It’s been just over a year since I started using the Mars Hydro 300 watt grow light. Taking a look inside it clearly needed a quick disassembly…
A few LED’s were visibly melted/burned. To be specific, 4 of the blue ones towards the center. I’m not sure if there’s something different about *those* particular 4 (compared to the other blues in this unit), but needless to say, these ones were toast. Before anyone asks, I did check to make sure there had been thermal paste behind them, and there was.
I ended up replacing them with a few cheap white LED’s that I had kicking around, along with some fresh thermal paste.
As to other internals, I found out where the fruitflies had started going once my fly tape was full:
Air compressor took care of things there. I was tempted to replace the thermal paste behind the heatsink (5 screws to detach heatsink from PCB), but the existing thermal paste had it glued on pretty good so I decided to leave things be.
Looking at the fan and drivers…
This had stayed pretty clean and dust wasn’t an issue. I gave it a quick shot with compressed air afterwards since I had it open anyway.
With the new LED’s soldered in, I plugged it in and was up and running….
…except for 1 other white LED that I’d missed.
I pulled it apart later and swapped out that final LED. All was now well!
Thoughts After 1 Year of Ownership
This grow light’s seen various degrees of usage over the last year. Periods of running 24/7, periods of sporadic use on a timer, and periods where it hasn’t been in use at all.
I’m a bit disappointed that 5 LEDs went. It’s generally been well ventilated (ambient temps of 18-24 degrees C), and I really wouldn’t have expected LEDs to start dying. If Mars Hydro cheaped out somewhere, I’d guess it was in the LED brand/quality.
I do really like the rest of the unit itself though. The fan is fairly quiet, I haven’t had a power supply (LED driver) die yet (despite losing a few other drivers in other systems), the thing has a heatsink against the PCB, bypass diodes are used to keep the good LEDs running when an LED dies, and some of the “little things” are there too (bushings to electrically isolate PCB from screws that go into unit, cables nicely zip tied, ample thermal paste used, easy disassembly, etc).
February 2019 Update
A number of years later here… half the LEDs in the unit stopped lighting up. My first thought was a dead driver. However, I swapped drivers around and it turned out both drivers were fine! I pulled out the bench power supply and found 5 of the white LEDs out. Replacing the LEDs had the unit up and running again.
As to why the LEDs being out stopped the whole string from working? Two guesses here:
- It’s possible one of the zeners (bypass diodes) run in parallel with a dead LED also happened to die.
- It’s possible the sheer number of dead LEDs pushed up the required voltage so high that it ended up out of the driver’s max output voltage (since the zeners are undoubtedly higher voltage than the white LEDs).
In any case, if half your LEDs go out on this panel, before assuming it’s the driver, swap the drivers to make sure you aren’t simply losing half the string due to a number of dead LEDs. For those who do need a driver (a couple have asked), I did come across eBay listings for 50W/60W 600mA 54-105V drivers which would likely work as a drop-in replacement around the $15USD mark. You may have to search around a bit as high-voltage drivers aren’t super common, but they do seem to be out there.
-Dan.
The drivers do have a voltage range and different LEDs are driven by different voltages to reach the desired current - you have a bit of flexibility when it comes to swapping in LEDs of a different color but only to a point. Red 1W/3W LEDs tend to use around 2.2v each, and white/blue 1W/3W LEDs tend to use around 3.4v each. If for example you replaced *all* your leds with reds, the voltage might be too low for the driver's range in which case it's possible the driver might overdrive them and start blowing LEDs. Going the opposite route, if you replaced *all* your leds with a combination of blue & white the required voltage might be too high for the driver - in this case LEDs might start strobing or the driver might blow.
The easiest way to minimize any potential issues is to replace reds with reds, and consider blue/white to be the same (replace blue/white with either blue or white).
The other option is to look at the driver voltage range, and then add up the voltages of all your LEDS on that circuit - so if 1 of the drivers is running on it's circuit... say... 25 reds, 5 blue, and 7 white, add up (25x2.2v) + (5x3.4v) + (7x3.4v) = 95.8 volts. As long as that falls within the voltage range printed on the driver you're good - if it's getting near the minimum/maximum of the driver's voltage range you may want to adjust (swap to some reds to push down the voltage, or swap to some blues/whites to push the voltage up).
All that said, if you're just replacing a few LEDs, most drivers have enough tolerance that you can probably swap in whatever colors you want and get away with it. If you get to the point where you're replacing a *lot* of LEDs though, the color is going to start to matter a lot more and you may have to do a bit of math to ensure you're not trying to force the driver to operate outside of it's range.
