Cree and Bridgelux comparison

I'm planning an LED panel build and was originally going to use 18 Cree CXA-2530s. Running them at 40 watts (65% of max) allowed me to easily work out a light grid that delivers good coverage in a watt range appropriate for my space. I had briefly looked at The Bridgelux Vero series. I decided the Vero 18 would have required more emitters and the Vero 29 would require less, causing grid layout problems and light mixing issues (same reason I decided against the CXA3050).

Then after I had my mind all made up I read an article showing how Bridgelux has a slight par advantage over Cree for their respective emitters.

https://www.rollitup.org/led-other-lighting/743645-cree-cxa-analysis.html

So I take a closer look at the data sheets. It doesn't give a max watt rating like the Cree data sheets but it does state they have a 50,000 hour rating at 2X the test current. I'm pretty sure this means the Vero18 is a 60 watt unit, which would make it a direct comparison with the 60 watt Cree CXA-2530.

Can anyone confirm that the Vero18 is a 60 watt emitter?

If I find this to be the case I'll end up using Bridgelux. The link I provided shows the Vero series putting out more light per watt and they're a few bucks cheaper to boot.

Rahz said:

Can anyone confirm that the Vero18 is a 60 watt emitter?

Click to expand...

It would be more than that. I have Vero10's which have 18 dies of 1W (35 mil) in a 2x9 configuration. The forward voltage per die is about 3V. Extrapolating the voltage and current, the Vero18 would have 80 1W dies. The Chinese would call it a 80W COB.

Thanks for the reply. That defiantly makes the Vero look nice. I notice in the alternative drive currents the voltage for the Vero is always under 32, while 37V is typical for the Cree CXA. This presents a problem because I've found a driver for the Cree that's $22 an it runs 38V 1.05A. I can't find anything in that price category for the Vero at the suggested 30.4V 1.4A.

What would happen if I ran the Vero 18 at 38 volts? I'm new to electronics so that may seem like a dumb question.

for LEDs, you want constant current drivers. take another look at the driver you found -- if it's constant current (1.05a), then generally there will be a min/max voltage. Are you saying it's minimum 38v?

Rahz said:

Thanks for the reply. That defiantly makes the Vero look nice. I notice in the alternative drive currents the voltage for the Vero is always under 32, while 37V is typical for the Cree CXA. This presents a problem because I've found a driver for the Cree that's $22 an it runs 38V 1.05A. I can't find anything in that price category for the Vero at the suggested 30.4V 1.4A.

What would happen if I ran the Vero 18 at 38 volts? I'm new to electronics so that may seem like a dumb question.

Click to expand...

Rahz said:

I'm planning an LED panel build and was originally going to use 18 Cree CXA-2530s. Running them at 40 watts (65% of max) allowed me to easily work out a light grid that delivers good coverage in a watt range appropriate for my space. I had briefly looked at The Bridgelux Vero series. I decided the Vero 18 would have required more emitters and the Vero 29 would require less, causing grid layout problems and light mixing issues (same reason I decided against the CXA3050).

Then after I had my mind all made up I read an article showing how Bridgelux has a slight par advantage over Cree for their respective emitters.

https://www.rollitup.org/led-other-lighting/743645-cree-cxa-analysis.html

So I take a closer look at the data sheets. It doesn't give a max watt rating like the Cree data sheets but it does state they have a 50,000 hour rating at 2X the test current. I'm pretty sure this means the Vero18 is a 60 watt unit, which would make it a direct comparison with the 60 watt Cree CXA-2530.

Can anyone confirm that the Vero18 is a 60 watt emitter?

If I find this to be the case I'll end up using Bridgelux. The link I provided shows the Vero series putting out more light per watt and they're a few bucks cheaper to boot.

Click to expand...

What's the BOM run on something like that?

Rahz said:

What would happen if I ran the Vero 18 at 38 volts? I'm new to electronics so that may seem like a dumb question.

Click to expand...

Here's a nice intro to LED drivers. Consider running the vero 18 in series at 700mA.

Are you saying it's minimum 38v?

Click to expand...

The driver in question has a max voltage of 38. With it's current of 1.05A at max it would provide 40 watts @ 38V. I have found the max voltage table for the Vero18 and it is in fact lower than 38 so I would need to choose another driver.

