Asus B250 mining motherboard component issue

[doctorman]

I have a Asus B250 mining expert mother board that does not post,
it turns on, CPU fan and LED is on, no buzzer sound, and reboots after 25 seconds or so.

The components have been tested in another of the same model motherboard.

no visible issue at all

cpu pins are good , capacitors look good

but when I was testing continuity of all capacitors on circuit using regular multimeter , I found one that has continuity while all other did not,

so I checked the same capacitor on a working motherboard and see that it is not supposed to have continuity.



so I decided to take it out and see if the motherboard runs,
unfortunately it did not fix it with the capacitor removed...

Does that mean I have another issue beside the capacitor?
I was going to order

  Wurth Elektronik 865250157010  
 mouser.com/ProductDetail/710-865250157010

what are the chances that replacing that capacitor will fix it...

blue dot close to the place capacitor was removed

[doctorman]

Re: Asus B250 mining motherboard component issue

checked the continuity on the 24pin ATX port to the ground
compared it to a working motherboard of same model

following pins are abnormal on the 24 pin shorted or cut out

Pin 9 +5v is shorted and gets 12ohm
Pin 14 (-12) gets 10ohm but should be no connection
Pin 15 G gets 10ohm but should be no connection
Pin 16 PS-on gets 10ohm but should be no connection

there is certainly a short somewhere but it is impossible for me to figure it out specially without the schematics

anyone can help?

[momaka]

Re: Asus B250 mining motherboard component issue

checked the continuity on the 24pin ATX port to the ground
compared it to a working motherboard of same model

following pins are abnormal on the 24 pin shorted or cut out

Pin 9 +5v is shorted and gets 12ohm
Pin 14 (-12) gets 10ohm but should be no connection
Pin 15 G gets 10ohm but should be no connection
Pin 16 PS-on gets 10ohm but should be no connection

there is certainly a short somewhere but it is impossible for me to figure it out specially without the schematics

Were these tests done with the PSU disconnected?
Also, do a resistance test on lowest resistance scale (typically 200 Ohms for manual multimeter - for Auto-ranging, don't worry about this) and NOT continuity (which is often tied to diode check on cheaper multimeters and can cause some confusion when measuring.)

The "shorted" capacitor you removed was probably not shorted.

Pin 9 is the 5VSB (5V standby) and not regular 5V rail. In any case, 12 Ohms is way too low... but again, make sure you do this measurement with all PSU cables disconnected from the motherboard.
The -12V rail on pin 14 should indeed be open / high-resistance. Since nothing recent uses the -12V rail (or at least that I know of), a short-circuit on that just doesn't make sense. So please confirm how the measurements were done... or preferably do them as I suggested above and report back. If the -12V is indeed shorted like that... well, that would be the first one I see like this.

On a related note, are you testing the motherboard outside of the PC case and on an insulated surface? Sorry if these seem like "basic" questions. I'm not trying to assume your troubleshooting background, so that's why I am being as explicit as possible (hope you didn't mind that. )

[doctorman]

Re: Asus B250 mining motherboard component issue

Were these tests done with the PSU disconnected?
Also, do a resistance test on lowest resistance scale (typically 200 Ohms for manual multimeter - for Auto-ranging, don't worry about this) and NOT continuity (which is often tied to diode check on cheaper multimeters and can cause some confusion when measuring.)

The "shorted" capacitor you removed was probably not shorted.

Pin 9 is the 5VSB (5V standby) and not regular 5V rail. In any case, 12 Ohms is way too low... but again, make sure you do this measurement with all PSU cables disconnected from the motherboard.
The -12V rail on pin 14 should indeed be open / high-resistance. Since nothing recent uses the -12V rail (or at least that I know of), a short-circuit on that just doesn't make sense. So please confirm how the measurements were done... or preferably do them as I suggested above and report back. If the -12V is indeed shorted like that... well, that would be the first one I see like this.

On a related note, are you testing the motherboard outside of the PC case and on an insulated surface? Sorry if these seem like "basic" questions. I'm not trying to assume your troubleshooting background, so that's why I am being as explicit as possible (hope you didn't mind that. )

PSU disconnected,
Cmos battery in place
the motherboard is on insulate surface out of the case

I used craftsman No 130810958 multimeter , it has
Auto Ohm, Diod v , Continuity with beep Ohm, Auto Cap nF
modes

I used Continuity with beep for testing.

