Converting Double Insulated Light To Faux-Case Grounded

[dave9]

Someone gifted me an LED Work Light. Specifically a Utilitech Pro 12W # MPL1022-LED12K8040
https://www.lowes.com/pd/Utilitech-Pro-1000-Lumen-LED-Portable-Work-Light/1000260469

The casing is entirely plastic (+ glass front), no vent area and double insulated with only AC 110V live/neutral power cord.

It runs hot and other owners report short lifespan. I'd like to fix that.

Inside the mains 110VAC power cord is soldered directly to a thin (1mm?) mPCB that both the driver circuit and LEDs (15 in series) are mounted on. This mPCB sits on a heatspreader sheet of aluminum barely thicker than tin foil, held together by metal screws. I have some much thicker sheet aluminum that I could swap in and would be entirely encased in the plastic case, -OR- I could go with an aluminum heatsink sticking out the back of the plastic casing after a hole is cut for it.

The concern is with that line voltage present at the mPCB, if that mPCB were coupled to an aluminum heatsink exposed in the rear, it seems like it would be unsafe. I could add a huge silpad (but I don't have any and ordering that could cost as much as the light did) and still there's the metal screws connecting them.

Would you consider it safe enough to replace the original power cord with a grounded cord and earth ground an added, rear exposed aluminum heatsink, plus put a fuse on the live mains input? -OR- would you just replace the approx 0.35mm thick original heatspreader pictured with one 2mm thick? -OR- would you just use it till it failed then throw it away?

I'd like to not drill vent holes in the case, but could seal around a heatsink and keep it [poorly] water resistant. There is no gasket or conformal coating from the factory, it was never waterproof.

I suppose pics are worth a thousand words.
https://imgur.com/sXdq5Hq

https://imgur.com/a/i1LQY

 

[Bluejets]

Forget it or you will get yourself into problem areas you cannot imagine.

The heat dissipation is more than likely, allowed for in the design, and/or it has a certain duty cycle.

The label mentions nothing about being waterproof, just suitable for damp locations, not a recognized IP rating.

Best advice would be buy something suitable for your application.

 

[dave9]

Get myself into what problem areas I can't imagine? Seems like I'm imagining the problems already.

The heat dissipation might be allowed for but only within the context of outliving the warranty period. Obviously that ~0.35mm sheet with non-vented, low surface area is insufficient for 12W long term. Obvious to me because I've worked with LEDs and other power electronics a fair amount and as already mentioned, others have reported short lifespan (early failure).

Is this not also evidenced by other work lights of similar wattage having more robust heatsinking? The typical design has cast aluminum housing fins on the back. It can't be that these LEDs are some new alien tech with very high efficiency if they're getting only 1000lm out of 12W.

I don't feel like buying something else because then it's just more landfill waste, more cost, thicker, and I happen to already have various sheeting and heatsinks. I have done this sort of thing before with good results, just not to this particular product with a mains AC connection on a mPCB. They usually have a modular driver with low voltage DC output, 36V or less and that current limited, a lot less to be concerned about than AC mains at (whatever limit the circuit breaker or a fuse I add, has).

I could throw the thicker aluminum sheet in and take a thermal reading and will probably do that anyway before deciding whether to put a heatsink on the back, as the sheet would provide a good foundation to mount a heatsink, but if I'm going to throw a heatsink on, it would take less than half the time to fab the sheet if I only make it as big as needed to do that, instead of trying to maximize surface area with it including the cutouts around the cord, switch, two notched corners, and possibly even 90' bends along the top and bottom.

I am fairly certain that this alone has no issue I can't imagine. The only ones being needing ~ 1.5mm longer screws or to use alternate attachment points through-case with insulators, and to reduce the plastic reflector depth by a similar 1.5mm. The reflector doesn't hold the rest together, is more for show (concealment) than reflecting. Maybe I should just leave the reflector off then it would be easier to see when heat darkens it.

 

[Bluejets]

Regardless of how much you think you know, one cannot simply start hacking into mains powered equipment willy nilly and making modifications that can create a dangerous electrical situation.

