Fish Tank LED Power Supply Gets Hot then shuts down

[stj]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

it does not do esr, it's a basic compact clamp meter with a 100A limit.
it's fame is that it can read DC with the clamp with a 1mA resolution!

[CapLeaker]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

Nope. ESR meters aka LCR meters are their own, and these functions are not to find in a DMM.

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

No, I'm curious why it's drawing 28W on the 5V line

i think you need to check the current on the 12v line.

it could be mostly using 5v,
those old AT psu's had mostly 5v output, the 12v was only for the harddrives.
in that case an ATX is useless because their main output is 12v

OK, I cracked open the LED panel itself, which turned out to be only 4 screws then slide out the plastic that covers the LEDs and it exposes everything ... as you can see... I saw a problem immediately... You'll have to zoom in on the second picture to see it but the plastic on whatever that DC power connection is, is totally fried off of the wire though the wire is not disconnected.

[petehall347]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

looks like you found the very slow blow fuse .

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

looks like you found the very slow blow fuse .

Jokes ... love it!

Up next ... I removed the circuit board and there are some interesting places on the board with obvious over heating ... even traces where the solder mask is burned off but not even near any component ... ill have pictures soon.

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

So I took the circuit board out and as you can see ... several places got damaged by excessive current draw ... the side by side is the board flipped against itself so you can see that one side of the board got burnt while the other side didn't except in that one general area.

The housing has FOUR fans, two on each end... the red and black wires connect those fans and thats all those wires do.

[stj]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

i bet the RTC doesnt keep time anymore.
how the hell do you use it - serial port??

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

i bet the RTC doesnt keep time anymore.
how the hell do you use it - serial port??

As a matter of fact, there is an RS-232 chip on the board ...

Also, I forgot to show this ...

These are the same ... looks like a voltage regulator...



Also not sure what this means, but this chip has like a volcano ... zit looking protrusion coming out of it ...

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

Also, the way that these two caps are sitting next to this blue thing that is right next to the microcontroller makes me think that the blue thing is a crystal oscillator .... is that correct?

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

OK, here's what I THINK is going on with this thing...

this image shows that there are THREE LED "Arrays" which consists of 5 rows of LEDs that are screwed down to a heat sink.



The wires that come in from the two PSUs are wired straight up to the circuit boards that are mounted next to each LED array. Then, we have this controller board...



That feeds each panel by those connectors at the top. ALSO, each panel has FOUR MOSFETS on the underside of the circuit boards attached to each panel and the top of those boards simply have what looks like 2 or 3 watt resistors ... something like 6 per board I think...

I BELIEVE - based on one of the chips I looked up, which was a PWM chip, that the LED panel MOSFETS are being fed a PWM voltage.

There is a wide connector that feeds the display / control screen, and a COM port that was unoccupied.

Heres where it gets kinda weird - at least to me it seems weird ...

The voltage that is powering the main controller board is taken from the end LED panel and it feeds directly into the input of a 7805 (which has burn marks on the PCB at its solder points) and THOSE are the wires that got their sheathing melted off. And take a look at these two MOSFETS where that power connection feeds from...

The red arrows point to the PSU wires and the two green arrows are pointing to those fried wires which are the power source for the controller board.



As you can see ... there is major discolleration at those two MOSFETS and it KIND OF LOOKS LIKE RUST TO ME like it got wet ...

Here is one of the other panels MOSFETS and you can see that they have no discoloration at all...



So this is where I'm not sure what to think ... clearly the main board was drawing too much power to that 5 volt regulator but yet the MOSFETS at that LED panel that feeds the power to the main board look suspect to me but they would have no role in powering the main board, they would only be powering the LED panel that they drive ...

So im not sure where to proceed from here...

[stj]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

if the main board locks up the led panels could be 100% duty cycle instead of pwm.

if you isolate the controller and feed 9v into the regulator, does it output 5v? and how much current does it draw?