Pin 14 (-12v) beeps on continuity shown 11ohm
on Auto Ohm mode shows OL
1 probe is on USB case Ground, 1 probe in pin 14.

Pin 9 is the 5VSB I get 11ohm on Auto ohm mode.

I bought a LPC motherboard Debug board that I will try see if I get a code.
Yeah figured it was not the Capacitor so I put it back

for testing there is no CPU or Ram in the Mobo.

[momaka]

Re: Asus B250 mining motherboard component issue

PSU disconnected,
Cmos battery in place
the motherboard is on insulate surface out of the case

OK, that's good.
Try without the CMOS battery installed, though, just in case. I doubt the results would change, but you don't need it at this point... and that way, every time your remove power from the system, mobo should start with loading CMOS fresh from BIOS chip, which may be helpful in some situations.

I used Continuity with beep for testing.

It's better to use resistance / "auto-Ohm" setting when doing this kind of testing.

Continuity is useful for when you want to quickly find if (maybe) two points are connected together. Note, however, the beeping does not necessarily mean that two points in the circuit are connected together, but rather that there is a low DC resistance path between them - that is, relatively low to what the meter "thinks" is low. Most multimeters will trigger the beeping at around 50 Ohms, give or take 25 Ohms. So in other words, it's not a precise method for finding a "short-circuit", and also does not typically give as high resistance resolution as resistance / Ohm mode on lowest scale. I've seen this confuse a lot of people into thinking they have a shorted MOSFET in the CPU or GPU or chipset section of a motherboard / video card, when in fact that wasn't the case.

So for quick low-resistance path connection finding, you can use continuity. But once you find a *possible* low-resistance path, confirm how "good" that low resistance is with resistance check.

Pin 14 (-12v) beeps on continuity shown 11ohm
on Auto Ohm mode shows OL

One of these two must be a fluke.
Continuity beeping implies low resistance (under 100 Ohms typically), while OL implies open-circuit / infinitely high resistance (or at least, resistance higher than what the multimeter can measure.)

So if one of these is true, then the other cannot be. Either double-check your measurements or check your multimeter is functioning correctly. If you have another multimeter to confirm, that would be even better... though if you don't, that's OK.

My first suggestion is to touch both of your multimeter probes together and set the MM to continuity. On the screen, what resistance / voltage drop do you get? Now switch to resistance mode with the probes still connected together. What resistance do you get now? Both of these should be very low values, showing you what is the lowest resistance / voltage drop the meter can measure on each of these settings.

Also, another thing to keep in mind is that continuity test applies higher test current and voltage, typically enough to bias semiconductors. So that's another reason to use resistance when checking for short-circuit on a motherboard, as the resistance setting typically will not provide enough current to bias semiconductors.

Pin 9 is the 5VSB I get 11ohm on Auto ohm mode.

Somehow, I also find that unlikely (as in, perhaps your MM not reading something correctly and not necessarily something you're doing wrong), since this should indicate there may be something shorted on the 5VSB rail, yet you are able to turn the motherboard On and Off, which means the On-Off logic (supplied by standby power) is functioning.

The only thing I can think of... when you are doing these continuity / resistance tests, how long do you keep the probes connected before you take a reading for granted? Ideally, you should wait at least 1-3 seconds for the resistance to settle down to a stable value. Power rails with large capacitance can make the meter think there is a short-circuit for a brief moment otherwise.

I bought a LPC motherboard Debug board that I will try see if I get a code.
Yeah figured it was not the Capacitor so I put it back

for testing there is no CPU or Ram in the Mobo.

For doing resistance checks, that's fine.

But if you want to see what codes you get with the debug card, you will need to have CPU and at least one RAM stick. Maybe GPU too, if the board doesn't have any onboard and might complain otherwise.

