DAX1BDMB6F0 Power Issue

[piernov]

Re: DAX1BDMB6F0 Power Issue

If the battery is known good and charged, it should boot if you don't plug in the AC adapter. If AC adapter is plugged in and there is an issue with the center pin, it may not work.
Btw I've got a board that won't boot (even on battery only) if I don't pull up AD_ID to B+ with a 2k (IIRC) resistor. Couldn't figure it out, replaced the EC and same. So weird stuff can happen with this center pin identification.

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hello,

Got the new adapter today, plugged it in and the computer started right away. Turned it off and back on right away. The third time, pushed the power button and it blinked once, pushed the power button again, it blinked again and started on the third push of the power button. The fourth time was the same as the third.

I measured the voltage on the inside of the adapter and it is 19.9V. The pin voltage is 12.5V. I gather there has to be some reason for the lower voltage on the center pin that is logical.

My battery is fully charged.

I guess I will have to live with that quirk until it acts up again and does not want to start.

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hello - you will have to excuse me but the center pin voltage is not 12.5V but actually 19.9V. I must have had a bad connection when I took the reading the first time.

Also guess what, I left the computer be with the new charger plugged in for a while and it is now completely unresponsive no matter what I do.

Back to square one...

[mcplslg123]

Re: DAX1BDMB6F0 Power Issue

Look for IO thermal protection on page 35 of schematic nad measure voltage on resistors. I faced this issue on a quanta U83 motherboard .

[BlueMidnight]

Re: DAX1BDMB6F0 Power Issue

Btw I've got a board that won't boot (even on battery only) if I don't pull up AD_ID to B+ with a 2k (IIRC) resistor. Couldn't figure it out, replaced the EC and same. So weird stuff can happen with this center pin identification.

That sounds like a fun little problem. Though I would consider that more of a phantom adapter detection than a smart pin problem, as the latter seems to be working as intended. There might have been a shorted reverse blocking FET or bad charge IC, etc. making an AC adapter appear to be present.

The adapter resistors seem to start around 500k for the 45W (my measurement), 390k for the 60W (also my measurement), and 200k-300k for the 90W (from internet sources).

Using a 2k pull-up should work, but would be constantly forward biasing the protection diode to clamp the ID pin voltage at ~3V. The lower half of the resistor divider (the one on the motherboard) always seems to be 12.5k.

Look for IO thermal protection on page 35 of schematic nad measure voltage on resistors.

That's not a bad idea. Measurements can be done at the EC (iTE chip). The relevant pins have "THRM" in their names. H_PROCHOT# is interesting as well.

And you don't have to abandon taking measurements when the board is temporarily working. Whether it's something like the thermal protection, RSMRST#, or one of the other signals, what you are looking for is something that has a different reading when the board wants to power up compared to when it doesn't. And comparing readings you've taken under various circumstances is certainly one way to spot something useful.

[...] the center pin voltage is not 12.5V but actually 19.9V.

The center pin is connected to 19V via a resistor inside the AC adapter. If you just measure it while it's not connected to anything there will be no voltage drop across the resistor, and you should measure the full adapter voltage.

However, things change once you plug it into the motherboard. The simplified circuit looks as follows:

19V --> Resistor (in AC adapter) --> Center pin --> Resistor (on motherboard) --> Ground

This creates a voltage divider that will place a much smaller voltage than 19V on the center pin. By having different resistors in different wattage AC adapters, the center pin will take on different voltages. The board will read this voltage and know the wattage of the adapter.

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hello,

The adapter is 19.5V, 2.31A, and 45W

From what you said, if the center pin voltage is altered because of the two resistors as per your diagram, can the resulting voltage be measured, that is of course while the adapter is plugged into the board?

Also, not quite sure what "using a 2K pull-up" means.

I will take the motherboard out and check the voltage on the ITE again at the THRM (& Prochot) pins.

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hello again,

I measured the THRM Pins with the motherboard still unresponsive (no charging light or anything...)

Pin 70: 2.97V
Pin 71: 3.63V
Pin 72: 2.93V

Also measured:

Bat_I pin 66: 3.61V
Ad_Type pin 67: 4.00V
Sys_I pin 68: 1.79V

Also measured:

TempMBAT pin 120: 2.67V
H_PROCHOT#EC pin 124: 0V, that is zero volts
DGPUPROCHOT# pin 77: 0.22V this pin did not go anywhere so it was difficult to take the measurement...

