Booting from USB3.0 PCI-E card?

[bigbeark]

I have an ASUS M2N-E motherboard which seems to have no working USB ports. I don't know if the USB chip(?) is dead or if it's a BIOS problem.

Of course, these boards update the BIOS using a USB Thumbdrive.

So the question is do I buy a USB3.0 PCI-E card that installs into a 1x or 4x PCI-E slot and costs $15 or do I buy a new BIOS chip which is about $20.

If the USB3.0 card does not boot the board I can use it on another board. If I buy the BIOS chip and this doesn't fix the USB functioning, it's a loss.

Has anyone successfully booted from a USB PCI-E card?

[ChaosLegionnaire]

Re: Booting from USB3.0 PCI-E card?

nope. this being a nvidia shitset bumpgate board, its probably the mcp sb that is fucked. thats why the usb ports are no worky. u can ask momaka on how to do a ufc mma punch on the mcp to temporarily fix it and get the usb ports useable albeit temporarily. he has done this kind of ufc punch fixes before on these dying nvidia shitset boards.

or try reflowing the mcp sb. see if that fixes the usb ports.

[stj]

Re: Booting from USB3.0 PCI-E card?

you put a spare harddrive on another machine and install freedos on it together with the bios and flasher - then plug it into the asus board.
it's not your bios though.

boot a Linux live-disk from dvd,
then try to scan or access the usb ports - linux does not use the bios.

[momaka]

Re: Booting from USB3.0 PCI-E card?

u can ask momaka on how to do a ufc mma punch on the mcp to temporarily fix it and get the usb ports useable albeit temporarily.

Lol.
My method has only been successful on this one ECS MCP61PM-AM motherboard (from a Gateway GT-5656 PC), so far.

USB still works after those punches and board seems to be stable overall.

That said, I also give that PC regular kicks too, now - not because there are problems with it, but because it's under my desk and I often conveniently use it as a foot rest. The nice curves on the GT-5656 case make for a great foot rest, actually.

On a related note, I do also have an Asus/Pegatron M2N-E (or similar M/N anyways) motherboard from an HP Pavilion PC that locks up in BIOS with any SATA devices attached to it. No SATA devices and I can boot. Very likely I suspect the issue is with the GeForce 6150 / nVidia MCP61 chipset there.

So I think Chaos may be onto something here - your board may have chipset issues. Try the live Linux boot DVD, as sjt suggested, to see if you can access the USB ports (though also check in CMOS that they aren't disabled.) If that doesn't work, then maybe try the USB 3.0 PCI-E card method (though keep in mind that a failed MCP61 chipset can also have issues over PCI and PCI-E slots), as this will at least allow you to reuse the card in another PC.

In any case, don't invest too much into that M2N-E motherboard, IMO. Unless you make modifications to the chipset cooling (at least add a 40 mm fan once you determine if/where the issue was and that it wasn't from the chipset itself), the board will likely die from chipset issues, eventually. There really should have been a class-action lawsuit against nVidia on these, because way too many had issues.

[bigbeark]

Re: Booting from USB3.0 PCI-E card?

Thanks everyone for your suggestions. I'm going to use an old 10GB IDE HDD and the Freedos method to update the BIOS.

So I just boot a Linux disc and do a LSUSB command to see if the device is found?

If LSUSB detects it then what?

If the USB3.0 card lets me use a USB keyboard and mouse, I will be satisfied.

I know VIA cards are no good for Linux (not recognized)

I have a heat gun and a handheld digital temperature reader. Any tips on a ghetto reflow?

Is this a soldering issue or the chipset itself degrades and loses functionality?

Thanks for your patience.

[stj]

Re: Booting from USB3.0 PCI-E card?

I know VIA cards are no good for Linux (not recognized)

i doubt that, Linux has newer drivers than windows.

[bigbeark]

Re: Booting from USB3.0 PCI-E card?

I made multiple mistakes trying to update the BIOS on the board with the dead USB ports. I burned the FREEDOS img file fo a USB stick, then copied this plus the other FREEDOS files plus AWDFLASH and the 5001.BIN Bios file to the IDE HDD. This would not boot. The USB ports had miraculously come back to life, so I tried the USB stick I was able to flash the BIOS successfully,restart and edit the BIOS via the DEL key, but when I restarted the board it would not boot the OS.

I removed the BIOS chip from the board and put it into my good board. Same results. So I tried the onboard EZYFLASH2 option. It wanted a CD, so I burned the 3001 bios version to a CD and was able to repair the BIOS.

[momaka]

Re: Booting from USB3.0 PCI-E card?

Well, glad to hear you got it resolved.

[stj]

Re: Booting from USB3.0 PCI-E card?

i personally wouldnt trust that chip too much - consider a new one, blanks are only a $ these days.

[televizora]

Re: Booting from USB3.0 PCI-E card?

No boot rom = no support. At least for the PCI/PCI-E storage controllers.
I have never investigated a boot from a PCI/PCIE USB card.
The bios must recognize the controller, so it can use it as a valid device from which it can boot.

