TESmart HDMI Matrix - blinking power

[CaliMiller]

TESmart 8x8 HDMI Matrix
Refurb unit purchased in 2021, now out of warranty.

Amazon.com

I was using it daily in my studio, outputting to 3 panels. I left the matrix run when I wasn't using it, only powering off the panels, and putting the input devices in standby. It worked fine for 2 years. Then one day while I was using it with my computer, the image went blank on all panels, and the blue LCD panel on the Matrix started blinking on and off.

I tried turning it off and back on.
I tried letting it sit to cool down and tried it later. Same symptoms.

I figured maybe that the issue was the caps. I opened it up, and to my surprise, there were very few caps, and from what I could see, they all looked good. I don't see anything burnt, nothing smells burnt. I turned it on when I opened the case, and touched the caps and the chips (except the ones under the heatsink) to see if any were abnormally warm. Nothing seemed out of the ordinary. There was no dust in the case.

I had it sitting on a shelf, close to a wall (not the greatest ventilation) with an Xbox and a Switch on top. My guess is that it overheated and something got fried.

I've done some minor soldering work in the past, but nothing that involves microsoldering. I was Apple Certified to work on iPhones... but I'm kind of a noob when it comes to troubleshooting circuit boards. I have a cheap multimeter, and a few tools from iFixit. I would appreciate all the help offered, explanations and tutorials are welcome.

 

[Delta Prime]

Did they refurbish the power supply adapter as well? Probably the original was passed on to you? I would think that would be the culprit. 2 years is a long time. That's what happened to mine...
But I kept mine in the basement it had real bad dust bunnies and stuff living in it.

 

[73's de Edd]

Ms CaliMiller . . . . .

As is being revealed in your photo #1289 . . . .
Bottom left corner is main power 12VDC input and its associated E-cap filter . . . .it seems to be of a bluuuuuuuurrrrrrry ¿ 100 ufd ? value
It is having meter probe access to its top right side - tab and left side + tab, so . . . . . . vewy-vewy cawefully*** . . . . . . . place meter probes to those points and be expecting ~ 12VDC. Then lift off probes and place your meter in AC mode and retest to see what, if any, developed loads ripple voltage might be present.

Now, look upwards to see 4 clusters of DC-DC voltage converters with each using a controller / driver I.C. , driving into a 4.7 uh ringing inductor, for its back EMF voltage, it then being harvested by a schottky diode rectifier and finally storing that accumulated voltage in another bluuurry single ¿ 100 ? ufd E cap.
All 4 of those separate ¿ 100ufd ? E-caps are having adequate probe access to their upper side - tab and bottom side + solder tabs. Run the same DC voltage and AC ripple testing on each of those 4 sub supplies at their 100 ufd caps.

We are looking for an inoperative supply or one with time developed high AC ripple content .

I'm suspecting those sub-supplies to be for generating 5VDC . . . . . 3.3VDC . . . . . . and 1.3 or 1.8VDC . . . . . . . and the one on the far right for a supplemental +5VDC audio or digital supply.
Any failures / deficiencies in those power supply areas, could thereby account for your flashing LED indications.

NOW DON'T BE A BUMBLEFINGERS . . . AND THUSLY . . . LET YOUR METER PROBES SLIP . . . . . .OR . . . . . .BE FULLY PREPARED FOR A FATE WORSE THAN DEATH ! ! !

( On my meter, that is why I made 2 slip on probe extensions with actual needles soldered on their ends . . . . they will then pierce solder blobs . . . .and WILL THEN lock in that DEVELOPED divot FIERCELY, until you lift them out.
I additionally heat shrinked on conforming Red and Black insulative covers to externally insulate them . . . . . leaving only ~ 1/16th in of the needle tips peeking out..)

With ONLY you being able to see the mini 6 pin I.C.'s . . . pass the part numbers on to us so we can consult their data /application sheets . . . .if so being, further required.

*** a Fudd . . .'sterism

Thaaaaaaassssssssit . . . . . .

73's de Edd . . . . .

