Solved AC86U USB not reading 5 volts

[blitzkrieg]

Hi all

Recently managed to revive my AC86U by replacing MT3125.
It was working well for 5mths~ till I noticed the USB storage isn't detecting and my connected 5V fan is spinning 'louder'.
Tested the USB voltage (both 2.0 & 3.0) and it is not reading 5V, more like 10V~, and obviously fried my storage .
Apart from that, everything else is working OK.

Anybody know which component(s) do I replace/resolder for the USB 5V? (Yes I'm aware AC are EOL, but I'll like to keep it longer)

 

[blitzkrieg]

Upon further probing, both USB2 and USB3 is reading 18~V!

Think i found the culprit, with a burnt mark just above the USB2. Anybody knows what is 'D2' component?

 

[Tech9]

I don't know the number, but this looks like SOT-143 package Transient Voltage Suppressor (TVS) Diode (Dxx on the PCB) used for circuit protection.

Something like this:

https://www.bourns.com/pdfs/CD143A-SR05_12.pdf

Blown means this router was hit by high voltage at some point. 18V on the USB port means you have to check upstream step-down convertor as well.

 

[blitzkrieg]

I don't know the number, but this looks like SOT-143 package Transient Voltage Suppressor (TVS) Diode (Dxx on the PCB) used for circuit protection.

Something like this:

https://www.bourns.com/pdfs/CD143A-SR05_12.pdf

Blown means this router was hit by high voltage at some point. 18V on the USB port means you have to check upstream step-down convertor as well.

Cool, i'll try to get the TVS in 5V, nothing to loose.
Anyways do you know which component can i probe for the step-down convertor? I don't really have an electronics background.
Thanks!

 

[Tech9]

I don't really have an electronics background.

You won't be able to fix it yourself then. There are multiple VRMs on this PCB, you can follow the traces to find which one is involved, but ICs are SMD type soldered with lead-free RoHS solder with around 220C melting point. Some are QFN packages with 16 leads underneath in 3x3mm size. This means you have to have skills and equipment. The PCB has to be preheated, components around shielded with Kapton tape, IC removed with temperature and flow controlled hot air gun, the contact surfaces have to be cleaned with disordering wick and restored with soldering paste, etc. You may need microscope camera. There is barbarian way as well, but with high risk of damaging the traces.

 

[sfx2000]

You won't be able to fix it yourself then

Hey, let them do what they want to do...

 

[wouterv]

I do have an electronics background from ancient times and won't try to repair the router myself....but from my last 20 years work experience I do know some electricians on e.g. offshore ships can do magic and surprise even the original manufacturers by the repairs they manage to do to outdated PLC cards or other professional electronics (often by making one working card out of 2 or 3 failed cards). Magical fixes they are often kind of forced to make due to lack of support (and lack of investments by the ship owner).
So please support and encourage anyone willing to repair their router.

 

[Tech9]

So please support and encourage anyone willing to repair their router.

With no knowledge and proper tools attempting repair will cause permanent damage. I would give it to repair shop instead. For this specific model repair doesn’t make much sense. It’s EoL and has multiple hardware failure points plus unresolved software issues.

 

[L&LD]

Learning has no price. No sense is what you spread, continuously. And markedly opposite of what anyone else may state, or want.

 

[thiggins]

Learning has no price. No sense is what you spread, continuously. And markedly opposite of what anyone else may state, or want.

Uncalled for and unhelpful. Refrain from further attacks or you will get a timeout. I've just about had it with your attitude.

 

[blitzkrieg]

Hi again
I bought a CD143A (5V) and resoldered it, obviously it fried again upon power on.
So, i traced the REF pin (4) of the CD143A onto pin 3 (Output?) of IT76630M Step-down converter.
Decided to buy and replace the IT76640M as well, but the TVS got burnt again too! (pin 3 of IT76640M is reading 18V~).
Anybody knows what/how else can I trace the culprit to? Attached photo is what i found; red outline is connected, including the USB power and 16V cap. While green is pin3 and and 2 respectively of the TVS.

 

[Tech9]

If the output of the step-down converter reads 18V even after replacement - focus on the components around this IC, use reference design and compare values. With the IC removed I would measure all the capacitors for short first. It's very common for SMD capacitors to short when they go bad.

 

[blitzkrieg]

With the IC removed I would measure all the capacitors for short first

Ah yes, I actually tested the ceramic caps but none were shorted.
I re-traced and tested resistance for the components around IT76630M as attached.
My multimeter doesn't have capacitance mode, and at this point I'm learning as I go along. All caps are grounded except for the one connecting pin 1 to 3. Reading the IT76640M schematics I think R1 connected to pin 5 is the culprit, supposed to be 44.2kΩ (page 11), but I'm reading 8.36kΩ. And the resistor is smaller than the IC legs! Anybody know how to decode IT76640M?
*Edit: or, anyone willing with an AC86U can verify the resistors values?

 

[wouterv]

Checking electronic components while they are soldered in place is a challenge: for sure you will never be able to exactly determine their nominal resistance or capacitance because the way they are part of the schematic changes the value you will measure.
Also applying the voltage of your multi meter in resistance reading mode to the circuit board can cause damages to e.g. integrated circuits.
Resistors and ceramic capacitors will hardly ever fail while electrolytic capacitors (like the shown 16V / 100 uF) are very well known to degrade and fail - but again that is hard to determine while still soldered in place.
When a resistor fails it usually shows burning marks and there is often an underlying cause for the damage.

 

[blitzkrieg]

I got around into unsoldering the CD143A (3rd time) and IT76630M, remeasured the resistors; it is of the correct values for R1 & R2, 45.2kΩ & 9.84kΩ respectively!
Then i resoldered another sample of IT76630M, and it now outputs 5V (pin3)! FeedBack (pin5) correctly reads 0.925V instead of 3.24V! The 16V cap is also reading 5V~ I might have shorted the IC earlier in resistance mode then?
Anyways, both USB is getting 5V BUT only USB3 data works/detect; i think the data line for USB2 got fried from the beginning, I'll revisit that in future.

I'll call this repair a 90% success and learnt quite a lot, with no electronics background! Instead of dumping as e-waste and 'just buy a new router'. I've also changed all 4 electrolytic caps to Nichicon 'audio grade' i had as spares; does it makes a difference, dont think so but definitely better than factory. Oh, i needed to drill the casing to accommodate the taller caps

 

[wouterv]

As longs as the capacitance (uF) is equal and the nominal voltage (V) is equal or higher, you can replace capacitors.
I do somehow like your practical approach