[bigbeark]

Re: Booting from USB3.0 PCI-E card?

Lol.
My method has only been successful on this one ECS MCP61PM-AM motherboard (from a Gateway GT-5656 PC), so far.

USB still works after those punches and board seems to be stable overall.

That said, I also give that PC regular kicks too, now - not because there are problems with it, but because it's under my desk and I often conveniently use it as a foot rest. The nice curves on the GT-5656 case make for a great foot rest, actually.

On a related note, I do also have an Asus/Pegatron M2N-E (or similar M/N anyways) motherboard from an HP Pavilion PC that locks up in BIOS with any SATA devices attached to it. No SATA devices and I can boot. Very likely I suspect the issue is with the GeForce 6150 / nVidia MCP61 chipset there.

So I think Chaos may be onto something here - your board may have chipset issues. Try the live Linux boot DVD, as sjt suggested, to see if you can access the USB ports (though also check in CMOS that they aren't disabled.) If that doesn't work, then maybe try the USB 3.0 PCI-E card method (though keep in mind that a failed MCP61 chipset can also have issues over PCI and PCI-E slots), as this will at least allow you to reuse the card in another PC.

In any case, don't invest too much into that M2N-E motherboard, IMO. Unless you make modifications to the chipset cooling (at least add a 40 mm fan once you determine if/where the issue was and that it wasn't from the chipset itself), the board will likely die from chipset issues, eventually. There really should have been a class-action lawsuit against nVidia on these, because way too many had issues.

So I installed a Phenom X4 9500 quad core CPU which booted fine. The chipset runs hot though. That fancy copper heat pipe thing is pathetic. Worst of all the video card partially obscures the chipset heatsink so the largest fan that can be fitted is 25x25mm.

I've got a pair on order. I will use some hi-temp silicone to attach the fan to the heatsink. There are two fan headers right there near the chipset heatsink.

[televizora]

Re: Booting from USB3.0 PCI-E card?

I will use some hi-temp silicone

You mean thermal glue?

[Uranium-235]

Re: Booting from USB3.0 PCI-E card?

Lol.
My method has only been successful on this one ECS MCP61PM-AM motherboard (from a Gateway GT-5656 PC), so far.

On a related note, I do also have an Asus/Pegatron M2N-E (or similar M/N anyways) motherboard from an HP Pavilion PC that locks up in BIOS with any SATA devices attached to it. No SATA devices and I can boot. Very likely I suspect the issue is with the GeForce 6150 / nVidia MCP61 chipset there.


In any case, don't invest too much into that M2N-E motherboard, IMO. Unless you make modifications to the chipset cooling (at least add a 40 mm fan once you determine if/where the issue was and that it wasn't from the chipset itself), the board will likely die from chipset issues, eventually. There really should have been a class-action lawsuit against nVidia on these, because way too many had issues.

I had a A8N-E (nforce 4 Ultra). Fan on the ICH and it worked for several years but eventually totally died. Good temps or not, those nforce chipsets eventually die (bump). Nvidia did get sued, but getting any money from them depended on who you bought your laptop from (didn't include desktop MCP's)

https://www.itproportal.com/2010/10/...ails-revealed/

[televizora]

Re: Booting from USB3.0 PCI-E card?

Good temps or not, those nforce chipsets eventually die (bump).

Nvidia made AMD chipsets. I have always used boards with Intel chipsets and processors, because they have always been much more trouble free. You will extremely rarely see a defective Intel chipset. They know how to do their stuff.
In this case, it's not so much the Nvidia-s fault, but the tech itself.

[Uranium-235]

Re: Booting from USB3.0 PCI-E card?

Nvidia made AMD chipsets. I have always used boards with Intel chipsets and processors, because they have always been much more trouble free. You will extremely rarely see a defective Intel chipset. They know how to do their stuff.
In this case, it's not so much the Nvidia-s fault, but the tech itself.

yeah back in the day when we had dead dv6000's we would order replacement motherboards with intel chipsets. Never had a problem with them

It's pretty much irrelevant now. Nvidia/Via/Sis no longer bother making chipsets. For everyones good

[momaka]

Re: Booting from USB3.0 PCI-E card?

So I installed a Phenom X4 9500 quad core CPU which booted fine. The chipset runs hot though. That fancy copper heat pipe thing is pathetic. Worst of all the video card partially obscures the chipset heatsink so the largest fan that can be fitted is 25x25mm.

Oh, I see now. That M2N-E has one of those heatpipe'd chipset heatsinks, where the heatpipe runs to some fins back over where the CPU VRM is (and is also used to "cool" the CPU VRM. Sure....)

Yeah, most of those designs were gimmicks for gamers - look fancy with heatpipes, but don't cool worth a crap. I have several different mobos here like that - one of them being an EVGA (Intel) 780i-SLI. The chipset on that one will not run below 52-53C, even in the winter when it's nice and cool in my house. In the summer, it runs hot and it's a crash-fest. Also have an MSI P965 Platinum with similar gimmick chipset cooler. Luckily that one has Intel chipsets and just doesn't run as hot, so I'm not too worried about it.

