Actions
Forum Thread Details
  • State Suggested Answer
  • Replies 57 replies
  • Answers 7 answers
  • Subscribers 303 subscribers
  • Views 34536 views
  • Users 0 members are here
Related
See a helpful answer?

Be sure to click 'more' and select 'suggest as answer'!

If you're the thread creator, be sure to click 'more' then 'Verify as Answer'!

Repair Challenge: XBOX ONE S! (Beep on then beep off)

lilithelotor

CLARIFICATION

YES, BEFORE STARTING WORK ON THIS OF COURSE I GOOGLED AND WATCHED SEVERAL VIDEOS (MOSTLY USELESS) AND READ RESOURCES ON COMMON XBOX ONE AND ONE S PROBLEMS.

 

  1. I tried the HDD in a PC, and it does show up as several NTFS partitions with files that all look fine.
  2. Blu Ray drive at least powers up and can insert and eject a disc.
  3. Tried the "cold mode hairdrier" trick to spin the fan manually. It's not the fan. The fan works.

All other components of the board seem to work. Just the system started up with a startup sound, and shut down with a shut down sound immediately after. Likely a protection being triggered.

 

Help meee!

 

Trying to figure out this XBOX One S issue: https://www.ifixit.com/Answers/View/523582/PSU+11... Any ideas?

 

It makes the start and then stop sound. Southbridge gets voltage. All 12V test points get voltage. Power rails to CPU suspiciously low resistance. Details in iFixit link.

 

I have an XBOX ONE S that I bought as “for parts” on eBay and the owner says it was kept in a clean house, etc. It is indeed clean inside. No signs of water damage.

The PSU is also fine. The APU had a TON of thermal grease all over the surface mount capacitors. I didn’t remove that yet but I did re-apply some arctic silver and I used a more reasonable amount.

The symptoms I see:

Console when plugged in does a few quick pulses to the APU fan. Is it trying to init?

Console when eject is pressed makes the eject sound, then the shutdown sound immediately after.

Console when power is pressed makes startup beep and then shutdown beep immediately after.

I don’t know what kind of issue I’m getting yet. Is there a JTAG interface or some way to do further diagnosis on this thing?

11.94V sounds perfectly fine for a 12V rail. Should I simpy try plugging a PC power supply into the motherboard to see if it will boot?

Any links or info on diagnosis would be appreciated.

Thanks in advance!

later

Aha! In the image you can see the CPU filter capacitor is registering just 1 ohm! The others are pretty low too but I bet if I remove that one it’ll at least boot. I have a server motherboard with a broken socket that I can steal capacitors from. Might replace the whole row if necessary.

image

later still

Looks like I was wrong. That capacitor I removed still offers up almost exactly the uf value on the tin. So that was another dead end. Hmm... One weird thing, all of the test points register 60Hz but not 110V. So thankfully the PSU isn't pushing wall socket voltage through. But that frequency is weird. My multimeter needs a charge though so I'll measure that again after it has.

The entire power area for the APU offers almost no resistance to the ground plane which is worrying. Again, need to charge that multimeter and re-check.

later again

Power on with no heatsink, the processor doesn't even warm to the touch. CPU, GPU, and other test points around the board pulse to 1-2V and then turn off fast.

Can't help but think it's related to the very low impedance between ground and positive on the V rails for the CPU/GPU. Is that normal?

USB port is charging the multimeter image

So that 5V rail from the southbridge is working.

Lots of 12V test points are consistently 12V (11.94V)

So...

Does any one have a diagram of the test points around the motherboard? What voltages / values should I be seeing? What are some likely culprits?

Thanks in advance! You're all wonderful!

 

This is the original XBOX One. Similar, but not exactly what I need. I can still maybe work from it.

 

image

  • 0 in reply to lilithelotor

    If you get a chance, could you check the measurement directly across the capacitor holes (i.e. by flipping the PCB)?

    Since each red location on the photo appears to have via stitching (I can't be sure, but it appears that way), it is possible to all be ground plane connections. Some circuit boards will do that (they will add more layers to be the ground plane, and some may use the top side partially for this too).

