Leaf boot loop after deleting a system app

[frankbonatelli]

Originally Posted by Renate

So, what do you have for experience/tools?
Voltmeter (digital)?
Resistors (220 ohm to 2200 ohm)?
USB UART with 1.8V logic level?

That's an easy question to answer...… Sadly no experience so no real tools that would make you happy.

Originally Posted by Renate

A better photo of just the circuit board would help.

Hopefully this with tape removed will display the pins needed.

[Renate]

I'm drawing a blank.
I can identify many of the test points, but I don't see anything unless it's close to the battery connector.
This is a shot of my Poke3 (same SDM636) to show you how obvious they made it.
The glass reed switch is my addition. The UART is on the (also added) white connector.

You can try disconnecting the battery and reconnecting, see if anything changes.
That's on a "mezzanine" connector. You can pull it with your fingernails on each side (knuckles against the board). Pull evenly on both sides.

While the battery is disconnected, you might look under the Eink panel connector.
Manufacturers try to be sneaky sometimes.
It's a mezzanine connector too.

If your device WAS working it would be easy to find the EDL points. Since it's the same SDM636 I know that it's GPIO57 and even have a program for detecting it.

I think that you have to get a resistor or two. Got any neighbors?
If you had a 1000 ohm and a 1.5V battery you could just test all the points.

P.S. If you look under that Eink flex don't bend it too much.

P.P.S. You're really, really sure that holding one of the buttons while resetting doesn't get you to fastboot?
My Poke3 doesn't have any buttons.

[frankbonatelli]

So I've popped the battery connection and photographed the connections a little better. There are three more hidden under the connector.

I've dug out two different volt meters, not sure which if either will work. Any thoughts on if either will do the needed task?

[Renate]

I've updated the Leaf1-2.png (the "map" diagram).
There's the test points near the LED connector marked "Danger".
Don't mess with them at all. (It's only ~20V but it could burn something up or burn itself out).

Stage one:

Get a piece of paper to keep track of your results.
With the device plugged into USB and the black probe measuring off that nice bit of tin on the WiFi/BT module measure some voltages.

With your DVM (the yellow looks nice) in DC volts mode:
You're looking for one spot measuring 1.8V and the adjacent spot measuring near 0 (like <0.1).
Mark them down on paper and notate which were zero volts.
If something measures over 1.8V, like 3.3 or 5V it's nothing that you want.
When you're done, did you get any suspicions?

Step two:

Time to try stuff. Put your meter in the diode mode (blue triangle bar)
Black to tin.
Now you pick a likely zero voltage point and measure.
If it says 0.0 (or close) it's just a ground point.
If it say OL, it's probably not what you want. (but might be worth checking later if you get nothing)
If it says numbers, that's a possibility.
With the probe held on the spot, do a long >10 second power push and see if something shows up on USB, like 9008.

Theory: The diode test puts out a safe voltage and current which we can use to trigger the EDL (if it's the right pin). Except for that LED area you're not going to break anything.

Just to reiterate:
You can measure voltage on any (non-LED) point in stage one.
Only do diode test in stage 2 on things that you've already measured as zero-ish.

[frankbonatelli]

Originally Posted by Renate

I'm drawing a blank.

P.P.S. You're really, really sure that holding one of the buttons while resetting doesn't get you to fastboot?
My Poke3 doesn't have any buttons.

Sorry I missed this question earlier....
As to that i am not 100% confident that the case buttons actually do anything. Ive tried up reboot down reboot and both reboot. nothing seemed to work except the timed power button series of hold and clicks. I reached out to boox and they sent me a video for getting into recovery
Onyx Boox Leaf boot to recovery with just power button

[frankbonatelli]

Originally Posted by Renate

I've updated the Leaf1-2.png (the "map" diagram).
There's the test points near the LED connector marked "Danger".
Don't mess with them at all. (It's only ~20V but it could burn something up or burn itself out).

Stage one:

Get a piece of paper to keep track of your results.
With the device plugged into USB and the black probe measuring off that nice bit of tin on the WiFi/BT module measure some voltages.

With your DVM (the yellow looks nice) in DC volts mode:
You're looking for one spot measuring 1.8V and the adjacent spot measuring near 0 (like <0.1).
Mark them down on paper and notate which were zero volts.
If something measures over 1.8V, like 3.3 or 5V it's nothing that you want.
When you're done, did you get any suspicions?

Step two:

Time to try stuff. Put your meter in the diode mode (blue triangle bar)
Black to tin.
Now you pick a likely zero voltage point and measure.
If it says 0.0 (or close) it's just a ground point.
If it say OL, it's probably not what you want. (but might be worth checking later if you get nothing)
If it says numbers, that's a possibility.
With the probe held on the spot, do a long >10 second power push and see if something shows up on USB, like 9008.

Theory: The diode test puts out a safe voltage and current which we can use to trigger the EDL (if it's the right pin). Except for that LED area you're not going to break anything.

