need a good COB

The efficiency goes up slightly as current goes down, but some cobs are also just more efficient than others as well... I don't know why.

The driver efficiency is independent of the cob efficiency..

If you have a 90% efficient driver and a 38% efficient led array, the efficiency of the entire system is 34.2%. (0.38 * 0.9)

The bins for cxa are actually absolute. AB always means the same amount of lumens. It's not relative to a certain type or color temperature of cxa, but absolute to the entire cxa line.

So any suggestions on a substitute if the vero cannot be run by itself? I would like the best possible option for the driver thanks

https://www.rollitup.org/t/c-o-b-s-s-l-news-and-articles.848298/
Have you read this thread. You don't have to read all of it. Lots of COBs to choose from. Some reading won't maim or kill ya . I think the 3070 will run, maybe some CXMs.

churchhaze said:

Actually, the vero 18 might not work because of the 30V minimum. At 650mA, they'd probably drop more like 29V.

Maybe there's another cob between the voltage range.

Edit: You could use monos to drop a few extra volts dropped!! problem solved!

If you add just 1 XP-E photo red to the series string, it will drop enough volts to operate within the voltage range!

Click to expand...

Maybe I'm too stoned nowadays but I'm pretty sure that's not gonna work with that driver. The COB will still want 29 and you would be running the driver at max capacity and it will almost certainly blow the driver or an LED after a bit. So if you put a mono in front of the COB and the mono uses let's say 4 volts there's 30-42 volts available to the COB with this driver, not 26 - 46 volts available. It can't do 1 through 29 volts just 30 - 46 volts DC and will do it's best to supply that COB 30 volts till something in the circuit fails. I'm pretty sure it works that way in this case Church.

FranJan said:

https://www.rollitup.org/t/c-o-b-s-s-l-news-and-articles.848298/
Have you read this thread. You don't have to read all of it. Lots of COBs to choose from. Some reading won't maim or kill ya . I think the 3070 will run, maybe some CXMs.


So if you put a mono in front of the COB and the mono uses let's say 4 volts there's 30-42 volts available to the COB with this driver, not 26 - 46 volts available. It can't do 1 through 29 volts just 30 - 46 volts DC and will do it's best to supply that COB 30 volts till something in the circuit fails.

Click to expand...

An excellent explanation. If what you say is true, then maybe Thorhax should consider playing with the V29?

http://www.digikey.com/product-search/en?v=976&s=44064&FV=fff40008,fff80354&mnonly=0&newproducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25

It could handle the voltage. Do COBs typically have a current minimum? I always see a max but never a minimum.

Yes they do but usually you just see the typical and the max voltages with LEDs and if you look at something like this graph you'll see a minimum. Under-driven LEDs are probably not considered dangerous to the die and really you just need to know the maximum voltage but I guess designers would want to know more.

If the mono drops 4V at 650mA and the vero 18 drops 29V, the voltages add since they're in series. So the series combination of that mono + the vero 18 would drop 33V with a draw of 650mA, which would put it in the 30-42V supported by the driver.

FranJan said:

Maybe I'm too stoned nowadays but I'm pretty sure that's not gonna work with that driver. The COB will still want 29 and you would be running the driver at max capacity and it will almost certainly blow the driver or an LED after a bit. So if you put a mono in front of the COB and the mono uses let's say 4 volts there's 30-42 volts available to the COB with this driver, not 26 - 46 volts available. It can't do 1 through 29 volts just 30 - 46 volts DC and will do it's best to supply that COB 30 volts till something in the circuit fails. I'm pretty sure it works that way in this case Church.

Click to expand...

Yeah they add but how does it change the COBs profile? So the voltage needed for the circuit would be 4Vf + 29Vf = 33 volts but the COB will still want 29 volts and get 30 instead. I don't see how the mono changes the fV of the COB at 650mA or any mA at that matter? I understand I'm not Mr Electrical Engineer and might just be dead wrong but I just can't get my head around what you're saying.

The voltage across the cob is dependent on the current running through it. If you put 650mA through it, the cob's characteristic curve will determine the voltage drop across the cob at that current.

The driver will vary the voltage between minimum and maximum in attempt to keep current at 650mA, and since the series combination of the cob and the mono would be between the min and max supply output voltages, it would be able to maintain the current at 650mA.