Can you tell me if that's possible with the mars hydro 300 ?
Thanks,
Ben
In various LED units that do have them, if they aren't hooked up to a physical switch the ideal option is generally to replace the LED with another (generally replace UV with either blue or white and IR with red to keep voltages similar).
The less ideal option is to flat out remove/desolder them. If there's a bypass diode the other LEDs will still come on. If there isn't a bypass diode, you could solder a bridge in the line. Downside here, particularly if the unit happened to have a *lot* of LEDs is that this could push the voltage outside the range of the CC driver. Or if a CV driver is being used it'll dump a lot more current through the existing LEDs and/or might overrun the driver.
Edit: The 55-100V range is high enough to be quite dangerous and can hurt/kill, so keep that in mind if you decide to use a voltmeter to get an exact voltage from a working driver while powering the LEDs. Don't be touching anything while it's plugged in. If you splice in a voltmeter do that while it's unplugged.
Note that one of the driver enclosures also goes to the fan (possibly a separate driver contained within). So if that's the one that died, assuming you can't find another driver with dual outputs you'll likely need a separate fan power supply. The fan itself in mine was rated 12V 0.3A. The 12V/300mA drivers I've found are pretty hit and miss so I'd be inclined to look for a constant voltage 12v driver that can handle up to 1A.
Do a quick power-on test without the corresponding LED connected to make sure the unit works with the new zener.
If you have too many higher-voltage zeners activating (lots of burned out LEDs), you might push the CC driver's output voltage too high, but assuming you catch/replaced burned out LEDs before the death count gets too high, I wouldn't be too worried.
If Mars Hydro originally had 3.0-3.4v ones made, jumping to 6-7v might be a bit high for the driver depending on what voltage the other LEDs add up to. If the voltage is too high, the driver will likely start strobing when they're installed (unless the voltage is way too high in which case it may not even power on). Of course if the zener diodes Mars Hydro used activate at less than 6v, your blues simply won't light up.
The easiest thing to do might be to add up all the reds (should be ~2.2-2.4v each), and then figure out the total voltage if blue/white are ~6.5v and do the same assuming blue/white are 3.25v, and see what voltages fit within the driver range. A better option if you have a DC supply kicking around would be to pull a couple blue/whites and hook them up to the DC power supply and slowly ramp up the voltage until you find out where they light up to see whether Mars used ~3-4v or ~6-7v leds originally.
If it turns out the required voltage is too high, one option would be to use 1-2 of your new blue ones and use reds for the others (the eBay 5W reds tend to be 2.2v, so I'm assuming the chips are usually run in parallel there).
Another (perhaps more temporary) option if your required voltage is too high would be to install 1 or 2 of the blues and either rely on the zener to bypass the open slots, or solder in a jumper wire across each open spot.
One of my drivers has died and it would be great to find a replacement.
Love your post btw. Mars Hydro lights are great.
Thanks
Appreciate your help Matt.
I should probably also mention another option: you could use the current (working) driver and simply run the LED circuits in parallel. Each side would get approximately half the current. I've done this on some units before to get a little more LED efficiency and less heat (using 1 driver out of 2 or even 1 out of 4 to power all the LEDs). A decent interim option, as you'll get more total light than you do right now due to efficiency increases, though less light than having a dedicated 700mA running each circuit separately.
Either way, good luck!
I just had a look on the meanwell website. I hope you don't mind me asking your opinion on these CC drivers:
ELG-75-C700 or ELG-75-C500
HVGC-65-700 or HVGC-65-500
HLG-60H-C700
I'm new to this, but these all seem like good replacements? I will get my tape measure to check if they fit!
As a side note, before buying anything you may also want to ensure it's the driver which is dead. Sometimes if multiple LEDs die, the zener/bypass diodes push the required voltage up beyond the limit of the driver. To check this you'd want to swap the drivers.
Looking on Mouser, it seems to have 3 different driver types. A, B, and just ELG-75-C700??
My drivers are definitely the same as yours. I notice that the Mars Hydro drivers are labelled A and B, does it use both types in one light? If driver B is blown on my Mars Hydro, should i buy a Driver B ELG-75-C700?
It's definitely the driver that has died. I swapped them around when i opened it up and different lights came on.
The standard ELG-75-C700 looks to be a fixed 700mA. The A model has a potentiometer to vary from 350-700mA. The B and AB have a dimming function. So I'd either go with the standard model or the A model.
For the completely dead string, if I thought the issue was that enough failed that the bypass/zener diodes are now pushing the voltage out of the range, I'd likely start by replacing with red LEDs, as I would expect the string to come back up after popping in a couple reds.