Just to reiterate:
You can measure voltage on any (non-LED) point in stage one.
Only do diode test in stage 2 on things that you've already measured as zero-ish.

Thanks ill test this tomorrow and report my findings. I'm kinda excited to see if i have what it takes!

[frankbonatelli]

I've found two more underneath the battery ribbon
Question
1) can this be tested with out battery connected? Drawing power from USB C port
2) if not how far can i bend the ribbon and keep it connected?

-----

With the probe held on the spot, do a long >10 second power push and see if something shows up on USB, like 9008.

also for clarity when its time to try stuff, holding down the power button and touching red probe to the dot is the black probe still to make contact as well?

[Renate]

No, this may not boot at all without the battery.
It's better to check first what you can easily check.

The test points are there to test the main board without anything external connected.
That's why most of the test points are adjacent to connectors.
You have to test the reaction to the power button so there are test points there for the contacts and any little indicator.

If all this doesn't work we'll have to proceed otherwise.
As I said, there is a stage one, a step two and a procedure C

[Renate]

Originally Posted by frankbonatelli

Also for clarity when its time to try stuff, holding down the power button and touching red probe to the dot is the black probe still to make contact as well?

Yes, you always need two probes for anything.
If you have any alligator clips you can clip the ground to the black.
Since you'll have to be pushing the power button you'll want somehow to have the black connected without using a hand.

[Renate]

Mmm, shall I presume that you haven't had any success yet?

[frankbonatelli]

Sorry for the late reply, work gets in the way of play!
I've attached a list of my findings. I've done the tests four times now and the results are similar, none of which stand out as what I am looking for.

I'll keep trying till the magic smoke leaks out tho.
Thanks as always for everyone's efforts in this.

[Renate]

Originally Posted by frankbonatelli

Sorry for the late reply, work gets in the way of play!

OMG! Nobody can call you a slacker!
Your diagram is a work of art.

You got me confused though.
You were only supposed to diode check the testpoints that were about zero (<0.1V)

What is OL and 86.7 supposed to mean?
Is 86.7 showing millivolts? Is OL a voltage or the diode test?
I don't understand what I'm looking at.

On the ones that did a reset did they pop up as 900e again?

It looks like you only did a real test on 3-4 places?

I'm sorry if I caused you so much work.

Edit: You've got two voltmeters. You can cross check them.
Connect the black to black and the red to red.
Put one on diode test and look at the voltage with the other one.
Then do it the other way around.

When the yellow is in diode test and the "pen" reads 2.5V then we say that the yellow open circuit diode test voltage is 2.5V

[frankbonatelli]

OMG! Nobody can call you a slacker!
Your diagram is a work of art.

You got me confused though.
You were only supposed to diode check the testpoints that were about zero (<0.1V)

Originally Posted by Renate

What is OL and 86.7 supposed to mean?
Is 86.7 showing millivolts? Is OL a voltage or the diode test?
I don't understand what I'm looking at.

so OL is what is displayed when doing the second part of your suggested procedures, the "Step two: Time to try stuff. 86.7 is millivolts and only shows @test point 1 on my map and 12. These do fluctuate. Also worth a note that they drift down to a zero state if held long enough.

On the ones that did a reset did they pop up as 900e again?

Yes the the number 13 test is the only one that consistently triggers an immediate power off. Power off only. If I hold the power button before i make contact it will not reboot. If i contact it with out the power button push and hold it does not reboot. In either state I need to only hold the power button a count of 7 seconds to turn on, I surmise that this activates go to sleep type of call. And yes they all reboot into 900E

It looks like you only did a real test on 3-4 places?

I did both your prescribed tests on all 33 points 4 times twice in ascending order twice in descending order. I can see my mind map looks a little un structured, That's an old mans life time struggle with dyslexia and I apologies for the mess. There are two columns one on the left hand side of page and one on right hand side of page. the one on the left is Step Two first step one Second. The one on the right is step one first and step two second. I've also done random re re testes to be sure and eliminate the human error. all results are the same plus/minus a point or two.

Edit: You've got two voltmeters. You can cross check them.
Connect the black to black and the red to red.
Put one on diode test and look at the voltage with the other one.
Then do it the other way around.

!! now why didn't I think of this!! I will test this out tonight when i get home! Dam how goofy the brain works, this could be another good test.

Thanks again for all your patience and assistance. I've already order the onyx boox Nova air C as I need to read!! This however is a battle that I will happily fight till a binary 1 or 0 occurs!

[Renate]

Originally Posted by frankbonatelli

I did both your prescribed tests on all 33 points 4 times...

Wait. You did 33 points times 4 times 10 second power press?
I just calculated that out. It equals 1.07 fortnights. (Roughly.)
And you did the power press on each one???

Frank is awarded the Croix de Guerre.

I should have mailed you a couple of resistors a week ago.

[frankbonatelli]

I love math!
It is the only true universal language!!