FranJan said:

Yeah they add but how does it change the COBs profile? So the voltage needed for the circuit would be 4Vf + 29Vf = 33 volts but the COB will still want 29 volts and get 30 instead. I don't see how the mono changes the fV of the COB at 650mA or any mA at that matter? I understand I'm not Mr Electrical Engineer and might just be dead wrong but I just can't get my head around what you're saying.

Click to expand...

I've actually seen multiple points presented throughout various threads against using different COBs and stars under the power of the same driver. Something to do with the dissimilar amperage or current of each array not playing together fairly.

I'd advise against using a random mono/star or whatever in conjunction with a high-powered COB.

Also, I believe Fran Jan makes their point clear that, regardless of voltage draw from series, the driver's range goes no lower than 30V. However, after weighing both arguments in my head, it seems reasonable to think, theoretically, that if the range for a driver is 30V-40V, then 33V will be supplied to a combination of 4V mono in series with a 29V Vero 18 COB.

Here is some experience for you guys, take what you can from it. When I bought my IR stars, I tried a few things out for the hell of it. The stars ran at a max of ~1.8V each with a current around 600mA. As far as a driver, I chose one that went 3.2-4V at a current max of 700mA. I wasn't paying attention to these numbers at first, mainly because they advertised the stars as 3W, while the driver was made for 1x3W (StevesLED). To make the story shorter, I powered just one star. I instantly noticed black smoke emitting from the star and could smell something cooking. Then once I ran the 2x stars in series, no smoke. I ran the numbers with my meter and lone behold the series ran around 3.6V. This example demonstrates that the range of a driver can be manipulated using a series connection. It should be pointed out that the stars were of the same make and model and all. I believe it's clear who won the debate .

A great discourse to follow guys; I had to put on and take off my thinking cap a few different times haha.

Look up KVL, KCL, thevenin, norton, superposition, and really just circuits analysis in general. There are a few basic analysis principles needed to analyze circuits. Simulation techniques (transient, DC sweep, AC sweep, etc) are also a great sanity check on your design/analysis.

Take a look at how diodes work. It goes way beyond light emitting diodes.

It's really hard to treat such a hard science in a debate setting.

churchhaze said:

Look up KVL, KCL, thevenin, norton, superposition, and really just circuits analysis in general. There are a few basic analysis principles needed to analyze circuits. Simulation techniques (transient, DC sweep, AC sweep, etc) are also a great sanity check on your design/analysis.

Take a look at how diodes work. It goes way beyond light emitting diodes.

It's really hard to treat such a hard science in a debate setting.

Click to expand...

Did you go to college to learn these things or were these topics, mentioned above, things you came across while working with LEDs?

It all comes down to experience mate. I'd still be pondering on the theoretical possibilities if it weren't for my experience with those IR stars. I don't know if Fran has any experience with LEDs but I now know that my initial, minor mistake with the IR stars was a great learning lesson that helped prove your case in today's debate/discourse.

Cheers!!

churchhaze said:

The voltage across the cob is dependent on the current running through it. If you put 650mA through it, the cob's characteristic curve will determine the voltage drop across the cob at that current.

The driver will vary the voltage between minimum and maximum in attempt to keep current at 650mA, and since the series combination of the cob and the mono would be between the min and max supply output voltages, it would be able to maintain the current at 650mA.

Click to expand...

Church is correct, this is how it works. Multiplication can be wizardry to some (lol) but its no mystery, that's a nice graphic by the way.

AquariusPanta said:

I've actually seen multiple points presented throughout various threads against using different COBs and stars under the power of the same driver. Something to do with the dissimilar amperage or current of each array not playing together fairly.....

Click to expand...

You would be correct about running COBs and stars on the same string if we were talking about higher amperes. Say we had a 1400mA driver with a CXA3070 and a Cree XP-E photo red star together on a string, the XP-E(s) will burn out very fast. That is because the photo red XP-E's have a maximum current of 700mA. Since we are talking about drivers with a current of only 650mA you would be safe running them both on the same string with proper heatsinks. I think you need to read this, it was made a sticky for a reason. 30+ pages of good information, most of it is on the first few pages.

http://rollitup.org/t/diy-leds-how-to-power-them.801554/

Basics...
Series Circuit
Parallel Circuit
Combination Circuit

... the law behind

...and there is no way to break this law

using this law on a cob at 1A and 4 deep red at 500mA
driven by a Constant Current-Source

simple current balance is done here by the 2 resistors.

