Board not working: blinking red PWR_LED

[Silica]

Hello everyone,

I recently got a problem with my A20-OLinuXino-Micro Rev. J.

Description:

When the the power jack is plugged to the board, the red PWR_LED
blinks endlessly, regardless of whether an SD card is plugged in or
not. There is no other device plugged to the board. The board is not
able to boot. The same problem has already been reported here:
https://www.olimex.com/forum/index.php?topic=3804.0

In my case, the red PWR_LED blinks at a rate of one pulse every 0.4 s,
with a slightly varying light intensity: sometimes the PWR_LED may
blink weakly, sometimes it would blink with a greater light intensity
over time.

When/how it happened:

I have bought the recommended power supply from the Olimex website. I
did not use any other power source for my A20-OLinuXino-Micro. I was
also using an SD card that has been pre-flashed with the official
image by Olimex.

When I received my board, I was able to boot up my OLinuXino from the
SD card for the first time, powered with the official power supply,
and plugged to an Ethernet cable to interact with it through SSH (no
HDMI, no keyboard). I was able to play with it for like 20 min, then I
left it idle for a dozen minutes, and when I came back the system had
frozen. It did not respond to ping through Ethernet. First I thought
the board entered sleep mode, so I tried the combination of buttons
specified in the manual to wake it up, but nothing

Since then, when I plug the PWR jack in, the best I can get is a
blinking PWR_LED.

Diagnostic:

This is not a boot or SD card issue, but most likely a power supply
circuit issue.

The specifications written on my power supply are:

SWITCHING ADAPTER
MODEL: SYS1421-0612-W2E
INPUT: 100-240V~0.5A MAX 50-60Hz
OUTPUT: +12V⎓​0.5A
OUTPUT POWER: 6W MAX.

The power supply delivers an almost perfect 12.04 V tension measured
with a multimeter, so I assume it is working properly (I am a bit
unsure about checking the current intensity).

I have also tried to power the board through USB-OTG, with the
strongest USB power source I could find (1.0 A). Using different
combinations of PWR jack and USB-OTG I got these results:

PWR jack only: PWR_LED blinks
USB-OTG only: no lights
PWR jack + USB-OTG: PWR_LED blinks with a stronger intensity,
sometimes the green CHGLED would randomly blink as well, synchronously
with the PWR_LED.

When USB-OTG is plugged in I can measure a straight 5.0 V tension in
the board (between USB V_BUS and USB GND).

When measuring between PWR jack and ground, I get an tension varying
between 2.80 V and 3.40 V following a sine-wave pattern with a very
long period of 10 s, which is actually so long that I can effectively
use my multimeter as an oscilloscope to observe the shape of the
electric signal.

Fortunately this board is open-hardware, so I have been able to
download the circuit schematics and proceed further with the
diagnostic. The problem is located somewhere in the power supply
circuit. I have let aside the AXP209(QFN48) [U12] chip for the moment,
and mainly focused my research around the MP1584EN(SOIC8E) [U14]
chip. I have measured the tensions I get on all the MP1584 pins when
the PWR jack plugged in:

Pin 1 (SW): Varies between 0.07 V and 0.64 V, proportional to VIN
Pin 2 (EN): 1.0 V (but should be > 1.5 V if I understand the MP1584 datasheet correctly?)
Pin 3 (COMP): Varies between 0.05 V and 0.50 V, proportional to VIN
Pin 4 (FB): Varies between 0.01 V and 0.10 V, proportional to VIN
Pin 5 (GND)
Pin 6 (FREQ): Varies between 0.00 V and 0.07 V, proportional to VIN
Pin 7 (VIN): Varies between 2.8 V and 3.4 V (same branch as PWR jack)
Pin 8 (BST): Varies between 0.3 V and 1.3 V, proportional to VIN

I have measured the resistance of all resistors connected to the
MP1584 chip. All of them are fine, except R108 (connected to MP1584
[U14] on the FREQ pin) which should be 100 kΩ according to the
specifications, but which is between 16.5 kΩ and 18.7 kΩ (the value
depends on the mood of the day). I think this is likely to be the
problem, and since it is connected to the FREQ pin it may explain the
shape of the electric signal. The MP1584EN chip may be dysfunctional
as well, but it is more difficult to diagnose an IC.

