consolidating individual LED interior-lighting drivers?

[808Dave]

Amazon (etc) is full of LED lighting hardware, including the variant that I'm after for my renovation project. Sets of six to twelve LED 'wafer' heads, sold with drivers packaged in companion boxes, are not rare. The drivers are intended to be tucked unseen into a presumably-open ceiling space near the LED head, and the two wired up by means of mating pigtails that protrude from each component. Feed 120VAC to the drivers, and you've got light.

I, however, have 'cathedral' (raked) ceilings to work with, all solidly filled with foam insulation. I don't want 120VAC in them at all (even so, am still concerned about heat buildup from the wafer LED heads). Clicking around, it occurred to me that it SEEMS like I could buy the LED kits as described, NOT use the drivers, and instead consolidate/replace the drivers by means of one of the many LED drivers that are sold separately. I got all excited, thinking all I'd need do was pick out a dimmable one that had an ample wattage vs. the sum of my LED heads, and I'd be in biz. I'd parallel all my LED heads together, eventually consolidating all leads into one heavier-gauge twin-pair lead, and Bob would, in theory, be my uncle.

Only one thing: I've no idea what the voltage/current characteristics are for those individual drivers I'm trying to replace. Nobody wants to sell me a consolidating driver without knowing that, and rightly so. I'd been naively/vaguely assuming they'd all be 12V-max variable-voltage devices, but apparently it's not so simple. I gather they could require either PWM or analog drivers, and be...12V? 24? Some other voltage?

So does anyone know from experience what's IN those drivers, current- and voltage-wise, and whether it would be easy to emulate it with a higher-wattage driver sourced from a third party? Just buy the stuff and start poking around at it? Maybe. The hardware isn't terribly expensive, but I've already got too much stuff like that in my parts-graveyard from would-be customizations that didn't end up flying.

 

[808Dave]

addendum: I've tried contacting several seller-support types...the most they'll say is that I will void their warranty...no clue about data-sheets, etc.

 

[dave9]

As you've stated, there's lots of lighting hardware out there. You'd need to start with providing as much info and links to what you're considering as possible.

Decent drivers usually are current regulated, they can generate a certain range of output voltages, so you match that to the sum forward voltage of the LED array on each LED board, then the current regulation depends on both the ratings of the LEDs and the heatsinking capacity of the board. Often the generic Chinese manufacturers will rate something as "X" watts by multiplying the max rating of each LED * the number of LEDs, suppose 20 x 3W LEDs and they'll call it 60W, even if it has insufficient heatsinking so they set it up to run at 30W.

If you feel your ceiling insulation is going to be a problem with retaining heat, you might even need to run them at a lower wattage than the seller supplied drivers would, for good lifespan.

Running them in parallel, you will probably end up with some brighter than others because the forward voltage of each LED can vary a bit, so those with lower forward voltage get more current and vice versa. This is one of the problems running the modules in parallel without current limiting to each one.

Because of this, to meet your criteria of not having 120VAC through your ceiling, I would suggest determining the forward voltage of each module. Suppose it is 12V (DC), then you could run a 15VDC voltage regulated supply to all the modules and at each module have a 15VDC to 12VDC current regulating driver. Adjust the numbers to suit your parts, but leave enough voltage margin that you have a few volts between input and output of the LED driver.

Since you need to have wire strung anyway, is this avoidance of 120VAC runs just to avoid electrical code requirements?

 

[808Dave]

Thanks for the thought-out reply, Dave.

I did put up a link to one of the types of kits that I’ve been after, but you might not have noticed it because it’s hypertext.

I am indeed avoiding code and real-world safety concerns by keeping 120 V out of the ceiling, Honestly not sure what it would take to be compliant, while also not ending up filling the ceiling with, say, aluminum flex and a nightmare of individual armored runs. Hate spiral flex anyway...if you know of some non-metallic and code- compliant method, it seems like it would be a lot simpler that way. Plastic flex? I’d still be a bit concerned that at some point someone would screw or drill through it, but if it’s code approved I would be more likely to overlook that concern.

I’m in an airport lounge right now so not in a great frame of mind to ask following questions about the electronica details, but will check inlater when I get to my destination.

Mahalo...

 

[dave9]

I think the largest problem you're going to have is finding something dimmable that runs over a low voltage supply, unless it's some fancy smantzy smartphone controlled module or something, and I have no specific knowledge about those. If you must have them dimmable, and can't find a low voltage driver to do that, it might be a show-stopper forcing you to do high voltage runs to the lights.