[doctorman]

Re: Asus B250 mining motherboard component issue

.

so I used the ohm only , removed CMos battery, reset my ohm meter to zero
held it on place for few second till stabilized the reading

Pin 14 (-12v) shows 13ohm

Pin 9 is the 5VSB I get 13ohm


seams like pin 14, 15, 16 are connected together

If I had schematics perhaps I could have tried better , but I got nothing and no experience no advanced tool so I feel completely hopeless

Does anyone know of a good service I could send this in for repair?
Preferably local to New york
how much would repair cost?

I also tried the debug PCIE-LPC board , gave me error 58

Model KQCPET6

https://www.amazon.com/gp/product/B07JK4JGLM

https://www.pctestcards.com/download/

not sure if I used it right , put the CPU and RAM in and connected the board to my LPT and PCIE
I do not see any physical issue..

not sure what parts I can really check on circuit with a multimeter to narrow it down more

[doctorman]

Re: Asus B250 mining motherboard component issue

I might be using the wrong pinout for the ATX 24 pin...
none of this makes sense ...

what is a good pinout map, I wills tart over

[doctorman]

Re: Asus B250 mining motherboard component issue

ok so I was using the PCU side of the pinout , switched it over to the motherboard pinout and it all makes sense

looks like there is no short at least in the big picture

now the question is .. what the hell is wrong


could be a bad BIOS flash ?

https://imgur.com/a/Hp5hl13

[momaka]

Re: Asus B250 mining motherboard component issue

ok so I was using the PCU side of the pinout , switched it over to the motherboard pinout and it all makes sense

looks like there is no short at least in the big picture

Yup, figured something had to be amiss there with those resistances, but couldn't tell exactly what, with this being a mining board with multiple power connectors.

could be a bad BIOS flash ?

Perhaps... or maybe not. Did you flash the BIOS prior to that?
BTW, I can't view the last picture you posted above on imgur. Not sure what it is, but I've been having a problem with imgur for a while now. Please attach pictures to the forum using the forum's built-in function.

I also tried the debug PCIE-LPC board , gave me error 58

Can't find much info on this card, but it appears to be pretty much a standard POST card. If that's the case, then my next question is regarding that "error 58" you're getting - does that appear instantly on the LPC debug / POST card, or does it go through some other numbers first before that? If it's the latter, then the board is actually trying to boot, but something in the POST process is failing.
And if it's the former case, where the LPC debug card is instantly showing that error 58 number, remove the RAM, but keep the CPU and test the motherboard with the LPC debug card again. Now does the code change on the LPC debug card to something else? If yes, then that means the BIOS is getting read and the board attempting to POST. Just then have to find out what is tripping it out... which of course may be easier said than done.

But a good first step is to often check voltage on all of the power rails - that would be the CPU V_core, CPU V_tt, RAM V_dd, RAM V_tt, any NB/SB/chipset Vcc rails, 3.3V standby power (and possibly other standby rails, such as 1.5V, 1.8V, 2.5V, and etc. - varies with board design and boards in general.) A multimeter is typically all you need for that (so it's OK if you don't have any "advanced" tools - I don't either. )

[doctorman]

Re: Asus B250 mining motherboard component issue

Can't find much info on this card, but it appears to be pretty much a standard POST card. If that's the case, then my next question is regarding that "error 58" you're getting - does that appear instantly on the LPC debug / POST card, or does it go through some other numbers first before that? If it's the latter, then the board is actually trying to boot, but something in the POST process is failing.
And if it's the former case, where the LPC debug card is instantly showing that error 58 number, remove the RAM, but keep the CPU and test the motherboard with the LPC debug card again. Now does the code change on the LPC debug card to something else? If yes, then that means the BIOS is getting read and the board attempting to POST. Just then have to find out what is tripping it out... which of course may be easier said than done.