These are things I learned in EE school in a formal setting. I released my first diode smoke in a lab setting. lol.

Just remember, there's no way to put the smoke back in the diodes. I don't expect people to take a circuits analysis course just to learn to make lights, but it will answer definitively questions about design.

Simulation with a tool like LTSpice (free) is easy mode. (put your circuit in the simulator until it works, then build it irl)

AquariusPanta said:

Did you go to college to learn these things or were these topics, mentioned above, things you came across while working with LEDs?

It all comes down to experience mate. I'd still be pondering on the theoretical possibilities if it weren't for my experience with those IR stars. I don't know if Fran has any experience with LEDs but I now know that my initial, minor mistake with the IR stars was a great learning lesson that helped prove your case in today's debate/discourse.

Cheers!!

Click to expand...

epicfail said:

You would be correct about running COBs and stars on the same string if we were talking about higher amperes. Say we had a 1400mA driver with a CXA3070 and a Cree XP-E photo red star together on a string, the XP-E(s) will burn out very fast. That is because the photo red XP-E's have a maximum current of 700mA. Since we are talking about drivers with a current of only 650mA you would be safe running them both on the same string with proper heatsinks. I think you need to read this, it was made a sticky for a reason. 30+ pages of good information, most of it is on the first few pages.

http://rollitup.org/t/diy-leds-how-to-power-them.801554/

Click to expand...

What you included makes great sense. I don't think the first page of "How to power them" makes a better example than the one you brought forth.

churchhaze said:

These are things I learned in EE school in a formal setting. I released my first diode smoke in a lab setting. lol.

Just remember, there's no way to put the smoke back in the diodes. I don't expect people to take a circuits analysis course just to learn to make lights, but it will answer definitively questions about design.

Simulation with a tool like LTSpice (free) is easy mode. (put your circuit in the simulator until it works, then build it irl)

Click to expand...

I'm currently in a ECE program but have yet to take any EE classes. Instead their making me learn intro C++ this semester, which sucks. Probably going to drop out after Spring term lol.

Thorhax said:

So any suggestions on a substitute if the vero cannot be run by itself? I would like the best possible option for the driver thanks

Click to expand...

vero 29 or cxa 3070, there is still room for a couple of deep red (660nm) leds in series, which is beneficial cause it maximizes driver efficiency and is good for flowering.

Isn't intro to C++ just learning to use if() and while()?

I mean if you're doing good in your math and science courses, stay in ECE. If you hate those courses the most, change your major. lol.

If they aren't having you take math and science courses yet, find a new ECE school!

AquariusPanta said:

What you included makes great sense. I don't think the first page of "How to power them" makes a better example than the one you brought forth.



I'm currently in a ECE program but have yet to take any EE classes. Instead their making me learn intro C++ this semester, which sucks. Probably going to drop out after Spring term lol.

Click to expand...

http://www.digikey.com/product-detail/en/CXA1304-0000-000C00A430F/CXA1304-0000-000C00A430F-ND/4173027

Would these work?

5 of those per circuit would be 45v at 650mA= 29.25watts per circuit

Sound ok?

churchhaze said:

Isn't intro to C++ just learning to use if() and while()?

I mean if you're doing good in your math and science courses, stay in ECE. If you hate those courses the most, change your major. lol.

If they aren't having you take math and science courses yet, find a new ECE school!

Click to expand...

The if and while functions were the first four weeks. I don't even know what were doing now.

I've already earned my associates in mechanical and civil engineering technology and enjoy most math and science courses but wanted to give ECE a shot. This new school's engineering department is highly dysfunctional and the teacher I have for the previously mentioned C++ class is easily the worst teacher I've had during my college career.

I'd be concerned about is that they'd be running at about 120% of the test current at 650mA. I'm not sure how efficient that would be, but it might still be pretty good. Maybe Supra would have an idea of their efficiency at 650mA.

Thorhax said:

http://www.digikey.com/product-detail/en/CXA1304-0000-000C00A430F/CXA1304-0000-000C00A430F-ND/4173027

Would these work?

5 of those per circuit would be 45v at 650mA= 29.25watts per circuit

Sound ok?

Click to expand...