Post by PaulGiven one to troubleshoot: it has no beep codes. Yep, neither pins for a speaker on one on the mainboard.
Eff you Asrock, glad I didn't pay for this piece of crap.
The HDAudio chip has "BEEP_IN" pin and a block labeled
"BEEP-Gen". It's a RealTek ALC892. Alternately, when
stuff like this is missing, a beep signal can be capacitlvely
coupled into L or R of the Line_Out.
Connect your stereo computer speakers to LineOut
and try your test again. Remove RAM, so the board can give the RAM beep.
Naturally, if the CPU is not coming out of RESET, there won't be a beep.
The beep relies on the BIOS loading, to work.
The board does have a SuperIO. I couldn't find it, using the board
picture on Newegg, but the SuperIO is hidden by the PCI Express plastic
heel lock in the lower right corner of the board.
Paul
I found 2 problems. Bad RAM, and it is very fussy. I had to try 3 DIMMs before one that it liked. Secondly, it is overheating. It seems the CPU socket is too low and the cooler doesn't make contact. I tried 2 Intel cooler, the same issue. So one needs a very thick dob of heat-sink paste. Even then it is 46 degrees at idle at 90 degrees with 1 core going 100%. So needs CPU fan on full speed in BIOS.
I wonder how that happened. How would the datum be too low ?
The CPU list seems to have desktop processors in it. And it is
LGA 1150 and does not have a soldered-in-place CPU. It's
a land grid array. It should be the right height for the
stock cooler.
https://www.asrock.com/MB/Intel/H87M-ITX/index.asp#CPU
If you de-lid a processor, that makes it too low. That I
could understand.
This article doesn't cover the options well. One kind of
shim was a "bumper" that prevented rocking from side to
side on AthlonXP. That prevents die cracking. In your
case, you need some sort of shim to adjust the height
(on top of the IHS integrated heat spreader). That really
should not be necessary.
.
If the CPU had been lapped with grinding powder, the
legend (CPU part number) would be missing.
https://en.wikipedia.org/wiki/CPU_shim
*******
I was flipping though the manual, and it has an enable
for BIOS beep. And it says ("needs buzzer") when I cannot
see a buzzer on board, neither do I see SPKR header for the
function. I don't think the tech writer had a sample of the
board sitting in front of him.
*******
If you had XMP RAM, that's the fastest way to get a motherboard
to adjust itself. That's because the XMP table has the DRAM
voltage setting in it. But at that speed on DDR3, perhaps only
a DDR3-1600 would have XMP. Either DDR3-1066 or DDR3-1333 would
be the default in that case, and XMP makes it run faster. but
the other purpose of XMP, is to achieve stability, when a
motherboard BIOS was not designed well. Sometimes the XMP
setting is more stable than the rest of the code.
When I put some DDR3-2400 XMP capable in my Optiplex 780,
I had to leave XMP off, because... it doesn't have XMP.
And it did OK by selecting DDR3-1066 and ran fine. On machines
like that, you don't have any choice in the matter. The machine
picks 1066 and that is it.
There is nothing at all in the BIOS release notes, hinting
at RAM problems.
Whether you have DDR3 or DDR3L, both should run off the
same voltage, so that is not it. You can design a DDR3 board
for regular DDR3, and DDR3L is backward compatible and will
run in a DDR3 board. If a board was specifically designed
for only DDR3L, then the voltage might be too low for
stable "regular DDR3" operation. Check the voltage setting.
Sometimes the memory controller has a voltage
that needs a slight bump from default (System Agent Voltage Offset?
VCCSA maybe). Other than the basic DRAM voltage, that's about the
only kind of voltage I adjust here. I don't adjust terminator voltages
or thresholds or anything of that sort.
*******
I think a little tuning, and you will get this running right.
After all, it's not an NVidia chipset :-)
Some of those can be rat bastards, with respect to RAM.
Almost impossible to live with.
This is an Intel. And it should listen to reason.
Paul