But a good first step is to often check voltage on all of the power rails - that would be the CPU V_core, CPU V_tt, RAM V_dd, RAM V_tt, any NB/SB/chipset Vcc rails, 3.3V standby power (and possibly other standby rails, such as 1.5V, 1.8V, 2.5V, and etc. - varies with board design and boards in general.) A multimeter is typically all you need for that (so it's OK if you don't have any "advanced" tools - I don't either. )

1) the pic was just showing what I got on the 24 pin by ohm meter with the correct pin out numbers, seam fine to me https://imgur.com/a/Hp5hl13
it is working , but not much to see on it

2) some number show before 58 on diagnostic board
if I take out the RAM altogether and run it it gives me a different error for the RAM not present.

so the BIOS should be okay... and there is a bad component probably somewhere

I do not have schematics and have never done Desktop motherboards so no idea where to check for all those ... if you have any guide on that .. it would be great

also checking component on the board might be tricky so any advice as to what I should be looking for or how to test with multimeter would be amazing!

so I don't need power injector?
and I don't need the 24 pin power connected when checking?

many of these components are so small, would benefit from a microscope probably

CPU V_core,
CPU V_tt,
RAM V_dd,
RAM V_tt,
any NB/SB/chipset Vcc rails,
3.3V standby power
other standby rails, such as 1.5V, 1.8V, 2.5V, and etc.



well... I guess I am asking to learn how to diagnose a desktop motherboard component .. that should be short and easy enough lol

[momaka]

Re: Asus B250 mining motherboard component issue

2) some number show before 58 on diagnostic board
if I take out the RAM altogether and run it it gives me a different error for the RAM not present.

so the BIOS should be okay... and there is a bad component probably somewhere

Yes, that indeed confirms the BIOS is OK... or at least to a point.

Now, since you do get other numbers before the 58 with the RAM installed and yet another different number without RAM, that means the BIOS is able to see the missing RAM. Not only that, but for the BIOS to start checking for the RAM, that means the CPU must be initialized. And for that to happen, then the chipset(s) must also be ready... which implies that the board is likely getting power to all important rails properly, because that's one of the very first things that are checked for at POST.

So I suspect you probably don't have a power issue anywhere. If anything, there may be a signal issue somewhere. But whether or not that's caused by a bad component or bad connection, I think that's the part that we don't know.

I do not have schematics and have never done Desktop motherboards so no idea where to check for all those ... if you have any guide on that .. it would be great

I rarely bother to look for schematics for these things, because they rarely exist.

Unfortunately, I don't have a guide either.

Most motherboards are similar enough that I can determine fairly easily where everything is with just a multimeter and based on the voltages I get. Yet, there are too many differences between even motherboards from the same era (in terms of layout and where to find these rails), making it difficult to make a "universal" guide for them.

That said, 1st thing I do is try to identify all of the bulk-regulators... so that means find all inductors/toroids on the board. In the case of your motherboard, there are 4 inductors to the left of the CPU socket (these are probably all for the CPU V_core) and 2 inductors above the CPU socket (these are both probably for the CPU V_tt.) Next, there is a smaller inductor close to the top-right corner of the RAM slots, another to the left of the "ASUS" logo, and yet another close to the bottom-right side of the RAM regulators. From these three, not sure which one does what, but probably at least one of them is for the RAM V_dd, which should be about 1.2V (since this is a DDR4 board, if I'm not mistaken.) Then, the other of the two remaining inductors might be for RAM V_tt (approximately half of RAM V_dd... so about 0.6-0.7V probably.) One of these may also be for the B250 chipset Vcc... which I don't remember what voltage it takes, but it's probably around 1.x V too. And the last smaller two inductors are: one above the B250 chipset heatsink and another to the left of the chipset heatsink between all of the PCI-E x1 slots. Again, I can't say what these may be for, but one may be for the B250 Vcc if the other 3 inductors before that weren't. Or it may be a secondary/standby voltage for the B250 chipset.

Lastly, identify any possible linear regulators. These are usually MOSFETs or regulator ICs in TO-252 case or SOT-223 case (look up these numbers to see what they look like.) From the first picture you posted, I can identify two TO-252 devices: one to the right of the RAM slots but to the left of the "B250" text and another between the 2nd and 3rd row of PCI-E 1x connectors in the lower-left corner of the board. The latter is probably your 3.3V standby generator (takes 5VSB and generates 3.3V)... but could also be a regulator just for the audio or LAN ICs. Right next to it on the left (and slightly above) is a SOT-223 device - that's another linear reg circuit of some sort probably. Again, no telling what it's function is, but the area it's in suggests audio or LAN voltage.