Now that I see your Very Well Hidden link, I can see what the driver pictured has on its label.
Power:12W
Input: 110V-130V 50-60Hz PF>0.95
Output: DC27-42V 280mA (or 250, 260, something mA... close enough)
Triac Dimmable IC Type

Dimensions: 88mm x 85mm x 33mm (something to keep in mind if you want to reuse them with an open driver board but you will still have to device a mounting method, possibly drilling/tapping holes for standoffs)

Anyway, the above driver specs with a relatively low 280mA current imply the LED module is running a single series of LEDs and you'd need a boost driver rather than a buck (15V to 12V) example I mentioned previously. That may be easier to find if you settle for a 10W driver instead of 12W, possibly settling for a 350mA driver and modifying it to produce closer to 280mA if not lower (regarding heat buildup in your ceiling) but it may still be like looking for a needle in a haystack to find one that's dimmable using conventional dimmer controls, unless someone else reading this topic knows of some.

There's about where my help ends. If you can do without dimmable, next I would spend a day on aliexpress or ebay or wherever, for a boost LED driver, that accepts 24VDC or lower input (I don't know the max voltage in your area to stay within electrical code limits) and produces an output voltage range that goes up to 42V as the original driver did. If it is 280mA or a little less, it should work as-is.

If it is much above 280mA then you have to determine if you can, and are willing to, modify it to produce lower current, which typically involves reverse engineering the circuit by looking at the driver IC datasheet and then probably it has a current sense resistor value you can change to alter the current. Many datasheets even provide an equation for how to calculate the output current based on the resistor value used, which would help to choose a new resistor value.

 

[808Dave]

dave9 - thanks for your followup reply, though it left me a bit in the dust. I certainly didn't notice the specs you picked up on from that linked item's photos, but even so, I still need a couple of things spelled out for me, as I'm more from the 'caveman school of salt-water dimmers' than I am in the 'PWM LED driver's seat,' so to speak... Really your reply initially left me thinking it's time to just find a way to pipe 120VAC to all my LED heads and see what happens with the heat-buildup by running some tests, but that would truly be giving up too soon. (skipping comments on how grossly behind I am on this renovation work...)

I'm not sure I made it clear before, but I want to centralize & hide a single (ideally) driver that would use an input of 120VAC, close to an AC source. It would not be in the ceiling proper - instead, it would be in an adjacent storage loft-space on a short wall. From there, I'd envisioned low-voltage DC-carrying wires fanning out, sized according to the max current they'd be carrying in each (diminishing-size) branch, leading directly to my LED heads. The supplied drivers would go in a junk-box...

Dimming is vital, especially because of my uncertainty about how many lumens I'm really going to get out of each head (well...even if I knew precisely, really, just for flexibility in the room: game-time vs movietime). If that, alone, puts me at the point where I must abort this alternative, so be it - but I'm not sure that's where you were going. Wife would definitely shoot me (and I'd give her the ammo) if I were even caught looking at IC datasheets, let alone found trying to modify electronics at the component level to get this done (I should post my list of currently-backlogged projects in order to emphasize this, but you get the picture.)

I take it from those specs I missed that what I'd ideally be after, assuming I'm driving all 12 heads, would indeed be a constant-current driver that can provide a variable DC voltage from between 27 and 42. Some of what you wrote makes me think I'm wrong about the following (wishful?) conclusion - but anyway, left to my own, if I'd seen that label in the photo, I would have gotten from those specs that to run all heads from a single driver, it would need to be run a stable 280mA x [number of heads I need], so with my 12-head kit, it would need to provide 3.4A constant over that voltage-range...and so in my dreams, I'd be looking for a constant-current PWM dimmable driver that would output 3.4A, ranging from 0VDC at minimum dimmer setting to 42VDC at full throttle. But I think you're saying that'd probably be a rare bird, and I'd need to modify something to come up with those specs?

Poking around on Amazon again during my insomnia session this evening, it didn't take long to find a line of programmable PWM dimmable drivers that at least APPEAR to be able to meet these requirements, though I would have to buy their rare-to-nonexistent MSSL200 programming unit (let's assume I can get past that somehow, or someone else has similar hardware available). Apparently the $75 highest-current option in their driver line offers these specs - my red text emphasis indicates where I took hope:

• Input 90～305VAC 50/60Hz, Output 320W, 24～54VDC, 600～8000mA, the rated output current can be configured by MSSL200 so if you want to change the current please buy MSSL200
• Dimming Modes: 0-10V(Default),PWM,CLKS. Advanced communication & programming features including selection of the dimming curve.

Would you take from the above that with that driver, I'd have an off-the-shelf (once programmed) single central dimmable supply that could be tailored to my needs, then? The 320W rating amply covers what I could consume with all heads, yes? Do we think that that stated voltage-range is programmable (to limit to my needs), as would be the current?

Thanks for your help in thinking through this...

