HP Envy 17 - J184NA - No POST

[pjzz2000]

Board Model Number: SPS 720266-501

Inventec 17SBGV2D 6050A2549801

Apparently no schematic available.


Symptoms:

Charger light on,
Power button light on
Doesn't POST or boot to BIOS

Diags:

1. 19.5V rail present all over as input to all other power rails

2.. With charger connected (no battery), 3.3V and 5V present at respective co-located inductors, but after about 20s disappear. Weirdly 3.nV remain present at various locations, including BIOS and around SIO and power button. Where is that power coming from?

3. When short power button pins, power led lights and says lit

4. Only other sign of action, a 1s flicker of CPU fan

5. After that, no progress. If I short the power button again, the fan makes a kind of high-pitched buzz like it wants to turn but insufficient power. It has something like 1.7v on the + connector of fan

6. With power LED on, the 3.3V, and 5V inductors are reinstated and steady

7. Lots of 3.nV and lower present around and directly below the PCH,

8. CPU is cold, like it has done nothing, or next to nothing, which makes sense as no power to them through the various CPU-neighbouring Inductors.

So, any thoughts on my theories below, or any other ideas?

a) SIO and/or PCH fails to "read" BIOS so doesn't bother turning on the CPU power rail(s), in order to initiate POST.

b) CPU power rails dead for some other reasons (have not felt any heat around that area)

c) CPU dead (or other chips dead)

I have not been tempted yet to try re-flashing the BIOS chip yet because of the way the priblem developed at the beginning:

I got the problem sporadically, then it died for a significant time, then it booted to BIOS a few intermittent times. It felt more like a solder joint problem, working sometimes at different temperatures. Not sure how BIOS could have been corrupted and then fine again, then corrupted, then fine.

So Im pretty stuck on what to do next. Im a beginner but learning a lot and determined!!: Any help much appreciated!

Thanks...

[mcplslg123]

As far as your question#2 is concerned, the behaviour is absolutely normal. When you plug in power,3V/5V will appear for few seconds and go away and will come back only after power button press.
The reason is "DEEP SLEEP WELL" technology to save power.

My suggestions:
01. Measure voltage on all coils after pressing power button
02.Measure resistance to GND on all coils(no power,battery attached).Allow some time to discharge and then measure resistance.
03.Phsical observation of the whole mb out of casing .

Its not possible to conclude at this stage whether cpu is faulty or not. Cpu will be powered after all its requirements and communication with PCH is fullfilled. So lots of troubeshooting before you reach there.
Right now we dont even know whether your are reaching S4 stage not to speak of S0 stage.

[pjzz2000]

Many thanks!

Will come back to you.

Could I ask a question, not necessarily related, but just for for my knowledge.

I've watched good videos on 3.3/5V power rails generation. The 3.3v feeds various key stakeep co.ponents, SIO, BIOS, and Power button... so my question: what powers the voltage regulator ICs that manage the mosfets to generate these voltages.... ?

[mcplslg123]

First 19V will pass through the whole board when your dc-in mosfets are correctly biased(charging circuit). Then this 19V will reach the stepdown controller as VIN. Depending upon the stepdown controller,it"ll generate 3V LDO and this 3V LDO will power the SIO/P.BTN etc. Some step down like TPS51225C dont require anything else then VIN to generate LDO.Newer SYliergy stepdowns like "BAC" requires ENLDO to generate LDO besides VIN.

[pjzz2000]

Thank you!

So is the 3v LDO power rail separate from the 3.3/5v switched power rails?

And does the 3v LDO power rail provide power to SIO, Pwr button and to all the switched power supply ICs?

I'm finding it hard to find out the details of what powers the switched voltage regulator ICs when 3.3/5 switched rails do not yet exist.

(I took all the inductor resistance to understand but I lost them while editing. None were shorted to ground. The minimum were the ones closest to CPU @ 24 ohms. Will take measures again, and post the values).

(Some of the measurements started very high resistance and gradually reduced from M ohms, down to k ohms. I dont understand that when no power on board)

Several of the inductors have 0v (or mV) after power button. 19.5v available across the board, 3.3v, 5v present at inductors, but no voltage at the other 7 inductors

[mcplslg123]

3V LDo is separate from 3V ALW rail and as mentioned 3V LDo powers the SIO.3.3V/5V are present on coils as per your claim.Since power button light comes on i beleive the mb is switching as well.All other rails will be powered from 3VS/5VS .Locate the dc-dc converter on your mb and tell us the code on it. It must get the enable to switch power from 3.3VALW/5V ALW to 3VS/5VS respectively.

[pjzz2000]

Thanks for above info.

Sorry, I've had a family bereavement, so I'll be some time getting back to this.

I was looking at vcore vrm which all looked correct and the vcore ic's (integrated mosfets).

They look good with 1.nV supplies output but nothing at the neighbouring coils.

I will go back and look at all of this again to see how far it is getting.

One question, if you can help. Are SIO signals high/low and detectable with dmm, or are they sometimes transient, they go high, then back to low immediately once acknowledged by the recipient ic?

Thanks for all your help, I'm learning a lot!!

[mcplslg123]

SIO measurements can be done with DMM on all major pins barring a few exceptions like LPC lines.

[pjzz2000]

Hi,
Trying to get as much knowledge as possible so I understand what I'm seeing.

Most of the important power supplies "seem" to be present.

The ones where im seeing no output voltage at either side the inductors, do have 19v at their large input capacitors.

They mostly seem to be around peripherals that are not connected, display, keyboard etc etc.

Would it be reasonable to ask whether the power supply could be disabled, requiring a signal from the peripheral itself to enable it?

Or should the voltage be there independently?

[pjzz2000]

What an adventure I've had as a beginner.

Leg detached from EC chop, resoldered

One power controller chip exploded with a nice pyrotechnics display.

Used my new thermal camera to detect that the remains were shorted to ground and getting hot.

Dedoldered from board

Bought a semi working board for not much money so I could compare signals and power rails.

Was pretty sure in the end that either PCH or CPU was dead.

Tried CPU from bad mobo in good mobo and sure enough, CPU dead as a Dodo.

Trying to pluck up the courage to put good cpu in dead mobo to test, but worried will destroy it.

What do you think?