CAVEAT . . . . . before you give a colleague a piece of your mind, FIRSTLY, just be sure that you can spare it.

.

 

[bidrohini]

I suggest you perform capacitor test, short circuit test etc.

 

[CaliMiller]

Thank you all for the replies. I will test the caps when I get some time.

Did they refurbish the power supply adapter as well? Probably the original was passed on to you? I would think that would be the culprit. 2 years is a long time. That's what happened to mine...
But I kept mine in the basement it had real bad dust bunnies and stuff living in it.

Not sure which parts were refurbished.

Ms CaliMiller . . . . .

As is being revealed in your photo #1289 . . . .
Bottom left corner is main power 12VDC input and its associated E-cap filter . . . .it seems to be of a bluuuuuuuurrrrrrry ¿ 100 ufd ? value

Now, look upwards to see 4 clusters of DC-DC voltage converters with each using a controller / driver I.C. , driving into a 4.7 uh ringing inductor, for its back EMF voltage, it then being harvested by a schottky diode rectifier and finally storing that accumulated voltage in another bluuurry single ¿ 100 ? ufd E cap.
All 4 of those separate ¿ 100ufd ? E-caps are having adequate probe access to their upper side - tab and bottom side + solder tabs. Run the same DC voltage and AC ripple testing on each of those 4 sub supplies at their 100 ufd caps.

Just want to confirm that these are what you are referring to?

It looks like the text on all of the caps reads "100 16V UT"

 

[73's de Edd]

Just want to confirm that these are what you are referring to?
Yep . . . . . ABSOLUTELY !
Are you reading the U1,2,3 and 4 controllers marking code . . . . Eye See's . . . . as being an AGMJ ?
With the benefit of your now more detailed blow up . . . I am seeing no Schottky diodes, so they must be incorporated within the I.C.'s.
That AGMJ code . . .if I'm reading it correctly . . . . is not being found . . . unfortunately . . . . any reader up in the peanut gallery have any ideas ?
But . . . MAINLY . . . get back to us with those found voltages.

Be Excellent To Each Other . . . . . . . .Oh Tay . . . . . . but would NICE , equally suffice ?

 

[73's de Edd]

any reader up in the peanut gallery have any ideas ?

O.K . . . I FOUNT IT ! . . . . . it just required a slightly different search function definition term.
And then coming up with a MP1471A of 2012 design vintage . . . .now degraded against further use of it . . . .so roll out the MP1477A.
Looks like it is just dependent upon toggling the R2 divider resistor to acquire the different 5-3.3-1.2/1.8v output voltages from that 12VDC input.
If being just a support discrete failure . . .most likely would me the C3 . . . Bang-Bang*** . . . . 1 UFD charge up capacitor . . . . .although typically, those ceramics are QUITE RELIABLE !

HERE 'Tis . . . . .brace yourself . . . INCOMING !



That's switching at *** 500,000 Bang-Bang's per second . . . . . . OOOOOOOOOOHHHHHHH . . . .BABYCAKES !

73's de Edd . . . . .



.

 

[Delta Prime]

any reader up in the peanut gallery have any ideas ?

Always a pleasure to watch you work.
I learn immensely from you.

 

[73's de Edd]

¿ Is Takuro a work associate / acquaintance of yours ?

 

[bidrohini]

Check if the power supply unit (PSU) is providing the correct voltage and is not faulty. Use your multimeter to measure the voltage output from the power supply to verify it matches the rated voltage of the matrix unit.

 

[Delta Prime]

¿ Is Takuro a work associate / acquaintance of yours ?

He's wearing my overalls! “Kao wo tateru”
We met while attending a mandatory meteorology training course where I am
employed at the California institute technology for proprietary instrumentation calibration.
His overalls were a misplaced I offered mine. He barely benefits in them. I was on the night shift so no biggie. His family asked about it, after the media took a snapshot of him. Crazy thing is they wouldn't stop feeding me and I didn't want to offend them, I gained a couple pounds...
And a good friend!