XBox One S - No Power On - Missing 3.3V

[LiftedTrace]

There are many that are not turning on, but some of them I suspect will never turn on until the power button is pressed and the whole thing wakes up. I just do not know which ones those are that should be in all the time, and those that only come on at power up.

It may be time to take my working one apart and use for a reference.

As a side thing, there was clearly something wrong with the output of U3D6, or what ever makes it tick. Do you know the part number of that thing?

EDIT 1:
Ok well I am 
When it occurred to me that other power supplies were not turning on until power up, I connected the power button board, and low and behold! The 3.3v I was missing suddenly came to life 
The enables for that 3.3 chip ( U4C2 ) is not present, as well as the enables for the two on the lower left ( U3F1 and U3E1)
AND!, when I press the power button, I get a beep sound as it’s powering up, and the shuts down right away.  So it seems my problem is not what I thought after all, lol. 

Also that strange 3.3v issue seems random. It dropped back to .8v again and then back to 3.3.
There is clearly something wrong with that section. But while it is working and the board powers up, I see some power supplies spike for a second and then the whole unit powers down.
When I power up, I see no increase in current draw but 20ma, so I do t think I have a short. I wish I did, it might make this easier to hunt down.
So clearly something is not seeing the correct voltage and it’s shutting back down.
The main 12v supply for the APU doesn’t even hiccup at power up, so either that is a delayed start after the rest of the board chimes in as ok, or my issue is over there.
One mosfet ( Q9D1 ) goes low on the gate when powerup.
I see a spike on its output for a brief time. I’m going to try and track that down and see where it goes.
It doesn’t seem to be tied into the three groups of fets for the APU supply, but I suspect maybe it enables them some how.

[fzabkar]

U3D6 appears to be marked as "GEH", in which case it is in my BOM.

[LiftedTrace]

U3D6 appears to be marked as "GEH", in which case it is in my BOM.

Sorry, I missed that. Thanks.

It might help to compare this PCB with your water damaged one. A similar hires photo could help us to identify the date and batch codes from the part markings.

I no longer have that xbox. The only other one I have is my own working one and I am still on the fence for opening it up or not.

I don't think it is a good idea to backfeed the supply voltage into the outputs of your regulators.
Have you checked all the other supply test points? Is your 3.3V supply the only one that is missing now? I suspect that the 1.5 ohm short may have been due to an MLCC (ceramic cap).

I don't know how the two are tied together, the enable of U4C2 and the output of U3D6, but there is a shady looking cap on the back side of U4C2 (the 3.3V chip) that is connected to its enable pin, and I was pretty sure last I checked it, it measured short to ground. Maybe that cap is intermittent problem for why the 3.3v is sporadic. And every time the 3.3v chips gets a good enable, the output of U3D6 is also good.

There are still some test points that do not turn on. I am trying to track down where Q9D1 goes off to. That closes when it powers up and then the system shuts down. I am thinking maybe the power from that is not getting to its intended destination. That's my theory anyways. I follow it into the enables of the three small chips U9D2, U9E1, U9E2. Those I suspect control the power for the 12v for the APU and I do not even see a blip of turn on when the box is powered on. It might also be nothing.

Here is a higher res stitched together photo with more readable text. : https://www.dropbox.com/s/k1f7ol4rekuwrwq/XBOX%20Front.jpg?dl=0 Seems my browser wont open that file, probably to large. Might have to be downloaded.

EDIT 1:
So I located what controls U9D2, U9E1, and U9E2. Its the larger chip NCP4205 (U9C2 ). Its coming from pin 28 on that chip. I looked thru google results for any type of datasheet but cannot seem to locate anything and I didnt see it on you BOM. I was hoping to back track thru and find out whats going on, but without a description of those pins, I have no clue if its working as intended.

[fzabkar]

So I located what controls U9D2, U9E1, and U9E2. Its the larger chip NCP4205 (U9C2 ). Its coming from pin 28 on that chip.

Is it possible that the NCP4205's enable output is actually an enable input? That is, is the NCP4205 also being controlled from some other chip?