Dave

 

[dave9]

... all 12 heads, would indeed be a constant-current driver that can provide a variable DC voltage from between 27 and 42. Some of what you wrote makes me think I'm wrong about the following (wishful?) conclusion - but anyway, left to my own, if I'd seen that label in the photo, I would have gotten from those specs that to run all heads from a single driver, it would need to be run a stable 280mA x [number of heads I need], so with my 12-head kit, it would need to provide 3.4A constant over that voltage-range...and so in my dreams, I'd be looking for a constant-current PWM dimmable driver that would output 3.4A, ranging from 0VDC at minimum dimmer setting to 42VDC at full throttle. But I think you're saying that'd probably be a rare bird, and I'd need to modify something to come up with those specs?

The voltage range of the driver is more about allowable range, and yes it would need the sum of (n # lights) x 280mA capacity but that plus some margin for good lifespan is only the full brightness state. You want the opposite, a variable-current, though current regulated dimming driver. I can't tell if the unit you listed would work, the "Dimming modes 0-10V" alone is puzzling and this is often a problem with merchants providing incomplete information and some wrong or poorly translated to English.

The more I think about it, the more I think it won't work to make guesses about how multiple lights in parallel are going to react to a DIY attempt.

I don't want you to waste time or money towards an unworkable solution so I feel you need to visit a home lighting showroom and talk with the person there to find something engineered to meet your requirements, but I would keep the requirements loose and just be looking for 12 recessed lights that can be controlled from a dimmer and see what they have and narrow it down from there.

 

[808Dave]

Yeah, I think you're probably giving me reasonable advice, thanks. Lighting showrooms here tend to be staffed by low-tech-skilled people, and they're also pricey, so if I give up, my fallback is probably going to be 120VAC. So before I abort, I will see if that driver seller ("Moon's...") wants to jump in with support for this application.

My take on the basics, though, leaves me scratching my head again about your 'variable current' reference. The label you noticed on the drivers in the photo call for a fixed current, variable voltage (27-42), so...maybe you mean 'variable' in the sense that I'll want it to hit my sum total of i-requirement by my heads in parallel. Having read a primer on PWM dimming, I was just beginning to accept that the "variable" portion of this - whatever it is that allows the lights to dim - is accomplished by electronically chopping the current (somehow...) such that the square waveform produced by the driver still reaches my programmed height of 3.4A maximum with each pulse, while the wavelength of the pulses can be narrowed or widened such that...hmmm... the average voltage ranges between 27 and 42 (or between 0 and 42?), which I believe from this driver's specs is indeed programmable, and not limited to that 0-10V mentioned as only the default programmed option.

I'll see what, if anything, I hear back from Moon's about this and post back.

Thanks again - Dave

 

[dave9]

A typical LED driver has a current sensing feedback. Its output goes through a low ohm sense resistor, and measures the voltage drop across that and compares to its internal reference voltage.

It varies the voltage to achieve a specific current, thus current regulated. It lets the voltage rise or fall to end up at the desired current, but is only guaranteed to be stable doing this within the voltage range specified which is paired with an LED (or series on the same board) having a forward voltage drop within the same range at the desired drive current.

I found more information about what is probably the driver you're referring to including the datasheet and other specs on the "Download" tab on the following link. The dimming is a 1V-10VDC signal from the LED wall dimmer while it still gets mains 110-277(?)V AC input. I think it is possible it could work but don't know about the result of paralleled modules sharing current equally without an individual driver for each one. I could suggest that you experimentally put a low ohm series resistor on each LED module to share current more evenly but to what lengths you want to go to is offset by the other considerations.

https://www.moonsindustries.com/series/mu320hxxxaq-cp-series-a11010106

The following not-so-direct link to the datasheet PDF may or may not work, is convoluted instead of a direct link to a file, but the file is also linked on the page linked above:

https://www.moonsindustries.com/med...ZhMGIwMmRkYzlhY2M4YTAxOTViYTY&attachment=true

I'm not clear on what you hope to solve. This is a ~40-53V driver so it is not as though you are stringing LOW voltage DC runs using it. Code limits in some areas are 30VDC or lower. Such systems are generally set up at 12V or 24V which is just too low for the LED modules you linked.

I do not think this topic is simple enough to resolve in a few forum posts and that if your local lighting center cannot help, then you need a qualified electrician knowledgeable about LED lighting to do a site survey to determine what is going to work and stay within codes.

 

[dave9]

The other option is pick different LED modules that can operate from under 30V (though you never did tell us your location and what the code limit on voltage is there?) and then search for a dimmable low voltage system driver, at each LED fixture. I don't see a way to get around a per fixture regulation whether active or passive.

Suppose you picked LED modules capable of running from 12VDC, and put series resistors on each. In that case, you could hack together a setup where you have a variable voltage control to dim them, but it would need to have upper and lower voltage limits that are compatible with the dimming granularity needed, which seems more hands on than you want this to be.

It is time for me to exit the topic as I am probably just creating more distractions from seeking someone to look at what you have on site and determine the most cost effective option from there.