DIY DIY Upgrade to the Roleadro 400w COB

[4d-Rock]

I looked back in my emails to Roleadro sales/support. My second lamp only drew 175w from the wall which was greater than the +/-5%. The first drew 195w. That's why I tried to send the second back and then decided to hack it. I never expected it to draw 400w from the wall, no.

My buddy did a write up for me on the pro's. It's a complicated picture but I will post after work today!!

Ahh, gotcha. I don't have the tools to test mine so I guess I'll just have to take their word on it.

Thanks

 

[Banjacked]

Ahh, gotcha. I don't have the tools to test mine so I guess I'll just have to take their word on it.

Thanks

You only need a multimeter, a simple "clamp" type is perfect, you can then open the case, clamp around the "hot" wire inside from either the input plug or the switch, then turn on and get the current draw in Amps, no need to disconnect things and risk zapping yourself.

Watts = Volts x Amps so, if you are on 120VAC and draw 1A then you pull 120 watts.

 

[Screwauger]

Can't see the details on the drivers so can't be sure.

Here ya go

 

[Screwauger]

@4d-Rock

Here is my buddies reply to some or all of your questions.

I assume these are probably better quality COBs as well but are you saying that they are a better grow light due to the 4000K light temp. vs what is in there currently?

Extra special attention there needs to given to the fact that they are 4000k 90cri cobs and not just 4000k. Sure, we would’ve preferred 3000k or 3500k 90cri. But we were taking advantage of some low cost leftover cobs too. Here is distribution graph for gen 5 90cri 1212


E00F6FED-19EC-417E-98E3-AED3F107E842.jpg750x1334 154 KB
Pay extra special attention to the longer wavelength peak around 630nm and how it provides at least some light from about 415nm all the way out to beyond 780nm. Vs what manufacturer provides


B7190620-985F-4B3D-9F4E-A56B9B0B7366.jpg750x1334 142 KB
Every published light study from McCree, NASA, Oxford, and Emerson supports that you now have a more photosynthetically efficient light spectrum that will also trigger different plant responses.

Did you conduct any test measurements of the light output of pre- and post- upgrade? Is it significant or is the main benefit the reduction in energy draw?

The simple answer here is yes. Citizen gives us minimum and typical output, roleadro not so much. You can probably guess why? Since light spectrum is different you would need par meter to even give close explanation. And you could buy and mod 2-3 lights for cost of par meter. So it doesn’t make a lot of sense to drop that kind of money to test it.

However, it is pretty standard knowledge across led diyers that a 50% reduction in drive current generally works out to be about 10% increase in electrical efficiency (par watt vs heat watt). Nobody is paying for sphere tests on the roleadro, I don’t really blame them either. But several tests have put the 1212 at 2.3 umol/joule at 1050ma. So we can probably estimate at 1200ma we’re around 2.2 umol/ give or take. The lights using epistar leds that have been tested usually come out between.9 and 1.4 umol/joule. Assuming this is inline with rest of their products we’re talking a range of about being in the range of 50-100% increase in par per watt. There’s obviously a lot of speculation there. But turning it on, your eyeball can probably validate that it’s at least better.

I’m not sure if it matters to you but I noticed that the fan noise is a bit on the high end.

I’m not sure, but I feel like we may have potentially discussed swapping fan in planning stages? If not, it was definitely with one of the other guys hacking. But yeah, inexpensive light is going to use inexpensive components. No surprise there. Amazon sells higher quality pc fans for $5-$15 depending on size. You can certainly upgrade them at any time if noise is a problem or you’re worried about longevity. Just verify voltage and current are within specs of the provided power supply.

but do you know why they use 2 drivers per COB? And why are you able to only use one driver per COB in your setup (more efficient COB?)? Could this reduce lifespan in anyway as using 2 drivers may share the load?

That was a choice not a requirement, and efficiency is absolutely one of the reasons. This chart explains that.


A1496534-8B81-48B5-B3A4-5FEE390664A7.jpg750x1334 120 KB
But there is a bigger reason.


43E769E3-0FB2-47E5-8E5B-797FC0F6B542.jpg750x1334 136 KB
If I remember, they’re constant current drivers which means the driver will provide priority to supplying current and then let the cob take the voltage it needs, provided it’s eithin the rating the driver. If we supply too much current the cob would require a voltage beyond the drivers rating. Obviously that cob isn’t rated for twice the voltage either, so running off a single driver with $9 cob is a no brainer in my opinion. We tossed around some other options, all at significant increase in cost that would be trading more power for a lot more money. Plus, not being exactly sure if they were doubling current or voltage there could potentially cause other issues. And we’d really need a data sheet on the factory drivers to see if it would go cv/cc mode if we needed it to.

The part about reducing lifespan is actually completely the opposite. The driver regulates power differently than something around the house. So that part is a non issue. The fact that we’re running at half the power means half the heat. And heat is everything when it comes to leds. The life expectancy of leds is directly reflective of the current/temp relationship. The higher the current, the case and junction temps will go up. The higher the temps go up, the less life expectancy will be.

 

[Banjacked]

Here ya go

View attachment 957829

Each driver runs a max of 42w (36v x 1.2A) so that explains them doubling up to ensure the original cob had the juice it needed, otherwise, like I said earlier, they would have needed bigger drivers to run one per cob which would affect the casing size, cooling, etc.