FWIW, here are datasheets for chips with a similar part number:

NCP4200, ON Semiconductor, Programmable Multi-Phase Synchronous Buck Converter with I2C Interface, 8-bit VID DAC, 4 x PWM, Vin = 5V, QFN-40:
https://www.onsemi.com/pub/Collateral/NCP4200-D.PDF

NCP4201, ON Semiconductor, Programmable Multi-Phase Synchronous Buck Converter with PMBus, 8-bit VID DAC, 4 x PWM, Vcc = 5V, QFN-40:
https://www.onsemi.com/pub/Collateral/NCP4201-D.PDF

In fact, if there is a capacitor between the enable pin and ground, then this would suggest that the enable is configured like a simple RC power-on reset.

Code: [Select]

V+ --- R --+--> Enable
           |
           |
          --- C
          ---
           |
          _|_
           =

[fzabkar]

Just for completeness, could you identify the markings of these 5 chips on the back of the PCB?

U5N1 is associated with L6B1 on the top side, so it must be a stepdown buck regulator.

U1P1 DFN-10

Q1P2 SOT23-3

U5N1 SOT23-5

Q6R2 SOT23-3

U6U1 SOIC-8 audio amp

[fzabkar]

These would be my power supply test points:

[fzabkar]

More test points ...

[LiftedTrace]

I really appreciate the time your taking to help me figure this out. Its a learning experience for me. I think I took care of the strange 3.3V output with the flaky enable. I replaced a cap and a resistor that were connected to the enable of U4C2. I cannot believe how challenging it is to replace these little buggers. Now that this enable has a good voltage to it, the 3.3v is working well. That being said, here is a list of all the test points you were asking for in them pics.
The ones that state "only after power on" go right back to zero after the unit shuts down.
TP4D1 = 3.29
TP4D2 = 5
TP4D4 = 1.83
TP4D5 = 1.135
TP4E1 = 1.11 (Only after power on)
TP4F1 = 1.81 (Only after power on)
TP4F2 = 1.83
TP5F1 = 0 (Even after power on)
TP5D1 = .6 (Only after power on)
TP5D3 (v_fuse) = 0
TP6F1 = 0 (Even after power on)
TP8B1 = 12
TP9D1 = .72 (Only after power on)
TP9D2 = 0 (Even after power on)
TP9F1 = 0 (Even after power on)
RT3F1 = 5
RT6B1/2 = Both 5V
L6B1 (Vout) = 12V
C9C5 (Vout) = 12V
U5D2 (pin 2) = 3.29
C5D4 = 1.84 (Only after power on)
U3D6 (Pin 5) = 1.83

The Chips on the back - Identification
U1P1   1206            Q1P2    KFD                  U5N2    C08F                                  Q6R2       KFD                         U6U1     8104SY1             
            87K                           829                  (C has a line over it)                                       829                                     1814
           ADRG                                                 (08F have lines under them)                                                                     28053010           

Is it possible that the NCP4205's enable output is actually an enable input? That is, is the NCP4205 also being controlled from some other chip?

I am not sure. I know the data sheets for NCP4200/NCP42001 do not have the same number of pins, but they list the enable pins as pin 6.
The pin that is connected to the three chips, U9D2, U9E1, U9E2, is 28. Pin 28 on UC92 is connected to pin 3 on all three of them chips thru a resistor.
UC92 (Pin 28) -----------|-----(R)------U9D2 (pin 3)
                                        |
                                        |-----(R) -----U9E1 (Pin 3)
                                        |
                                        |-----(R)-----U9E2 (Pin 3)

If NCP4205 is similar, then it looks like the enables are all tied together to the bigger chips PWM(one of them) outputs, or the SW(one of them) output. Thats assuming they are similar enough where the PWM and SW are on the same sides.
Also, Pin 2 on U9D2 is connected to pin 27 of UC92.

So one thing I noticed when checking these test points is two of them were low when it powered up.
TP5D1 = .6 (Only after power on)  It should be 3.3v
TP9D1 = .72 (Only after power on) Not sure what this should be.
TP9D1 seems to be coming from the power supply in the lower right corner (U9F1). This seems to start up a bit but I dont know what it should be at. The test point is labeled MEMIO, so maybe it should be this low?