So based on the voltages you get for the above items can help us determine what they are. For the inductors that are through-hole, it may be easier to flip the board upside down when testing and measure the voltage right on the solder point that sticks out of the board. For the SMD inductors and SMD TO-252/SOT-223 regulators/MOSFETs, you can measure the voltages right at their pins. For the TO-252/SOT-223 parts, look up their part number online to see if it's a MOSFET or regulator. Common regulator parts are "1117", "117", "1084", "1085", and "1086", among others. For the 1117/117 parts, input is usually the rightmost pin and output is the tab. For MOSFETs, the tab is usually the input (though not always) and the rightmost pin (Source) is usually the output.

And that's more or less my "guide" on how I check the voltages.

so I don't need power injector?

Probably not for the problems you seem to be having with this board, which seems more of a "signal" issue than a power issue (where a bench PSU may come in handy.)

and I don't need the 24 pin power connected when checking?

Only when checking voltages, as obviously you would need the motherboard connected and powered on by a PSU. But for resistance tests, no, it's better to keep the 24-pin (and other power connectors) disconnected to avoid erroneous readings.

well... I guess I am asking to learn how to diagnose a desktop motherboard component .. that should be short and easy enough lol

There are several checks that I'd say are "short and easy" that I like to do before even digging too deep into voltage measurements and whatnot.

I find more and more issues lately stemming from bad CPU socket BGA. So the first easy check against that... though not 100% accurate or easy to replicate... is to put the motherboard in a computer case and bolt it down. Then put a CPU heatsink (with thermal compound) on the CPU - but don't clamp it down! After this, try powering On the board (with RAM and GPU.) Watch the numbers on the POST card / POST display and note what number they get stuck on. Do this a few times - a few from "cold" boot and a few from pressing the reset button. Now apply pressure with your hard over the CPU heatsink and repeat the same tests. Do the numbers on the POST card change at all to some value that you didn't get before with the previous tests? If yes, you may have a bad CPU socket BGA issue (or bad contact/trace/solder joint somewhere else.)

That's about all of the simple tests I do. I rarely go further if a board has proper power on all of its rails and probably good connections/BGA.

[DynaxSC]

Re: Asus B250 mining motherboard component issue

Hi,

First of all connect the powersupply to the mobo, turn it on and check whether the +5V standby voltage is present (look for the correct pin, as most of the pinouts in the internet are shown mirrored from the cable connector side, not from the motherboard socket - not really useful).

If the +5V standby is low (not 5V, eg 1v or lower), then most probably the chipset has a short and needs to be replaced (this is a BGA chips, replacing not so easy). There should be a power regulator on the board somewhere, and if this gets hot in standby mode, this is the next indication this is the chipset. You can also induce with a separate regulated Power Supply the 3,3V at the output of this 3,3V regulator. Start with very low voltages (up to 1V), if the current is very high, at low voltages then the chipset is gone, the regulator most of the cases is OK as it has thermoprotection. You may need a powersupply capable to generate up to 10A at 1V to check this definitely, above 5A you should be able to see that the chipset becomes hotter, if so then it's definitely broken. Do not induce more than 7-8A, as the mobo paths may burn, and then you can use the board only as donor of parts.

Check also all voltages on all coil-resistors (always in the neighbourhood of mosfets, such grey quaders), generally does not matter on which side you measure. You should find at least 3,3v 5V and some third voltage (something 2V or so). If any standby voltages are missing, you need to look for demaged PWM modules on the board (mosfets, drivers, controllers). Best would be to have another working board to compare.

Next look if the Power On signal from the mobo start pins is arriving at the Super I/O chip (download a specs of the Super I/O. If yes check whether the Super I/O give the output signal to the PSU (PS ON), if not there are probably one or more standby voltages missing, or the Super I/O chip is broken.

Chipsets and Super I/O you can buy on Aliexpress, or you buy another board with such chips, remove the chips, raball it, and install on your mobo. Requires some hardware, trials and excellence to do so.

Check also the whole mobo in detail visually, sometimes you find paths cracked or parts missing, compare with pictures from internet, as may places for parts are not occupied in factory stage already. Check also all connectors if not broken/shorted (eg USB's).

Check also all mosfets for shortages, eg in the CPU VRM section.