 

[dave9]

Well, that's part of why I asked what others would do and I appreciate your feedback. Case grounded equipment isn't rare. I could probably even manage to shoehorn in a ground fault circuit though most case grounded equipment doesn't have one.

Willy nilly though? I can't even count the # of mains powered things I've repaired, and sometimes modified to fix the initial design flaw so it wouldn't need repaired again as soon.

That's kinda the direction I've gone in recent years, that I'd rather identify equipment weaknesses and fix them BEFORE the first fault, warranty be damned. Everything these days that's reasonably priced is considered disposable which I strongly disagree with doing if it can be helped.

The sad part is when the design improvement would have cost less than 10% of the purchase price and more than tripled the lifespan. I'm fairly sure I'm going to need light for longer than this unit will last if it isn't heatsunk better...

 

[Bluejets]

Ok...to begin with, the low voltage generated in the device is not isolated supply.
Then there is a possibility the heatsink plate may be connected to the line indirectly.
All the above is acceptable in a double insulated, encased, away from prying fingers, environment.
Bring any of it out to the real world will remove it's rating and it's certification, and more than likely make it an electrical danger to any and all who pass by.

 

[dave9]

Right, not isolated which is why the topic talks about earth grounded heatsink. The topic started out already a step past that. The mPCB backing and heat spreader plate are not connected to AC, would not be unless heat delaminated the mPCB tracks, or the solder joint for the mains cable broke and the wire flopped over onto it but this latter risk is present on any metal cased LED light.

I wasn't really interested in vague danger will robinson type replies. Well to be fair I did ask "Would you consider it safe enough" but was aiming for improvement ideas, not just avoidance of the topic of improving something mains powered.

If you feel that something MORE is needed to make it safer, or would heatsink better without large addt'l size/weight/cost, go right ahead and make that suggestion. I've decided this is getting a heatsink upgrade whether it be internal or external... wasn't really an option to just run it as is till it fails or I wouldn't have posted at all, would've just done that.

Explain how it will more than likely be an electrical danger to any and all who pass by compared to any other case grounded equipment? Non-isolated supply is present in tons of power tools, appliances, etc. This is old school stuff.

I already have a viable idea. If you have a way to IMPROVE upon that, it would be constructive. "Prying fingers" could yank any old incandescent factory lamp cord from a non-case grounded metal lamp and make it far more dangerous than a grounded heatsink, yet society deems that construction to be "safe enough", even high end metal lamps usually don't have case earth grounding.

 

[duke37]

Go for extending the internal heat sink, Use as thick a sheet as possible and cut it into offset fingers so the air can pass through from one side to the other.

The plastic case will produce considerable thermal resistance but I woud not modify this.
To add an external heat sink will involve very good eletrical insulation which would have a large thermal resistance.

 

[dave9]

I went with a simple flat plate 'sink, scrap material at that. I didn't feel there would be a lot of gain cutting fingers or even bending up the ends at 90', might slightly warp the middle anyway which I wanted to stay flat.

Not much else to say, I used 10 ga. 6061, a couple mm longer screws, marked the back of the reflector and sanded it down to fit.

I had some plate aluminum at least 1/4" thick but opted not to use it because I'd have to take so much more material off the reflector, and then it might hit the screws or SMD resistors, and at some point the pseudo-sealed enclosure is the next weakest link but I'm leaving it intact for at least a little water resistance.

It still gets hotter than I'd like but I can tell that the heat is spread out more than the prior piece of tinfoil 'sink allowed.

 

[(*steve*)]

If you can find a way to add a small fan to blow air over the heatsink, the effectiveness of the heatsink will be greatly enhanced.

 

[dave9]

^ Yeah there's the thing. It's a cheap light and very slim, no room for much more in it. I have several little DC fans but since this isn't set up with any LVDC output it'd need a power supply too or for me to buy a small AC fan that probably costs as much as the whole light does.

if it ever burns out I may cannibalize 2 x 9W LED light bulbs for their drivers and LED modules so I have the more traditional arrangement driving the LEDs from sub-30VDC, then I'll also have that DC available to power a fan I already have, though I'm not sure if the drivers will fit in an enclosure that thin, maybe if I got rid of the reflector which is such a wide angle that it doesn't really do much except hide the circuitry, and then more room for a fan too.