The cheapest 40 watt driver option I can find for the Vero18 is about $10 more than the 38 volt driver I mentioned, which negates the cost savings over the Cree units. For $10 more than that I can find 50 watt drivers. That would justify the extra costs but I'm not sure I want to add the extra power (seriously considering it though). The Cree build would be 720 watts. For about $350 more I could build a 900 watt panel with the Vero18s. hmm...

What's the BOM run on something like that?

Click to expand...

Still working with rough estimates, but I can provide a close figure for the emitter/drive/heatsink. For the Cree setup, around $24 for an emitter, $22 for a driver, $15 for a heatsink, $61 total, x18 = $1098 for 720 watts. That figure doesn't include connectors, wiring, or frame assembly.

For the Vero18 $22 for an emitter, $42 for a driver, $15 for a heatsink, $79 total, x18 = $1422 for 900 watts.

Grow space: 15 square feet. Cree setup would provide 48 watts per foot. Vero setup would provide 60 watts per foot. Thoughts?

Here's a nice intro to LED drivers. Consider running the vero 18in series at 700mA.

Click to expand...

Thanks for the link. I might save some money running in series, but was planning on discrete drivers for minimizing cost in case of a driver malfunction. Like I said I'm new to electronics and don't have much idea of how much advantage that would provide over the long run. This electrical crash course has been driving me batty but it would probably be wise to do a cost analysis for a series setup as well.

Rahz said:

Thanks for the link. I might save some money running in series, but was planning on discrete drivers for minimizing cost in case of a driver malfunction. Like I said I'm new to electronics and don't have much idea of how much advantage that would provide over the long run. This electrical crash course has been driving me batty but it would probably be wise to do a cost analysis for a series setup as well.

Click to expand...

I dunno man. Might be hard to compete with some of the pre-made units cost-wise. They're getting a greatly-reduced rate on those components. Then factor in all the hours you are going to put into this. I can see the reward if it's for hobby-sake though.

I was a huge fan of DIY for my first 2-3 grows. After that I've largely abandoned the practice in favor of my free-time

Hi rahz
I think you should be able to use the driver you planned on using with the cree chips with the vero. You need the drivers minimum voltage to be below the chips minimum voltage and the driver maximum voltage to be above the chips maximum voltage. As long as your chips operating voltages fall inside the drivers output voltages you should be ok (assuming they're constant current drivers.)
im sure someone more knowledgeable than me will be along soon to help.
Sounds like its going to be a great light when its done, similar to what I plan but much bigger.
Hope this helps.
Andy

The driver would work, but at the chips max voltage and a constaint current of 1.05A it would only provide about 34 watts.

I'm leaning towards the 720 watt Cree build again. 900 watts would be nice but the 50 watt heat sinks would just barely get the job done, and a heatsink upgrade would further increase the costs.

Might be hard to compete with some of the pre-made units cost-wise

Click to expand...

I thought so too when I first started looking. But I can probably finish it for around $1500. That's for 720 watts to the emitters using top bin parts.

The driver I found (part# PLM-40-1050) is a real steal compared to most other 40 watt drivers. The Chinese made heatsinks are pretty cheap too and there is a 'make an offer' option so I might be able to get them a little cheaper than $15 each.

Rahz said:

The driver would work, but at the chips max voltage and a constaint current of 1.05A it would only provide about 34 watts.

I'm leaning towards the 720 watt Cree build again. 900 watts would be nice but the 50 watt heat sinks would just barely get the job done, and a heatsink upgrade would further increase the costs.



I thought so too when I first started looking. But I can probably finish it for around $1500. That's for 720 watts to the emitters using top bin parts.

The driver I found (part# PLM-40-1050) is a real steal compared to most other 40 watt drivers. The Chinese made heatsinks are pretty cheap too and there is a 'make an offer' option so I might be able to get them a little cheaper than $15 each.

Click to expand...

I'm fascinated to see your results. Subbed all the way. If you can build a better-quality product for near or under the price of the pre-mades, I might do one myself.

After more research I've managed to up the watts and lower the price over the Cree build. I found a Mean Well 60 watt driver for under $20. At 30.4 volts it will deliver 42.6 watts to a Vero 18. Part# LPC-60-1400

The Vero 18 is also cheaper than the Cree module and comes with connectors. I have also secured my heat sinks for $13 each.

Cost for emitters, drivers and heat sinks is now $920

Total watts has jumped from 720 to 767 and the build should now be less than $2 per watt. I'll post updates as they happen.