I have also cleaned my board with great love and passion using pure
isopropyl alcohol and a brush (following the recommendations of
LUB/Olimex), insisting especially on the R108, but did not result in
any noticeable change.

Conclusion:

I wouldn't mind spending a little bit of time and money to buy
soldering tools and to try to replace the R108 (even if it is unlikely
that I get it right the first time, I would still like to improve my
soldering skills anyway). However, if the board is covered by a
guarantee then I would prefer to get it replaced. Olimex may also be
interested to get the board intact so that they can investigate on
their side as well.

If you have any suggestion regarding a new non-destructive test I
could perform, I will consider doing it.

Have a nice day!

[JohnS]

I suggest contacting support as it may be quicker than hoping they see the message here.

John

[LubOlimex]

Hey,

Of course, we can replace anything that was faulty but it is important to determine what is the root of the problem to prevent further similar problems. However, based on your data I don't think the issue is in the board itself. This part is quite telling:

> Pin 7 (VIN): Varies between 2.8 V and 3.4 V (same branch as PWR jack)

VIN on U14 should show the voltage that you have applied to the power jack, this means 12V +/-5%. The whole input from the power jack goes straight to U14. Refer to the excerpt below that shows the connection between the power jack and the regulator: https://imgur.com/DbcKgz6

This leads me to believe either that the switching power adapter was/got faulty/damaged or that you are not measuring correctly. My advice is to measure the voltage available on the adapter while disconnected from the board - is it 12V? It can also be something like poor contact between the adapter and the power jack.

Best regards,
Lub/OLIMEX

[Silica]

I suggest contacting support as it may be quicker than hoping they see the message here.

Hi John, you are right indeed! Actually I have contacted the Olimex support first, and Lub suggested to start a forum thread.

My advice is to measure the voltage available on the adapter while disconnected from the board - is it 12V?

Thanks for your reply, Lub! When I plug the switching power adapter to a wall socket, and leave it free (i.e. disconnected from the board), I can measure the tension between the inner and outer metal contacts of the jack: I get an almost perfect and constant 12.04 V.

However, when I plug the power jack to the board, I get a tension periodically varying between 2.8 V and 3.4 V. I obtain the same tension by measuring at various points on the PWR - VIN branch: on the U14 VIN pin, on the PWR socket legs, and on the accessible metal plates on the PWR socket. I have tried with different ground points as well. I always observe the same behavior, so I am quite confident with this measure.

Do you think the switching adapter may be faulty because it is not able to deliver a 12 V tension while plugged to the board?

It can also be something like poor contact between the adapter and the power jack.

Regarding the possibility of a poor contact between the adapter and power jack, I can try a continuity test but I would need to disassemble the switching power adapter to get access to both + and - wires of the jack.

[JohnS]

Apologies - and I'm pleased to see Lub is on it!

John

[LubOlimex]

> Do you think the switching adapter may be faulty because it is not able to deliver a 12 V tension while plugged to the board?

Probably. And I hope this is the case. But it can also be a very faulty board that draws so much current from the adapter that the voltage from the adapter drops from 12V down to 3V

The options for moving forwards are:

1) Borrow or purchase another 12V adapter from elsewhere and test with it - this would clearly show if only the adapter is faulty. If you are going this path the important things about the adapter are the voltage and the amperage try with a more powerful one for exmaple 12V 1A one, the jack's dimensions can be found here: https://www.olimex.com/wiki/PWRJACK
2) Return both the board and the adapter to us so we can test it here. We will determine what is the problem and what caused it.

If you decide to return both the board and the adapter please drop an e-mail to support@olimex.com and include a link to this thread so we can provide a RMA number.

Best regards,
Lub/OLIMEX

[Silica]

I'm pleased to see Lub is on it!

Indeed, I feel like Lub is my hardware troubleshooting hero! <3

1) Borrow or purchase another 12V adapter from elsewhere and test with it

Sounds like a plan! Unfortunately I have not been able to find a suitable 12 V adapter, so I will have to buy one. I will post again in a few days as soon as I get new results.

[Silica]

I have bought a 12 V adapter, now the board seems to work fine. So in the end the problem is coming from a faulty switching power adapter. I will mail the support to get it replaced.

At least today I have learned that testing a switching power adapter with a voltmeter is not enough to tell whether it is working properly or not.

And I still wonder about this R108. I still find it odd that the resistance value I measure does not match the specifications. But anyway.

Problem solved!