[BlueMidnight]

Re: DAX1BDMB6F0 Power Issue

First, I have to make a correction... I somehow had the unpopulated diode stuck in my brain. It's actually the zener that clamps the smart pin circuit at 5.6V. Which makes sense, as 6V is a the max voltage for many I/O pins on these boards.

And unfortunately, I dug around the H_PROCHOT# circuit and found it's pulled up by the "+1.0V" rail. Since that doesn't turn on until later, the 0V is expected.

Yeah, DGPU_PROCHOT# seems to be only for boards with a discrete GPU.

It's worth measuring LID_EC# (pin 123) as well. It should be 3V, otherwise it thinks the lid is closed.


What do look weird are BAT_I and SYS_I. The charge chip takes the voltage difference across PR177 or PR156 (both 0.01 ohm resistors), multiplies them by some number and places that many volts on the appropriate "_I" pin. This can be used to calculate the current through those resistors. But I keep calculating currents that are really high.

For SYS_I, it can be the voltage across PR156 multiplied by 20 or 40 (configured by the firmware). The lowest case would be x40, so 1.79 / 40 = 0.04475V. So there is at least 0.04475V across PR156. That divided by 0.01 ohms = 4.475 amps coming from the AC adapter.

For BAT_I, it can be the voltage across PR177 multiplied by 8 or 16 (configured by the firmware). The lowest case would be x16, so 3.61 / 16 = 0.225625V. So there is at least 0.225625V across PR177. That divided by 0.01 ohms = 22.5625 amps coming from the battery.

Those readings are absurdly high. Even if they were an order of magnitude lower, they would still be too high for a board that is off. It would be interesting to see if they are different right before you are able to power up successfully vs. when it won't power up.

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hello,

When I plug in the adapter with the center pin (charging light off) the brown wire coming into the board has 19.9V. The yellow and white wires both have 5V. When I plug in the adapter without the center pin, the brown wire read 0V and both the yellow and white wires have 5V.

Before the tests above, I plugged the adapter with the center pin and the charging light came on, so I made some measurements:

Bat_I pin 66: 0.81V
AD_Type pin 67: 5.57V
SYS_I pin 68: 0.84V

So both Bat_I and SYS_I are lower whereas AD_Type is higher. The board does not seem to want to power up anymore so I could not take those pin measurements while powered up.

Pin 123: 3.22V

I then pushed the power button and the charging light went off and the board went unresponsive.

[BlueMidnight]

Re: DAX1BDMB6F0 Power Issue

Well the BAT_I and SYS_I readings were much better that time at least. The AD_TYPE voltage is getting clamped by the zener diode. It might not be very good for the EC if the zener decided to fail open circuit one day. Along with the slightly high 19.9V, it seems the new adapter has an ID resistor that is too small. But it doesn't really affect the present problem..

Just to clarify, were the measurements in this post and your last post done using the same AC adapter?

The brown wire is for the center pin. The white and yellow ones are for the 2 LED colors.

Another EC signal to check:
DSWROK_EC (pin 87)

Do keep an eye on that one and also RSMRST# (pin 119) and note if they are different right before a successful power up attempt vs. an unsuccessful one.

[piyal]

Re: DAX1BDMB6F0 Power Issue

thanks

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hello,

I think it the same power supply but I can't say for sure...

For the Pin readings, I plugged the power supply to the card and the charging light went on. Here are the measurements:

Pin 87: 0V
Pin 119: 0V

Then I pressed the power button and the charging light went off. Here are the measurements:

Pin 87: 0V
Pin 119: 0V

I haven't got a successful power up for a while now with the new charger.

& by the way, pin 87 does not go anywhere, unless it goes underneath (i.e., to the other side of the board) and I can't see it.

Thank you

[BlueMidnight]

Re: DAX1BDMB6F0 Power Issue

For the center pin voltage to change with the same adapter would be hard to explain. If it's with the same adapter, it's worth examining further. But if the 4V reading was with the original adapter, then that's fine.

It's convenient when traces come off of the pins where you can see them. But they can just as easily go the other direction, under the chip. They can even go straight down into a lower layer under the pad the pin is soldered to.

Without some new information/behavior, we can't be too specific, but the problem still seems to be centered around the EC. The randomness usually points to a hardware problem, but the firmware can't be ruled out.