Nice job researching the build. Can't wait to see what the finished product will look like and how it will work.

planning on putting dimmers? With that much watts it probably help a lot being able to adjust the output

Rahz said:

After more research I've managed to up the watts and lower the price over the Cree build. I found a Mean Well 60 watt driver for under $20. At 30.4 volts it will deliver 42.6 watts to a Vero 18. Part# LPC-60-1400

The Vero 18 is also cheaper than the Cree module and comes with connectors. I have also secured my heat sinks for $13 each.

Cost for emitters, drivers and heat sinks is now $920

Total watts has jumped from 720 to 767 and the build should now be less than $2 per watt. I'll post updates as they happen.

Click to expand...

That's pretty much what I did, but with the LPF-60d-36 at 1667mA - I paid $32 for the dimming driver. Gets just over 50w per emitter at full speed, but the heat output at top end is definitely a bit of a thing to deal with. Paid $20 for my heatsink too, so that makes the emitter the cheapest components of my build. Can't believe the price of these things

Everything with this gear is totally bullet proof - late one night i stupidly crosswired my two v18s between the two ballasts. Things got a bit funky but the drivers just shut themselves down with everything OK.

Can you link the heatsinks? I am always in the market for more but rarely can beat Heatsink USA prices. If this helps you choose your drive current, here is a general idea of the current droop for Vero 18 3000K.

Radiometric efficiency assuming Tj 50c
.35A - 37.8%
.7A - 34.4%
1.05A - 31.9%
1.4A - 30%
2.1A - 27.6%

The harder you run them the more heatsink/watt you would need to maintain a given Tj. Also the harder the current the taller the canopy could be. The new Area51 all white 3700K panels will be about $4/watt and approx 30% efficient so it seems like your build is a good value at $2/watt.

plan on putting dimmers?

Click to expand...

I haven't planned on it. It will be about 50W per square foot and well spread so I hope it won't be an issue.

Can't believe the price of these things

Click to expand...

Yea, I avoided the 60 watt drivers because I assumed they would all be more expensive than the 40Ws.

Radiometric efficiency

Click to expand...

I've been shooting for 40 watts, so 1.4A or 2.1A seems to be my only options considering the voltage requirements. I don't really grok the voltage except as means of determining total watts. I need to do some more reading.

50 watt radial heatsink: http://www.ebay.com/itm/2-pcs-up-to-50W-Watt-LED-Aluminum-Heatsink-Round-XL-10W-20W-30W-45W-50W-/251138638555?pt=LH_DefaultDomain_0&hash=item3a79078349

There are other sellers and you can find the same thing in a 100W version. Friend says they're razor sharp so watch out. I offered $13 and it was accepted so you might be able to get them for less.

Thanks for the link. I think the radial ebay heatsinks will be best for a passive cooling situation. Besides that I've already ordered them

Found some quality terminal strips on Amazon. http://www.amazon.com/Amico-Position-Terminal-Covered-Barrier/dp/B005I03WOI/ref=sr_1_3?ie=UTF8&qid=1387754837&sr=8-3&keywords=terminal+block One will split the AC into three lines and 3 more will supply 6 drivers each. Should keep the wiring to a minimum.

Since the lamp is going to be flower only I plan to keep it bare bone. No dimmer, no switches. I could add a couple switches later if I feel the need.

There is one piece of advice I could use. The Vero 18s don't seem to be bin rated by lumens. Apparently they either make the cut or they don't. There are however a couple CRI options. The 80 CRI have the most lumens but are the cheapest. Is there any reason to go with a higher CRI? My understanding is that they might have a slightly better spectrum profile, but I'm not guessing it's worth the loss of luminosity.

I think you would have to look at the relative intensity and wave length (nm) chart. I think the ideal red is around 670nm, the decor put out fewer over all lums, but hit the red spectrum much much better.

http://www.bridgelux.com/wp-content/uploads/2013/10/DS32-Vero-18-Datasheet-2013.07.18.pdf

I think there is a thread around here where the math was done about it...

https://www.rollitup.org/led-other-lighting/743645-cree-cxa-analysis.html

I think I found the thread you mentioned and the absolute values indicate the high CRI profile isn't as nice as the datasheet suggests as percentages. At exactly 650nm the low CRI is still outperforming the décor. Beyond that the decors advantage doesn't seem all that big compared to the loss.

https://www.rollitup.org/led-other-lighting/743645-cree-cxa-analysis.html