Anyways, regarding the fans that you got... you probably could have put a 40 mm fan halfway on top of the chipset heatsink (with the other half sticking out) to avoid the clearance issues with the GPU. At least that's what I did with some of my M2N68/78 boards so I can still use a dual-slot GPU. Though I don't see any cooling fins on that chipset heatsink, so the fans might not really help it. If anything, try cranking up the CPU fan speed so that it pushes more air through those fins/heatsink over the VRM (where the heatpipe from the chipset goes to.) I think that might have a better chance of helping the chipset run cooler... if that heatpipe is actually properly soldered/attached to both heatsinks and if it's also a sintered-Cu heatpipe. The ones on my MSI P965 Platinum are not soldered but just epoxy-glued instead.

It also probably doesn't help that those 1st gen Phenom X4 CPUs are pretty much like small burners. So maybe a good idea to crank up the CPU fan speed anyways.

I had a A8N-E (nforce 4 Ultra). Fan on the ICH and it worked for several years but eventually totally died. Good temps or not, those nforce chipsets eventually die (bump).

If it was a stock heatsink... fanned or not, it will run hot. I find the fanned heatsinks aren't much better, simply because they are often just a flat piece of aluminum with the sides bent up slightly as "fins". Often the base is too thin so they don't work too well either. Cheap corner cutting at its finest again.

Funny enough, I just set up a system the other day with an nForce 4 Ultra chipset for testing purposes (on a Biostar NF4U AM2G motherboard.) Chipset/ICH heatsink on that one looks fairly well made actually and has a fan too. But the fan has a shot BB and is louder than all of the fans combined in the system. Other than that, it runs stable-enough so far.

Nvidia did get sued, but getting any money from them depended on who you bought your laptop from (didn't include desktop MCP's)

Well, that's the sad part - they got a free pass for the desktop chipsets, which were also affected.

Nvidia/Via/Sis no longer bother making chipsets. For everyones good

SiS had buggy drivers sometimes, but were overall OK.
Via, though... they actually made some pretty damn dependable chipsets if you couldn't say anything else about them.

nVidia always had a lot of good features, but just weren't reliable. On that note, some of the late era ALi (remember those) chipsets for AMD were also in the same boat.

On the other hand, ATI's early chipsets weren't too bad, probably up until and including RD480 (Xpress 200 and 1300/1600) and RS690_. After that it can be a bit arguable... but generally they did OK and better than nVidia at least.

I have always used boards with Intel chipsets and processors, because they have always been much more trouble free. You will extremely rarely see a defective Intel chipset. They know how to do their stuff.

I've been following the same logic when changing my "main PC". I don't mind AMD, but the AMD boards that come with reliable chipsets and have all the features I want were very few. So I too have usually stuck with Intel for systems that needed reliability. But that's all with the older hardware. Not sure how the new stuff fares these days. After all, the choice is pretty binary with modern stuff: AMD board with AMD chipset or Intel board with Intel chipset. While Intel has had good reliability in the past, I think they started slipping a bit around their 7th gen CPUs - at least that's when I started seeing a lot more bad CPUs (typically due to poor stock heatsink cooling, but not only.)

[televizora]

Re: Booting from USB3.0 PCI-E card?

While Intel has had good reliability in the past, I think they started slipping a bit around their 7th gen CPUs - at least that's when I started seeing a lot more bad CPUs (typically due to poor stock heatsink cooling, but not only.)

Never seen bad CPU. Wasn't the throttling mechanism of the CPU there exactly to prevent overheat damage?

[momaka]

Re: Booting from USB3.0 PCI-E card?

Never seen bad CPU. Wasn't the throttling mechanism of the CPU there exactly to prevent overheat damage?

It is. But if you repeatedly abuse it enough, the CPU will fail.

I was working at the repair shop of a big local computer store here a few years back, and basically once the 7th gen Intel i3/5/7/ CPUs rolled it, we started seeing CPU failures on almost a monthly basis - even in desktops! Before that, I too had never seen a bad CPU, save for AMD socket 462 CPUs murdered by people not installing the heatsink properly (i.e. either cracking the exposed core or not mounting it right) or trying the system out for a "test" with the CPU but no heatsink on it.

I can't remember if it was the 6th gen or 7th gen when Intel decided to go back to regular thermal paste between the CPU core and the heatspreader plate on top, instead of keeping it soldered like it did on the previous generations. So as a result, core temperatures could jump a lot higher and a lot quicker than before, which is what I think was part of the problem.

Anyways, I'll try to look it up later. I do remember there were even memes about this, though. So it was definitely becoming a known problem.

[Uranium-235]

Re: Booting from USB3.0 PCI-E card?

7th in general runs a lot hotter than 6th

In my goodwill computer works days I saw a LOT of bad Athlon/XP's. Even in OEM systems. The paste becomes hard or the heatsink got clogged up, or both. Eventually heating the killing the die over time

I did smoke a t-bird 1.33 of mine by accidentally putting the heatsink on wrong