EDIT 1:
Forgive the hackery and ugliness of my skills. This: https://www.dropbox.com/s/ilu2rzqa8l6ucx5/XBOX%20Back.jpg?dl=0
is a photo of the back of the board. I seen this on the other xbox I had, the way the caps look. Like one side is missing the solder covered connection. Located on the left side in section "S" and "T" and "U".
"U" seems to be missing a Cap all together, and the one that is present seems to be missing its right solder connection. I remember this from the other board. I wonder how much of this is from water damage and how much is how it was from the factory.
The section in U1 also has some ugly looking caps, and when I tried to touch up the test points in that area, the board masking peeled off.
I wonder if there are traces inside the board that have eroded from the water damage. Is that even possible?

[fzabkar]

I would think that MEMIO is the supply rail for the SDRAM chips, in which case it should be 1.5V (according to the data).

I would also measure the resistances between ground and each of the suspect test points.

ISTM that the SoC (U4D2) is powered from the primary standby supplies (V_3P3STBY, 1P8STBY, 1P1STBY). When it senses the power-on switch, it enables several other supplies, and these in turn enable the supplies for the SDRAM, CPU, GPU, and NorthBridge. I suspect that one of these supplies, or its load, has a problem, and that this problem causes the power-up sequence to terminate.

I would concentrate on those supplies that never come up to spec, even briefly (the last 7 in the list). Check their inputs, outputs, enables, and load resistances.

Code: [Select]

TP4D2 = 5P0            5
RT6B1/2 = rear USB     Both 5V
RT3F1 = front USB      5

L6B1 =                 12V
C9C5 =                 12V
TP8B1 = V_12P0         12

TP4F2 = V_BAT          1.83

TP4D1 = V_3P3STBY      3.29
TP4D4 = 1P8STBY        1.83
TP4D5 = 1P1STBY        1.135

U5D2 (pin 2) =         3.29
U3D6 (Pin 5) =         1.83

TP4E1 = SB1P1          1.11 (Only after power on)
TP4F1 = SB1P8          1.81 (Only after power on)

C5D4 = V_SOC1P8        1.84 (Only after power on)

TP5D1 = V_3P3          0.6 (Only after power on)      should be 3.3V (from U5D4 ?)
TP9D1 = MEMIO          0.72 (Only after power on)     should be 1.5V from TPS51916 et al

TP5D3 = V_FUSE         0                              from pin 5 of U5D3 LDO reg (AP2127)

TP5F1 = SOCPHY         0 (Even after power on)        from U5F5 switcher
TP9D2 = GFXCORE        0 (Even after power on)        from NCP4205 switcher et al
TP9F1 = CPUCORE        0 (Even after power on)        from NCP4205 switcher et al
TP6F1 = NBCORE         0 (Even after power on)        from TPS53819A switcher et al

[fzabkar]

I just reread one of your earlier posts and I'm sorry for my misleading reply. Of course the NCP4205 chip enables each of U9D2, U9E1, and U9E2. Those are the MOSFET gate drivers for the CPUCORE.

Doh! (slaps forehead)

[LiftedTrace]

TP5D3 = V_FUSE         0                              from pin 5 of U5D3 LDO reg (AP2127)

Ok, So the output doesn't even hiccup on power up. The Shutdown control pin 3 i see a tiny spike when power on. I cannot track down where that pin 3 is connected. Its resistance to ground is 10K. I am suspecting its not getting enough to turn on. I could supply it with a small amount, but I am reluctant to, not knowing what it needs or where its going, so I am going to not do that. There are two diodes right next to it. (D5D2), the cathode goes to 3.3V after power on, the anode goes to .06. If I am not mistaken, should this not be around .7?

TP5F1 = SOCPHY         0 (Even after power on)        from U5F5 switcher

The input has a steady 3.3V and the output has nothing. I probed the other pins and found no other voltages present before or after power on. I cannot locate the datasheet on this chip either so I dont really know what pin does what.

TP9D2 = GFXCORE        0 (Even after power on)

This is coming from the three groups of mosfets that are not coming up. The chip that controls them getting its enable from NCP4205.

TP9F1 = CPUCORE        0 (Even after power on)        from NCP4205 switcher et al

Ok, so I may be getting closer to something.
TP9D2 = GFXCORE  & TP9F1 = CPUCORE seem to be tied together. I noticed something strange though.
When I measure the resistance between TP9D2 = GFXCORE and the output of Q9D3(larger mosfet by large first coil), I get .3_Ohms.
When I measure the resistance between TP9F1 = CPUCORE and the output of Q9E3(larger mosfet by large last coil), I get .3_Ohms.
When I measure the resistance between TP9D2 = GFXCORE  & TP9F1 = CPUCORE  I get 10.4_Ohms.