[redwire]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

J3 is just the DC input to the 7805 5V regulator. With the -ve black wire burning up I think would be caused by a ground fault.
So there is an overcurrent happening between the J3(GND) and power... somewhere. Guessing a fan or LED board or heatsink insulation failed and has shorted to power and the resulting high current is pulling ground return current through J3?
I would look for an open ground wire from the mosfets to the PSU.
Poke around with an ohmmeter and see if the mosfets are still insulated from the heatsink, ground wire did not go open-circuit etc.
It's a bit of a weird failure.

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

if the main board locks up the led panels could be 100% duty cycle instead of pwm.

if you isolate the controller and feed 9v into the regulator, does it output 5v? and how much current does it draw?

OK, so using the bench power supply at 9 volts ... set to 5 amps, I have the board removed from the unit with nothing else connected to it and then I apply 9 volts to the 5volt regulator ... the regulator output pin reads 5 volts and the current draw is only .1 amps at the power supply ... so that's .1 amps at 9 volts from the power supplies perspective ... Im assuming that we're powering the MC with that voltage and the other chips on the board.

That certainly seems normal to me ... which means that it would probably become necessary to connect the light panels to see the over-current draw at the regulator???

Remember when the unit was still not taken apart we saw 5.5 amps being drawn on the 5 volt line coming from the PSU

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

J3 is just the DC input to the 7805 5V regulator. With the -ve black wire burning up I think would be caused by a ground fault.
So there is an overcurrent happening between the J3(GND) and power... somewhere. Guessing a fan or LED board or heatsink insulation failed and has shorted to power and the resulting high current is pulling ground return current through J3?
I would look for an open ground wire from the mosfets to the PSU.
Poke around with an ohmmeter and see if the mosfets are still insulated from the heatsink, ground wire did not go open-circuit etc.
It's a bit of a weird failure.

OK, the MOSFETS are NOT shorted to their heat sink.

But what I find strange is that the red wire (the slow burning fuse) is actually soldered at the same point where the black wire is soldered coming from the PSU...and the ground wire (from the slow burning fuse) can be seen as the tip of the green arrow is touching it ...

I didn't see any kind of short at that circuit board, though with the wires having been melted, I can certainly see how the exposed wire could have been shorting but I didn't look for that before removing the wire from the main board and jostling it around.

The two red circles are the same solder point from opposite sides of the board

And NO, those white connectors next to the resistors never connected to anything ... in case you were wondering.

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

it does not do esr, it's a basic compact clamp meter with a 100A limit.
it's fame is that it can read DC with the clamp with a 1mA resolution!

Got this today ... by the way...

[stj]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

damn that was fast!

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

damn that was fast!

Amazon!

[redwire]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

OK, the MOSFETS are NOT shorted to their heat sink.

I'm thinking +power goes to the LED's, so I would make sure the LED base is not shorted to the heatsink. Or an open high current ground wire/connection going back to the power supply. The red LED's don't seem to have insulated mounting screw holes on the PC board.
The mosfets are on the low-side, switching to ground so a short (to heatsink) there would just give power all the time to the LED's, drain-gnd short.

[EasyGoing1]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

I'm thinking +power goes to the LED's, so I would make sure the LED base is not shorted to the heatsink. Or an open high current ground wire/connection going back to the power supply. The red LED's don't seem to have insulated mounting screw holes on the PC board.
The mosfets are on the low-side, switching to ground so a short (to heatsink) there would just give power all the time to the LED's, drain-gnd short.

The red and black wires go to the main controller board (not the red and black wires in this picture ... the ones that fried) and provide it power via the 7805 ... and here is a picture of one row of lights but they are all like this ... soldered on one side only, the other side is aluminum and screwed down to a heat sink.

[stj]

Re: Fish Tank LED Power Supply Gets Hot then shuts down

before we get sidetracked,
where does the 5v and 12v from the psu go to?