Explains why these things fail so quickly and easily, one driver drops and then the other gets a severe overload which fries it

 

[4d-Rock]

@Screwauger
Much appreciated! Very detailed and interesting info!

 

[hashead]

@4d-Rock

Here is my buddies reply to some or all of your questions.

I assume these are probably better quality COBs as well but are you saying that they are a better grow light due to the 4000K light temp. vs what is in there currently?

Extra special attention there needs to given to the fact that they are 4000k 90cri cobs and not just 4000k. Sure, we would’ve preferred 3000k or 3500k 90cri. But we were taking advantage of some low cost leftover cobs too. Here is distribution graph for gen 5 90cri 1212


E00F6FED-19EC-417E-98E3-AED3F107E842.jpg750x1334 154 KB
Pay extra special attention to the longer wavelength peak around 630nm and how it provides at least some light from about 415nm all the way out to beyond 780nm. Vs what manufacturer provides


B7190620-985F-4B3D-9F4E-A56B9B0B7366.jpg750x1334 142 KB
Every published light study from McCree, NASA, Oxford, and Emerson supports that you now have a more photosynthetically efficient light spectrum that will also trigger different plant responses.

Did you conduct any test measurements of the light output of pre- and post- upgrade? Is it significant or is the main benefit the reduction in energy draw?

The simple answer here is yes. Citizen gives us minimum and typical output, roleadro not so much. You can probably guess why? Since light spectrum is different you would need par meter to even give close explanation. And you could buy and mod 2-3 lights for cost of par meter. So it doesn’t make a lot of sense to drop that kind of money to test it.

However, it is pretty standard knowledge across led diyers that a 50% reduction in drive current generally works out to be about 10% increase in electrical efficiency (par watt vs heat watt). Nobody is paying for sphere tests on the roleadro, I don’t really blame them either. But several tests have put the 1212 at 2.3 umol/joule at 1050ma. So we can probably estimate at 1200ma we’re around 2.2 umol/ give or take. The lights using epistar leds that have been tested usually come out between.9 and 1.4 umol/joule. Assuming this is inline with rest of their products we’re talking a range of about being in the range of 50-100% increase in par per watt. There’s obviously a lot of speculation there. But turning it on, your eyeball can probably validate that it’s at least better.

I’m not sure if it matters to you but I noticed that the fan noise is a bit on the high end.

I’m not sure, but I feel like we may have potentially discussed swapping fan in planning stages? If not, it was definitely with one of the other guys hacking. But yeah, inexpensive light is going to use inexpensive components. No surprise there. Amazon sells higher quality pc fans for $5-$15 depending on size. You can certainly upgrade them at any time if noise is a problem or you’re worried about longevity. Just verify voltage and current are within specs of the provided power supply.

but do you know why they use 2 drivers per COB? And why are you able to only use one driver per COB in your setup (more efficient COB?)? Could this reduce lifespan in anyway as using 2 drivers may share the load?

That was a choice not a requirement, and efficiency is absolutely one of the reasons. This chart explains that.


A1496534-8B81-48B5-B3A4-5FEE390664A7.jpg750x1334 120 KB
But there is a bigger reason.


43E769E3-0FB2-47E5-8E5B-797FC0F6B542.jpg750x1334 136 KB
If I remember, they’re constant current drivers which means the driver will provide priority to supplying current and then let the cob take the voltage it needs, provided it’s eithin the rating the driver. If we supply too much current the cob would require a voltage beyond the drivers rating. Obviously that cob isn’t rated for twice the voltage either, so running off a single driver with $9 cob is a no brainer in my opinion. We tossed around some other options, all at significant increase in cost that would be trading more power for a lot more money. Plus, not being exactly sure if they were doubling current or voltage there could potentially cause other issues. And we’d really need a data sheet on the factory drivers to see if it would go cv/cc mode if we needed it to.

The part about reducing lifespan is actually completely the opposite. The driver regulates power differently than something around the house. So that part is a non issue. The fact that we’re running at half the power means half the heat. And heat is everything when it comes to leds. The life expectancy of leds is directly reflective of the current/temp relationship. The higher the current, the case and junction temps will go up. The higher the temps go up, the less life expectancy will be.

Epic write up man! This is exactly what im gonna do with my roleadro, except with citizen version 6 3500k 1212's. Could I ditch the holders/reflectors and just use termal paste? Line the new chips up with the original reflectors?

 

[hashead]

Did my upgrade, minus the holders and reflectors. Lined the chip up with the original reflector and used termal paste and kapton tape to secure the chip (citizen 1212 CLU048-353-M2M2-F1). Soldered the connections and sealed em with hot glue so they wouldnt short if the casing touched. Popped the case back on, plugged her in, and jobs a good un.

I definitely agree its way brighter! And at half the wattage, well worth the bit of effort

 

[4d-Rock]

Very cool @hashead !
Did you take any pictures while putting it together? I’d be interested in seeing how it differs as I think this is the way I’d also go if/when I do the upgrade.

 

[hashead]

Very cool @hashead !
Did you take any pictures while putting it together? I’d be interested in seeing how it differs as I think this is the way I’d also go if/when I do the upgrade.

I didnt unfortunately :/ but it was really straight forward! The original chip is mounted on a plate thats screwed to the heatsink. Just using the citizen chip gives ya about 2mm between the original reflectors and new chips. Ill pull the case off when I get home and post a pic!