This is an indication, hope this helps, but sometimes the search for issues may tak days, so you have to be patient.
Good help can also be found on russian sides, you needs to use an online transaltion tool, google this way: bla bla bla site:.ru, russians repair a lot, ans often have very useful hints.

[AMM]

Re: Asus B250 mining motherboard component issue

Did you ever resolve this? I am in the process of trying to get my B-250 ME (was new in box when I opened it 4/27/21) up and going and am having the same issue. I have bread-boxed it and can't seem to get any POST sounds with speaker attached and was in the process of ordering an LPC debug card when I came across this forum.

[mon2]

Re: Asus B250 mining motherboard component issue

@AMM, would Asus not have this new board covered under warranty?

If the board is new, it must work unless Asus has lowered their QC to zero. Personally had a very bad experience with another board company (need to look it up) from Amazon. They advertised the board can support 32GB but we spent 10+ hours in every combination under the sun, including other 'approved' memory suppliers and no combination worked to yield this 32GB. Only 16GB worked, rest would strap the logic board from POSTing.

Then we switched the main board to another and EVERY combination worked. The design of the board is everything.

Start with, is your memory on the 'AVL' (approved vendor list) ? The boards often feature a LED bank to display which stage of the boot cycle the board is at upon power up. Is that working for you? What is the LED sequence that lights up.

If you are not that far, then your power supply and cabling needs to be inspected. It is ok to even remove the video card (if present); all memory sticks and see if the logic board then boots.

As noted, consider to also ping Asus on these issues and post your results. We have built countless PCs in our lifetime with Asus boards and they are a solid company. There has to be a logical explanation to your new board not working.

Post your updates.

[AMM]

Re: Asus B250 mining motherboard component issue

@mon2 Thanks for your input. I don't recall seeing any LEDs on the board, but will double check this evening. I have tried 1 4 GB stick of 2400 Mhz RAM on both DIMMs as well as populating both with AVL approved RAM (LPC Vengeance) as well as a single stick of 2666Mhz, the CPU I tried (g3920) only supports up to 2133Mhz but I was under the impression that simply because RAM is rated at a higher speed it would clock lower for compatibility.

As far as RMA on the board it was manufactured some time ago and bought second hand from a friend who had it sitting on a shelf since he purchased it ~2 years ago and was never unpacked from the box until I opened it last week (same with the CPU)

I have tried 2 1300w EVGA PSUs that are currently in use on other rigs with the cables that are in use, as well as a few other cables that I had that the PSUs came with.

I have an LPC debug board on the way to see if it can point me in the right direction as well.

[momaka]

Re: Asus B250 mining motherboard component issue

Hmmmm.. if we are getting multiples of these boards with the same issue now, then I don't suppose the problem will be a random component going bad.

If anything, perhaps flashing the latest BIOS (or just re-flashing the BIOS anyways) might be something worth trying. Though with the board not posting, I suppose that means either buy a pre-flashed chip or DIY with a BIOS flasher tool.

[DynaxSC]

Re: Asus B250 mining motherboard component issue

Guy's, I'm afraid LPC card will not work on this board, these debug cards work only on very old mobo's, the current mobo's do not send the debug codes on the PCI bus, so no use of debug cards any more. What you also can do is an easy check: connect the power supply, and check whether any component on the mobo is hot, eg chipset (take off the cooler to check), or any IC/power regulator. You can also force the PSU to give the power to the mobo by shorting #PSON to ground, if the mobo does not start itself. Then you can check the voltages on all of the internal power coils, if all the voltages are present. However this is the last thing I would do, as sometimes it can happen, that something burns if it is demaged. Nethertheless this will indicate what, but it is risky. Another method is measuring all major components, eg mosfets, power filter coils, with ohm meter against ground, and checking against another same kind of board which is working, this is timeconsuming, but if you find differences, you can track the problem. With the same method, it is also possible to measure voltages on components (in stand-by and/or with power forced with #PSON).

[piernov]

Re: Asus B250 mining motherboard component issue

The point of connecting a POST debug card to the LPC bus (TPM header for example) is to get the POST codes directly from the LPC bus (which should always work) and not from the PCIe bus (which often won't work).