XPS13 all rails present, continuous shutoff cycle

[survivalbloke]

This machine was brought in because they did a battery replacement and shorted something out - They mentioned they saw a spark. Now what's happening is that I connect the power adapter and I've got 3.3, 5 and 20 rails all present, then, about every 8 seconds or so, the voltages all drop to zero for 1 second or less, then all go back to their normal voltages. So, it seems it's getting a random power-on signal as if the power button is being pressed, but it isn't. Can we safely say PCH is fried? No visible damage anywhere on the board. Drawing 117ma.

[mcplslg123]

Re: XPS13 all rails present, continuous shutoff cycle

What do you mean by saying "All rails are present"? Going by your statement u have ram supply as vcore which is impossible at consumption of 117mA.

Nothing is wrong in the basic circuit of S5 stage as you are getting +19/3/5V rails. Check whether any voltage appears on ram coil during the brief interval when its turned on. It seems you have some damaged rail in S3 or S0 stage.

Measure resistance across ram coil/vcore coil/vccsa coil with no power connected.

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

lol, I instantly feel like a noob.

I will do my best to locate those components and test them as you suggest. I'll report back with my findings. Thanks so much!

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

Going by your statement u have ram supply as vcore which is impossible at consumption of 117mA.

Can you explain this to me?

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

Check whether any voltage appears on ram coil during the brief interval when its turned on. It seems you have some damaged rail in S3 or S0 stage.

Measure resistance across ram coil/vcore coil/vccsa coil with no power connected.

resistances:

ram coil = .3ohm
+VCC coil = .4ohm
+VCCSA coil = .3ohm

On poweron signal, I am getting a brief flash of .245 vdc on the ram coil, PL601 in the boardview.

[mcplslg123]

Re: XPS13 all rails present, continuous shutoff cycle

Can you explain this to me?

U stated all rails are present. So that meand you do have ram supply+VCC core also available and thats impossible with consumption of 117mA.

[mcplslg123]

Re: XPS13 all rails present, continuous shutoff cycle

Ram coil resistance at 0.3 OHMS is too low. Are you sure about your measurement?? If yes, then proceed to de solder the coil and then measure resistance on both the pads of ram coil.

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

Ram coil resistance at 0.3 OHMS is too low. Are you sure about your measurement?? If yes, then proceed to de solder the coil and then measure resistance on both the pads of ram coil.

Removed the ram coil. resistance of the coil is .3ohms and resistance between the pads is OL.

[reformatt]

Re: XPS13 all rails present, continuous shutoff cycle

Resistance to ground is the required measurement. RAM rail usually measures about 100 ohms or more with RAM in, higher with RAM out.

A fried PCH usually results in a dead short across either 3.3V or 1.05V power rails.

[mcplslg123]

Re: XPS13 all rails present, continuous shutoff cycle

Removed the ram coil. resistance of the coil is .3ohms and resistance between the pads is OL.

I didnt asked for resistance between the pads. Measure resistance on the two pads of coil(after coil is removed). Black meter probe to ground and red probe on these pads,one at a time. Btw, @reformatt has already expressed his doubts and very likely that will come true. Still measure the resistance to be absolutely sure.

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

I didnt asked for resistance between the pads. Measure resistance on the two pads of coil(after coil is removed). Black meter probe to ground and red probe on these pads,one at a time. Btw, @reformatt has already expressed his doubts and very likely that will come true. Still measure the resistance to be absolutely sure.

Yes, thank you for clearing that up and it does appear true. I have 562ohms on one pad and OL on the other. Also, this board has soldered RAM.

Just so that I'm totally clear on this, I linked the schematic of this board above, and I am looking at PL601 which according to the schematic is labeled as part of the DDR controller circuit.

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

Resistance to ground is the required measurement. RAM rail usually measures about 100 ohms or more with RAM in, higher with RAM out.

A fried PCH usually results in a dead short across either 3.3V or 1.05V power rails.

Thank you, I was confused. I'm nowhere near as experienced as many on this forum. I'm pretty good at sniffing out shorts on the 19+ rail but stuff like this, to me, is much more difficult. Really trying to learn though and really appreciate the help of folks here - means a lot.

[Sephir0th]

Re: XPS13 all rails present, continuous shutoff cycle

How can you now have 562 Ohms and O.L., when it was .3 before?

Did you measure resistance across the coil pads before?

[reformatt]

Re: XPS13 all rails present, continuous shutoff cycle

The most important tools we have are our eyes. First thing is to do a physical examination of the board under a microscope. Most faults are corrosion based, but in this case you said a spark was seen so likely you are looking for a burnt component.

If the power is cycling this is usually either an overload, or the abscence of a required power rail. So if the physical inspection reveals nothing, you first measure all coils resistance to ground to check for shorts. CPU power rails all measure very low, so 2 ohms is not considered a short in that instance. If all good, you then go through your power rail sequence on pg 5 of the schematic, and power good signals for each rail.

[mcplslg123]

Re: XPS13 all rails present, continuous shutoff cycle

How can you now have 562 Ohms and O.L., when it was .3 before?

Did you measure resistance across the coil pads before?

I'm also puzzled by this measurements. please clarify

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

How can you now have 562 Ohms and O.L., when it was .3 before?

Did you measure resistance across the coil pads before?

OK, let me clarify, because originally I was measuring incorrectly because I misunderstood.

I'm hoping one of you have looked at that schematic and verified I'm looking at the right coil, but in the schematic it's labeled on the page marked "ddr controller" so I'm assuming it is, PL601.

Originally the coil was on the board and I measured .3ohms on either side of it. Then I was asked to remove the coil and measure resistance of the pads, but I misunderstood and measured from pad to pad, which was OL. After I was corrected, I measured each pad separately to ground, which showed one pad to ground at 562ohms and OL on other pad to ground.

video clip

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

The most important tools we have are our eyes. First thing is to do a physical examination of the board under a microscope. Most faults are corrosion based, but in this case you said a spark was seen so likely you are looking for a burnt component.

If the power is cycling this is usually either an overload, or the abscence of a required power rail. So if the physical inspection reveals nothing, you first measure all coils resistance to ground to check for shorts. CPU power rails all measure very low, so 2 ohms is not considered a short in that instance. If all good, you then go through your power rail sequence on pg 5 of the schematic, and power good signals for each rail.

Thank you for the brief lesson! I have already completed a visual inspection under my scope. I guess I'm wrestling through the next phase of that.

[mcplslg123]

Re: XPS13 all rails present, continuous shutoff cycle

Going by the video, resistance on PL601 is fine and you can solder the coil back. It seems you measured incoorectly resistance on the coil earler.

[survivalbloke]

Re: XPS13 all rails present, continuous shutoff cycle

Going by the video, resistance on PL601 is fine and you can solder the coil back. It seems you measured incoorectly resistance on the coil earler.

Much appreciated, thank you. What would your next step be?

[Sephir0th]

Re: XPS13 all rails present, continuous shutoff cycle

Since we still know to less about where it is Stuck, i'd continue with a overview which power rails are where present in the Brief moment. This means to measure voltage of each coil and share the values, maybe written in a picture of the actual board.