load Setup Defaults"
upon being powered up and resetting itself via the reset pulse. It does this
because the CMOS RAM checksum is wrong, and it sanitizes the contents via
a reload of defaults. On a UEFI BIOS, there might not be much in the 256 byte area,
as the "NVRAM emulation" in the BIOS flash chip, stores boot paths and such,
DMI and hardware info.
If you have modified the BIOS and used any of the dangerous or unreasonable
settings, then clearing CMOS is an option. Some BIOS will also reset the
settings, upon three power fails in a row. The BIOS clears an indicator, if
it "came up", and the indicator does not get cleared if it crashed, and three
of those power fails sequentially, may also reset the BIOS.
DO NOT clear the BIOS without turning off power and waiting 60 seconds or more.
Doing so will burn the ORing diode. Shorting the +5VSB path to VCC3, draws
excess current through the BAT45 (three-legged Schottky diode). +5VSb must be
off, before fiddling with CLRTC legs!
Your power supply could be the root of the problem, and checking the voltages
might be an idea to follow up on. If you can find a SATA Y-cable, cut the head
off one of the two output-side connectors, the five wires there give you
3.3V, 5V, 12V for monitoring, plus two black ground wires. If you connect
the "3.3V GND 5V GND 12V" from the PSU, to a 22TB drive, the drive won't
spin up. Some extender cables have only four wires "[NC] GND 5V GND 12V"
and if the cable you snipped only had the four wires, well, you can't
monitor 3.3V that way. But the SATA connector would also operate a 22TB drive
if the 3.3V wire was missing (on an extender power cable you used on purpose). .
Voltages:
12V (Vcore power, disk drive power, PCIe slot)
5V (Disk drive logic board, some NAND-based storage)
3.3V (Lots of stuff. PCIe slot)
[ -5V (has not been in a PC for decades... Pin could be missing.) ]
-12V (0.25 amps, used for RS232 buffer chip, RS232 sometimes quietly provided on an Asus mobo)
+5VSB (supervisor -- provided with fans off, runs USB ports, PS/2 ports, 3 amps max.
Powers the front-of-case power button to work. Shorting this out
makes a motherboard stop-and-drop.)
My older supply is running 11.5V on the 12V, and it is still running a PC
as we speak. High capacity HDD do not like this. Older HDD are good to
as low as 11V, before the drive will spin down. Modern drives over about 8TB,
may spin down at 11.5V, a tighter spec. Determined empirically. Below 11V,
I would hope Power_Good on the PSU, is deasserted, and the mobo won't start.
You would not get the single-beep if this was the case.
Otherwise, modern supplies are double-forward-conversion, the first stage
makes +12V only, this feeds a smaller board with the 3.3V and 5V regulator
(buck converter or similar). There could be a bit of cross-loading between
3.3V and 5V, but not nearly as much as in the past when conversion was
a single step. Power regulation is very good on modern PSU.
+5VSB is a separate SMPS. -12V implementation is unknown,
could be on 3.3V/5V board, or, just about anywhere in there.
The BIOS power load is the "one-core-max" number. On my other
machine, all cores is 224 watts. One core railed, is 115W or so,
roughly. The power supply on that system then, must be able to
source 115W, much of it on +12V, but certainly a significant
amount on 3.3V and 5V (mobo loads, SOC supply, PCH can be 5 amps,
and so on). The video card does not go to 3D-max while sitting
in BIOS. My fan-wont-spin video cards, don't spin the fan in the BIOS.
You're approaching the "find intelligent life" part of a build :-)
While you can take the PCB out of the case, I don't really
think a miracle is going to happen. My PSU is slightly out of spec,
but a couple of boards have run off it, no problem. Just a high
capacity drive, warned me the +12V was too low. That's when I checked it.
The grounding status of the standoffs, should not affect board function.
There are all sorts of sneaky ways they get the ground in there, such
as a couple "fingers" on the IO plate that touch the top of an I/O stack.
Paul
