Fish Tank LED Power Supply Gets Hot then shuts down

[EasyGoing1]

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

OK, some interesting new knowledge ...

When the other two fan ports wouldn't lite up the fans, I suspected that maybe it had something to do with the thermistor mounted above the MC and I tried warming it up but no luck... so I decided to see what this board does with the LCD screen attached to it so I pulled it out of the chassis and connected it.

When it boots, it says that it is Jan 1, 1980, and that it was 80 degrees ... so I then connected the other fan to the nonworking port and when I heated the thermistor this time, it kicked on the other ports.

There are two fans on each end of the unit, and they are mounted so that one side blows air into the unit and the other side blows air out so I'm sure that they connect it so that only two fans are moving air until it gets too hot then it kicks on the other set ... and it also kicks up the fans that were running so that they run faster.

When I was done testing, I cut power and felt the heat sink on the 7805 and it was TOO HOT TO TOUCH

Any thoughts on that? It's just a 7805, and I have a bunch of those with new heat sinks should I replace it?

Current draw was only half an amp on the 12 volt bench supply. So the 7805 is powering the MC, and the LCD screen but the fans are sourced from +12 outside of the 7805 and they take about 140ma each and I had three of them running so the 7805 was technically only taxing the power supply with less than 200ma and it got that hot.

[EasyGoing1]

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

OK, replacing the 7805 didn't help the heat situation. I know that the higher the input voltage on those regulators, the hotter they run, so here's what I think I might do next ... well ... one of two options:

1) I could run 12 volts into a buck converter and drive the board at around 6.5 volts. But I would need to cut the trace on the PCB so that I can still feed 12 volts to the fans.

2) Same as above, only instead of powering the board with a 7805, I use a miniature buck converter that would be a replacement for the 7805, driving the board at 5 volts and heat wouldn't be an issue since they can tolerate several amps. I've already got a grip of these and they should work fine for powering that MC and its related hardware.

[stj]

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

i use those buck modules all the time, they are really good - but dont trust the pot on them.
cut the track for the pot and link one of the fixed resistors

[eccerr0r]

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

It shouldn't get that hot with the heat sink. At 200mA of 12V to 5V it should get less than 1.5 watts and this is in the threshold of not needing a heatsink at all. Did you measure current and see what happens as you disconnect possible consumers of the 5V from the 7805?

[EasyGoing1]

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

i use those buck modules all the time, they are really good - but dont trust the pot on them.
cut the track for the pot and link one of the fixed resistors

I've never used the POT on them ... always the trace cut!

[EasyGoing1]

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

It shouldn't get that hot with the heat sink. At 200mA of 12V to 5V it should get less than 1.5 watts and this is in the threshold of not needing a heatsink at all. Did you measure current and see what happens as you disconnect possible consumers of the 5V from the 7805?

Well so the bench power supply tells me how much current is being consumed and yes ... when I run one fan it consumes about 140ma - but again, that's not going through the 7805 ...

When I have two fans only connected, the 7805 gets BARELY warm after about three minutes ... but as soon as I plug in the LCD display, it heats up RIGHT NOW too hot to touch ... the laser temp gizmo says its 67C and the LCD screen adds about 160ma to the pull from the PSU.


I should also note that when ONLY the LCD display is running, (no fans connected), it draws 260ma and it heats up that 7805 to 67C.

The PSU says that the circuit is consuming about 3.7 to 4 watts ... I didn't think you could heat stuff to 67C with just 4 watts can you?

[eccerr0r]

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

A 4W incandescent night light bulb gets to incandescent hot...
but that's a very small area...

[stj]

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

so lets see the display assembly

[EasyGoing1]

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

so lets see the display assembly

It's EXACTLY one of these

It has many options like controlling the intensity of different colors, manually controlling the fan, and where matters of time are concerned, you can set the date and time, but also some stuff about the moon phases, the time of day for sun up and sun down ... even the ability to set the offset for the RTC if it drifts ...

[EasyGoing1]

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

A 4W incandescent night light bulb gets to incandescent hot...
but that's a very small area...

so lets see the display assembly

... so given the size of the 7805 with its heat sink, and considering the fact that this 7805 is BRAND NEW ... and the fact that it's heating up to 67C including the entire heat sink ... while only drawing 260 ma from a 12-volt source ... DO WE HAVE A PROBLEM? Or can I just replace that 7805 with a buck converter and call it good?

[petehall347]

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

Junction Temperature
The maximum operating junction temperature is usually 150 ºC for 7805 voltage regulators, and operating at the maximum will certainly affect the reliability of the device. The recommended operating junction temperature is usually 125 ºC. This figure tends to be the same for the majority of silicon based voltage regulators in a TO-220 package.

Ambient Temperature
The ambient temperature is the maximum temperature of the surrounding area that the device will encounter. It is usually 25 ºC; however, you have to consider the worst-case scenario. In the summer, the temperature can go as high as 30 ºC, and if the airflow was poor then a figure as high as 60 ºC is more appropriate.
from here ...
https://www.petervis.com/electronics...Regulator.html

[redwire]

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

DO WE HAVE A PROBLEM?