So I started measuring the resistance between each of the larger mosfets.
1-2 = .3
2-3 = 10.4
1-3 = 10.4

1-G = 3.6
2-G = 3.4
3-G = 7.8

I find it strange these are not all the same which makes me think the issue may be on this power supply.
Maybe its fine? I do see a slight gap between the two top coils and the bottom coil. Maybe I am measuring the resistance thru the APU from the GFXCore and the CPUcore

TP6F1 = NBCORE         0 (Even after power on)        from TPS53819A switcher et al

The Vdd goes to 12v on power up. The enable and output gets nothing on power up.

TP5D1 = V_3P3          0.6 (Only after power on)      should be 3.3V (from U5D4 ?)

Input is seeing 5V, Output spikes on power up, Enable Spikes a small amount on power up.

After measuring all of these, the only one that seem odd to me is the one by the diodes (U5D3).
Would it be normal for the voltage before the diode to be 3.3, and after be .03-.06?

[fzabkar]

ISTM that MOSFETS Q9D3 and Q9E1 supply the GPU core, while MOSFET Q9E3 supplies the CPU core. In fact you can see "CPUCORE" adjacent to capacitor C8E3, and "GFXCORE" between C8D1 and C8D2.

The resistance that you are measuring "between the two cores" is actually the sum of the resistances from CPUCORE to Ground, and from Ground to GFXCORE, ie 7.8 + 3.5 = 11.3 ohms.

[fzabkar]

After measuring all of these, the only one that seem odd to me is the one by the diodes (U5D3).

Would it be normal for the voltage before the diode to be 3.3, and after be .03-.06?

I don't understand what you mean by "before and "after". The forward voltage drop would be limited to around 0.6V, while the reverse voltage rating could be 20V or more. A Schottky diode would have a lower Vf than a regular silicon rectifier.

[LiftedTrace]

I don't understand what you mean by "before and "after". The forward voltage drop would be limited to around 0.6V, while the reverse voltage rating could be 20V or more. A Schottky diode would have a lower Vf than a regular silicon rectifier.

My apologies. What I meant by "before" was the reading at the cathode and "after" was the reading at the anode.
Would a schottky diode have a forward voltage drop under .1v?
Maybe the only was I can check is to remove it and check out of circuit?

ISTM that MOSFETS Q9D3 and Q9E1 supply the GPU core, while MOSFET Q9E3 supplies the CPU core. In fact you can see "CPUCORE" adjacent to capacitor C8E3, and "GFXCORE" between C8D1 and C8D2.

I see that now. By the way, what does "ISTM" mean?

https://www.diodes.com/assets/Datasheets/ds30308.pdf I believe is the schottky (D5D1 & D5D2). This lists the forward voltage as .3 ish.
Am i on the right track thinking this should be higher than .03-.06?
 

[fzabkar]

ISTM - it seems to me

A forward voltage drop of 0.1V is quite possible for a Schottky diode if the current is very low.

FWIW, I checked the V_EFUSE references in the Xbox 360 circuits. In that particular design the CPU generates a VREG_EFUSE_EN signal which enables an LDO regulator. This regulator generates a V_EFUSE supply from the V_5P0 supply, and V_EFUSE then powers one section of the CPU's internal logic. I have no idea what V_EFUSE does.

The V_3P3 supply in the Xbox 360 is generated by an LDO regulator, and this regulator is controlled via a VREG_3P3_EN signal from the SouthBridge.

Did you check the MEMIO regulator? Its output should be 1.5V.

[LiftedTrace]

At this point, it seems I have quite a few power supplies not turning on because they are lacking their enables for some reason. Either they have no enable at all, or the enable never stays on. Other than the really low voltage on the schottky, I can not imagine what it might be. I could take that off and measure it just to see I suppose, but with a reverse break down voltage of 30v, I find it hard to believe this is the problem seen as its only seeing 3-4V.

I am trying to stay away from opening my xbox up, but I am thinking I may need a working one as a reference.....unless you may have any other options to check?

EDIT:

Did you check the MEMIO regulator? Its output should be 1.5V.

