New build won't power up

[danik56]

GA-B85M-HD3 with I3 4150 and ANTEC NEO ECO 450W psu.
when pressing the power button, the fans spin for 5 seconds and stop and then start again for 5 seconds and stop and this keeps repeating.
any idea what might cause this kind of behavior ?

 

[nickkarma]

Sort of a shot in the dark as I remember something similar happening to me as more of an oh duh moment on my first build. Did you by chance forget one of the power connectors to your motherboard? When it happened to me I forgot to plug one of the two in.

 

[DSF]

Any chance you have other parts to test with?

As nickkarma said, the first step is just to make sure all the connections are correct. Also, did you install the standoffs in your case before you installed the motherboard? (Or did you make sure they were there if your case comes with them preinstalled?)

Failing either of those options there are a lot of things you could try.

First, strip it down to the bare bones (CPU, one stick of RAM, motherboard and PSU). Don't even insert the graphics card if you're using one. Set the motherboard on the cardboard box it came in rather than seating it in the case and try to power it on. If all of the connections are correct and it still won't boot with just the bare essentials then you're mostly likely at the point of a dead part. If that's what happens then you'll need to swap things out until you find which part was defective.

On the other hand, if the computer is able to boot when it's resting on the cardboard then there's some issue with how you had it installed in the case. (Edit: Or it's an issue with one of the parts you added later like the hard drive or video card. It's unusual, but I have seen odd things like a hard drive causing a complete failure to POST.)

 

[danik56]

Let me answer few of the questions above.
I have already disconnected everything but the power cables RAM and CPU. the MB still in the case and it is sitting on 6 standoffs (5 built in and one I screwed in)
Next I will try to take the MB out and power it up outside the case. How does it power up if not connected to a power switch ?
I will also try to get another PSU and RAM (currently I have only a single memory stick)

 

[DSF]

Let me answer few of the questions above.
I have already disconnected everything but the power cables RAM and CPU. the MB still in the case and it is sitting on 6 standoffs (5 built in and one I screwed in)
Next I will try to take the MB out and power it up outside the case. How does it power up if not connected to a power switch ?
I will also try to get another PSU and RAM (currently I have only a single memory stick)

If the wires for the power button connection are long enough you could just connect it anyway even though the motherboard is not sitting inside the case.

If they won't reach you can just short the two PWR pins together for a moment with the tip of a flathead screwdriver. In reality that's all the computer's power button does - connect both leads with a metal plate for a moment.

 

[danik56]

Tried to power up outside the case using a small screwdriver to short the power pins.
same result. when I put my ear next to the tiny speaker I can hear 2 short ticks followed by another short tick and then the fan stops. this sequence keeps repeating until I turn off the PSU.
tried with different ram stick and also without CPU. same result as before.
this tells me the MB does not recognize the CPU so either the CPU is bad or the socket is bad.
I think I can see one pin in the socket that is not in place. it is hard to tell but in certain angel the light reflects off that pin differently.
You think a bent pin would cause this kind of behavior ?

 

[ViRGE]

You think a bent pin would cause this kind of behavior ?

LGA makes it hard to tell if a pin is bent or not unless it's badly bent, so it's entirely possible it's just the light playing tricks on you. But yes, if it really is bent then depending on what pin it is, a lack of contact could botch the entire operation.

 

[Termie]

You bought a new 4150 CPU that isn't guaranteed to work with anything but H97 and Z97 boards. That's probably your problem.

I'd return it and get a 4130.

 

[DSF]

You bought a new 4150 CPU that isn't guaranteed to work with anything but H97 and Z97 boards. That's probably your problem.

I'd return it and get a 4130.

Oh yeah, good catch.

 

[danik56]

The CPU support list for my board says 4150 is supported since BIOS version F9.
question is if I have the BIOS version out of the box or not.

 

[ViRGE]

The CPU support list for my board says 4150 is supported since BIOS version F9.
question is if I have the BIOS version out of the box or not.

Without plugging in an older CPU, there's no practical way to tell unless the board maker explicitly labels a box/board (as has sometimes been done in the past in these situations). A B85 board could have been sitting on the shelves for months.

 

[danik56]

Without plugging in an older CPU, there's no practical way to tell unless the board maker explicitly labels a box/board (as has sometimes been done in the past in these situations). A B85 board could have been sitting on the shelves for months.

I agree the board (revision 1.1) could have been sitting in the store.
What CPU model can I borrow which is guaranteed to work with the B85 regardless of the BIOS version ?
Would a G3220 work ?

 

[ViRGE]

Would a G3220 work ?

Likely. It was introduced fairly early into the Haswell lifecycle. Though Gigabyte's page isn't of any help here since it lists F9 for everything.

 

[danik56]

Is there any other reliable source for determining if a CPU is going to work with a given MB regardless of the BIOS version it comes out of the box with ?

 

[danik56]

Since B85M-HD3 is intended for Haswell cpu's, I assume there must be one that is supposed to work out of the box, right ?

 

[ViRGE]

Since B85M-HD3 is intended for Haswell cpu's, I assume there must be one that is supposed to work out of the box, right ?

All of the launch CPUs are sure to work. Though the G3220 was released only a quarter later, so it's not particularly newer.

 

[greenhawk]

One other idea is to check that the CPU fan is plugged into the fan header for the cpu and not one of the other fan connections. Though if this was the case, it should beep at you about a faulty fan.

 

[danik56]

CPU fan was plugged in correctly. I made sure of that.
I'll try to get another CPU and see if it makes a difference.

 

[danik56]

I managed to get a G3220 on loan to test my board and possibly upgrade the BIOS.
Will post results later tonight.

 

[danik56]

Good news.
With G3220 I can get the MB to post.
The Bios version is F5 and should be upgraded to F9.
Now I need to figure out how to flash a new Bios via a USB storage device.

 

[ViRGE]

Good news.
With G3220 I can get the MB to post.
The Bios version is F5 and should be upgraded to F9.
Now I need to figure out how to flash a new Bios via a USB storage device.

According to your mobo's manual it has Q-Flash support. So drop the BIOS on a FAT32 USB drive and you should be golden.

And a gold star for Termie for properly identifying this issue.

 

[Termie]

Good news.
With G3220 I can get the MB to post.
The Bios version is F5 and should be upgraded to F9.
Now I need to figure out how to flash a new Bios via a USB storage device.

Glad this worked. Awesome that you were able to get a loaner G3220 so quickly.

Intel should SERIOUSLY be more careful about warning people re: incompatibilities with its new chips. I can't imagine the number of RMAs that are going to result from this launch.

 

[danik56]

Thanks a lot to all of you here for helping to resolve this issue.

 

[ViRGE]

Intel should SERIOUSLY be more careful about warning people re: incompatibilities with its new chips. I can't imagine the number of RMAs that are going to result from this launch.

We really need a better way of doing this altogether. Either mobos need to be able to flash the BIOS without a CPU (go put a small Quark core in the PCH) or they need to be able to boot up new CPUs in some kind of failsafe mode.

 

[mfenn]

We really need a better way of doing this altogether. Either mobos need to be able to flash the BIOS without a CPU (go put a small Quark core in the PCH) or they need to be able to boot up new CPUs in some kind of failsafe mode.

Wow, using a Quark chip is a really good idea! Server boards have an ARM chip on the motherboard that drives the lights-out management which can help with this, but consumer boards don't have anything like that. A Quark chip is a great observation since the transistor budget should be minimal. All it really needs to do is be able to power the UEFI, so that's not a lot of power.