Yes, 260mA is way too high for that board and LCD backlight.
What does the power transistor next to the 7805 do?

[eccerr0r]

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

Is the whole system still drawing 5A on the 5V line? 260mA on the 7805 is not 5A on the 5V line...

Though 67°C is hot, and may start derating, it's probably fine and about what it's expected. 7805 are sturdy and usually won't kill themselves when they are overloaded. As long as you're getting 5V on the output, it hasn't shutdown.

[redwire]

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

OP has to find the short-circuit or this is a waste of time. If the ground trace to the controller board goes open-circuit it would damage the board with ground getting lifted up.

I think the 7805 current is high, not sure what the extra drain on that 5V rail is, the second TO-220? It might switch the backlight LED off.
But this thing isn't put together with the cooling fans running, so maybe it's OK.

[eccerr0r]

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

Since 5V from the PSUs are only connected to the LED boards, I suspect one of the boards have a shorted transistor+one of the two parallel LEDs on the board. Search for it... disconnect them to help find the defective boards?

[EasyGoing1]

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

Yes, 260mA is way too high for that board and LCD backlight.
What does the power transistor next to the 7805 do?

That's not a power transistor, that's a 3.3 volt regulator which powers the MC side of the circuit.

[EasyGoing1]

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

Since 5V from the PSUs are only connected to the LED boards, I suspect one of the boards have a shorted transistor+one of the two parallel LEDs on the board. Search for it... disconnect them to help find the defective boards?

I'll go through EACH LED row with the bench PSU since I now know how they are wired on the PCB. If any of them are shorted, that will show up immediately.

[EasyGoing1]

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

OP has to find the short-circuit or this is a waste of time. If the ground trace to the controller board goes open-circuit it would damage the board with ground getting lifted up.

I think the 7805 current is high, not sure what the extra drain on that 5V rail is, the second TO-220? It might switch the backlight LED off.
But this thing isn't put together with the cooling fans running, so maybe it's OK.

The cooling fans are only switched on or off by the +5V rail on the main controller board. Their actual power comes from the +12v rail which has no bearing on the current flowing through the 7805. They are also pulse width modulated based on the ambient temperature that is read with a thermistor mounted by the microcontroller, but switching them on or off or even pulse width modulating them has essentially zero impact on the current being drawn through the 7805 since that current is only in the context of the microcontroller and it simply switching pins on or off very rapidly.

So I'm not sure where to look for a short since the mainboard is working though I still have not re-assembled and tested with the LED panels connected, which is good since the issue here is manifesting without the LED panels being included.

Where would you suggest I look? Because what you seem to be telling me is that I SHOULD expect a 7805 to heat up to 67C when it's only passing 260ma from a +12v source... is that in fact, what you are telling me? OR are you telling me that 260ma is too high for that LCD display and that it should be drawing much less current than that?

Edit: ALSO, When I connect the main controller up to power with JUST the LCD screen attached and I turn the LCD screen backlighting to 0, it draws 170ma ... when I turn the backlighting up to max value, it draws 310ma total at the 12 volt source.

I also noticed that when I had the fans connected and they were merely in proximity to the 7805 such that there was a small amount of air flow around it, the temp of the 7805 dropped to about 45 to 50C and stayed there, and when this thing is all put together and operating there will always be air flow moving outside air in and through the entire enclosure and that cannot be changed via the interface. There is an option to override fan speed but your only options are Automatic speed control or all fans at 100%.

[stj]

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

check the lcd display for partially shorted caps and see if it has any resistors in series with the backlight

[redwire]

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

[...] Where would you suggest I look? Because what you seem to be telling me is that I SHOULD expect a 7805 to heat up to 67C when it's only passing 260ma from a +12v source... is that in fact, what you are telling me? OR are you telling me that 260ma is too high for that LCD display and that it should be drawing much less current than that? [...]

I think the LCD display backlight is a hog, 260mA at 5V is a 1.3W load which is lots for a character display. They are usually around 100-140mA. But maybe it's an older display or the backlight has a problem. This high current drain will naturally cause the 7805 to run hot, but it's got fan cooling so I think it's not a problem to go after yet.

I would say the original problem, the short-circuit causing overcurrent on the melted wire needs to be tracked down first. That's why I was asking to check the LED's for ground-faults, especially on the panel that doesn't light up and those powered from the main PSU 5V where you see the mystery 5.5A load. You can look at the LED's for burn marks or black spots too, shining a light in their lens.

Also, I can't see the end path for the hot pcb trace on the controller board, this is strange. Follow the blue line from J3 to whatever. It should have fused/melted the little traces if the LED driver boards were pulling high ground current - wait... I suspect it must be the first connector (board) that was drawing the high current. I suspect this one connector leads to the shorted culprit.

The LED's are likely a 1W or 3W part? So roughly 75 LED's which is around 75W or 225W and I can't see needing two AT PSU's to power this. It's gotta be really bright.
edit: you sure about a 3.3V reg? This board is from the 5V era, the Z8 etc. I think is 5V powered.