I did, it seems to be controlled by U9F1. Pin 13 (SW) spikes to about .7, same as MEMIO test point.
Pin 16&17 shoot up to 2.5 for a split second. I believe these are the enables of the chip. http://www.ti.com/lit/ds/symlink/tps51916.pdf (Page 17, State S0). Pin 16&17 are tied together, both are either Low (State S4/S5) or High (State S0).
So this chip too is not getting its enable held high.
Something else is triggering this whole thing to shut back down.

[fzabkar]

I am trying to stay away from opening my xbox up, but I am thinking I may need a working one as a reference.....unless you may have any other options to check?

MEMIO regulator?

You could check the voltages on U5D4.

Here is my guess:

pin #1 - enable from southbridge
pin #2 - ground or 5V input
pin #3 - 5V input or ground
pin #4 - V_3P3 output
pin #5 - not connected ?
pin #6 - feedback divider

[LiftedTrace]

Ok, well I looked a little deeper. The pins on the chip itself (16&17) are not connected, but a resistance check between them shows .3_ohms.
Right behind it, is the ugliness I was talking about in the high res image of the back of the board.
The fact that pin 17 is connected to this ugly looking cap makes me wonder.

[LiftedTrace]

You could check the voltages on U5D4.

Here is my guess:

    pin #1 - enable from southbridge
    pin #2 - ground
    pin #3 - 5V input
    pin #4 - V_3P3 output
    pin #5 - not connected ?
    pin #6 - feedback divider

Correct. This chip has 5V on it. The enable spikes up and back down, and Pins 4&6 seems to be connected in a feed back configuration. Pin 5 seems t0 be NC

[fzabkar]

It seems odd that you found and replaced one genuine shorted component, but now you have another problem. I'm wondering whether you introduced a second fault.

My other question is, can the CPU/GPU run without a heatsink? Does it throttle itself if it gets too hot?

[LiftedTrace]

It seems odd that you found and replaced one genuine shorted component, but now you have another problem. I'm wondering whether you introduced a second fault.
My other question is, can the CPU/GPU run without a heatsink? Does it throttle itself if it gets too hot?

It is very strange, I agree.
As for the heatsink issue, I would say no, it cannot be run without a heatsink, but its not even getting powered up to generate any heat.
When I press the power button, I go from 30ma to 40ma current draw and back down. It it is only powered up for 1 second top, maybe less.
The system shuts down almost instantly.

[fzabkar]

Ok, well I looked a little deeper. The pins on the chip itself (16&17) are not connected, but a resistance check between them shows .3_ohms.

I think we are both going blind. Pins 16 and 17 of U9F1 are connected on the top side of the PCB.

[LiftedTrace]

I think we are both going blind. Pins 16 and 17 of U9F1 are connected on the top side of the PCB.

hahahahah  I must be tired, lol.
I was mistaking pins 11&12 for 16&17 DOH!!!
Ok, then in that case, on the photo i uploaded, pin 12 is the one thats connected to that funky looking cap.

So 16&17 do still only get a short burst of voltage and turn off again. What troubles me is when I measured between 11&12, thinking they were 16&17, I did get .3 ohms.
Thats .3 ohms between pins 11&12. The data sheet lists them as V5IN and DRVL.
I dont think these guys should be measuring a short between them do you?
Unless they are connected internally? One being an input and another an output?

[fzabkar]

What troubles me is when I measured between 11&12, thinking they were 16&17, I did get .3 ohms.

Thats .3 ohms between pins 11&12. The data sheet lists them as V5IN and DRVL.
I dont think these guys should be measuring a short between them do you?
Unless they are connected internally? One being an input and another an output?

They should not be connected, either externally or internally. Something is wrong. Is it possible that you shorted the two pins while you were testing?

Otherwise, it could be that Q9F1 and Q9F2 are damaged.

[LiftedTrace]

They should not be connected, either externally or internally. Something is wrong. Is it possible that you shorted the two pins while you were testing?
Otherwise, it could be that Q9F1 and Q9F2 are damaged.

Its very possible I shorted them testing. The pins are very close together and I do not have a microscope. Only a small eye loop.
Measuring across Q9F1/2 doesnt show a short though.

EDIT:
OK, I think pin 11&12 were shorted. one of the pin solder was mushed. I fixed it. Same result though.
Pin 12 has constant 5V. now. Pin 11 spikes to about 3.2 then off again, and 16&17 spike up and down.
So it seems this one may me working too, but not getting its enable to stay on.