  • 0 in reply to shabaz

    I tested from the big heat sink hole rings to the positive and the negative on the capacitor holes and got short on both. I assumed that would likely be a good ground point.

     

    I can try measuring across tonight after work. image

  • 0 in reply to lilithelotor

    I see. In any case the red points are possibly connected to the ground plane I suspect - especially if this is the case across all three power sections. These are likely three separate rails, and it would be too much of a coincidence for all of them to have failed in this manner.

    X86 (at least Intel) have pretty complex supply requirements (assuming this is x86 - I know nothing about xbox).

    I don't know what the reliability of an xbox is, but the mechanical stuff (spinning drive) may have a very low reliability in comparison to the rest of the xbox - but I'm presuming that has been eliminated or confirmed working in some way (I didn't watch the entire videos).

  • 0 in reply to shabaz

    Of course the hard drive was the first thing I tested. It works great and the partitions show fine in a Windows machine. (5 NTFS partitions I believe)

    Yes XBOX is X86. It's literally just an AMD APU running Windows 10 specialized for a game console. image

    I just measured across the capacitor holes. (Caps are not present at the moment, tested them separately. They are fine. Around 820uF which is their marking.)

    The capacitor holes from negative to positive result in:
    C8E3: 1.7 ohms

    C8E2: 1.7 ohms

    C8E1: 0.8 ohms

    C8D3: 0.8 ohms

    C8D2: 0.8 ohms

    C8D1: 0.8 ohms

  • 0 in reply to lilithelotor

    Further testing:

    R9E5: 2.4 ohms

    R9E4: 2.3 ohms

    R9E3: 1.2 ohms

    RT9E1: 8000 ohms (8kohms)

     

    TP9G1 GND --> TP9F1 CPUCORE: 1.7 ohms

    TP9G1 GND --> TP9D2 GFXCORE: 0.8 ohms

  • 0 in reply to lilithelotor

    The small chip may well be a MOSFET driver, the chances of them going bad are pretty slim.

     

    As for your readings ... I'm not sure what that could mean - if not the MOSFETs, maybe that's actually current flowing through the actual CPU/GPU chips? I tend to avoid mindless probing across pads that aren't just across "one" component because there's always a chance of interference or current being driven into things which shouldn't be driven in such a way.

     

    Does the resistance hold stable or does it change after a few seconds? Are there no non-obvious solder-splashes, or board getting pinched/crushed at locations (especially important for multi-layer)?

     

    - Gough

  • 0 in reply to Gough Lui

    Indeed.. resistance readings are not something that make it easy to fault-find within the circuit. Especially if there's no known working xbox to compare to.

    By chance I've got a Intel x86 PCB on my desk (a working one) and although I don't see resistance values as low as those on the power rails, they are still in single-digit ohms.

    But still fairly meaningless.

  • 0 in reply to shabaz

    I mean, thanks but that doesn't help me move forward much, everyone.

     

    What can I do to positively affect this situation? image

  • 0 in reply to lilithelotor

    Without understanding what's wrong, all I see is bits being pulled off at random.

    This is not recommended, so better we stop that - there is a severe risk of causing more damage by powering up as you have done, with one of the rails deliberately unpowered - and if no concensus is reached on how to proceed, maybe there is no magic solution just yet.

     

    Perhaps if you come across another xbox, then it can be compared to see where the anomalies are. Or maybe some diagnostics exist - either something that plugs on, or some LED blinking pattern.

    Or maybe someone who has already come across the issue described in this thread may read it one day, and provide the answer.

  • 0 in reply to shabaz

    It isn't random.

     

    It might help to go through what has been said in the past.

     

    At Gough's advice I'm going to replace all of the MOSFETs. Thing is, I'm not convinced they're bad.

    I thought a good strategy might be to remove components that are likely to fail and test for grounding out again without them.

     

    If, as you say, the X86 rails practically ground out too, maybe that's a dead end and I should look for other common issues.