If I had to choose a course of action I would probably say to resolder the EC. It's pretty low risk if you have soldered dense pins like those before, and if you have a blade tip for your iron (to help prevent solder bridges).

Flashing the EC firmware requires having the right programmer since the memory is on the EC itself. And replacing the EC normally requires flashing it as well (unless you can find ones that are preprogrammed for this exact board).

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hi - I have a Weller soldering gun and tried to resolder the EC. I think I messed up as looking under the microscope it doesn't look to good...

I went on eBay and bought a new board; that was a dent in the wallet but hey, the computer is nearly $1K so I guess it is worth it.

I will let you know how that turned out...

Thanks for all your help BlueMidnight, I learned a lot.

[BlueMidnight]

Re: DAX1BDMB6F0 Power Issue

You're welcome. At least you have the old board for soldering practice...

Speaking of which... you should have a nice acute blade-like tip on the iron to solder chips like those (if you don't already). Then you can place it across multiple pins and slide it down parallel to the pins until it slides of the ends. Continue soldering groups of pins like this until all pins have been soldered.

The surface tension of the solder makes it pool up against the tip. If the tip is not acute enough (like the common conical or chisel tips), too much solder will be around the tip just as you are pulling it away from the pins. This causes solder bridges across adjacent pins.

Also, using flux is a must. The main function of flux is to clean surface corrosion and allow the solder to stick properly. And the more it wants to stick to the pins, the less it will want to hang out in the gap between them.

With the right tip and some flux, soldering a chip like that should take seconds. And there shouldn't be any solder bridges you need to clean up afterward. If one wants to form, you just clean the excess solder off your tip and give the pins another pass.

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Yes, I will surely need the practice.

One thing I notice though that made it difficult to solder was a layer of wax or epoxy on top of everything and that was probably why I came in with too hot of an iron for too long. I wonder then if there are products to disolve that without damaging the board.

Otherwise I will look into getting more accurate blades. I see they come in many different widths. The EC chip is something like 15.5 mm wide across the pins and it is probably a standard width.

[BlueMidnight]

Re: DAX1BDMB6F0 Power Issue

There normally isn't any sort of coating on these other than some fine corrosion on the pins, and flux takes care of that. Anything else can usually be removed with some isopropyl alcohol .

A tip with this kind of shape that is about 5mm wide or so seems to work well (though obviously you want to find tips somewhere that fit your specific soldering iron). There's no need to do all the pins on a big chip in one pass with a super wide tip. Not that that wouldn't work necessarily. But it's a lot of area to keep an eye on at one time. You don't want to accidentally wipe some nearby small components off the board by mistake.

There are also really large tips like this one that are perfect for cleaning the pads under a BGA chip (by sliding it around with some desoldering braid beneath it).

And the best general purpose tips for motherboard work are ones with a shape like this.

Also, soldering irons with the heating element and a temperature sensor built right into the tips work best. Otherwise, you constantly have to adjust the temperature to compensate for the size of the tip and the thermal mass of the area you are working on.

[Pierre-Yves]

Re: DAX1BDMB6F0 Power Issue

Hi & thanks for those "tips"...

Ordered a soldering iron with a temperature sensor, a desoldering braid and an ITE chip; hopefully will get these within a week or so.

Can't believe how inexpensive the ITE chip was ($4 to $5). I guess if you know what you are doing, are equiped (& skilled), you can save a lot of money.

[BlueMidnight]

Re: DAX1BDMB6F0 Power Issue

Removing a chip like that without a hot air station is a little tricky. Getting enough solder off the pins to be able to lift it cleanly off the board (without bending pins) is hard. Although, if you don't need the old chip, you can remove it in any destructive way you wish that won't harm the actual PCB.

Now the new chip.. that has its own problems. As I said, this particular one has flash memory inside it. So you need a chip that is preprogrammed with the correct firmware for that board (and possibly compatible with that board's current system firmware/BIOS/UEFI version as well), or you need a hardware programmer compatible with those chips (like the SVOD or the RT809F with the optional PEB-1 expansion board) along with a good copy of the firmware itself to use with it.

[piernov]

Re: DAX1BDMB6F0 Power Issue

On some laptops the ITE chip gets programmed automatically when you first apply power if it's blank. So there's a small chance he won't need an external programmer.