COB efficiency Spreadsheets
- Thread starter nogod_
- Start date
Spuzzum
Well-Known Member
Yes.. yes I do. It means the Mean Wells are 94% efficient at converting 110 ac to 1400mA dc. But they're the same efficiency at 80%.
If they calculate the lifespan of the product at 80% load, then you can be assured they calculated the 94% efficiency at the same load.
Look at the specs of your pc's psu.. it may say 650w or whatever, but if you add up all the 3.3, 5, and 12v rails.. at whatever current available to them... you come up with a bigger supply than what the manufacturer's labeled it as. Usually about 20% bigger. Hell.. just surf any serious diy pc forum.. they all say to give yourself headroom when buying the psu. Not because you may need more power down the road, but because of the safety factor.
When you buy components.. ie: resistors and capacitors, they're mass produced. They come with a certain +/- percentage beside their power ratings.. some say +/- 10%, while better quality components will be +/- 1%. That means some pieces in the assembly line get a little more of the active ingredients needed, and some get a little less. It's the ones that get a little less you have to worry about. Pushing them at 100% their rating will definitely make them die early. That's why if the circuit has a 1/4w signal, you use a 1/2w resistor. It's a safety factor.
nogod_
Well-Known Member
While I agree with your logic, any component has a chance of failure at any percent load. If you are that worried about your components failing and the potentially massive loss that could ensue, you'll have a backup on hand anyway. My drivers are maxed and remote and are easily swapped and ive had zero issues. I also have a few replacements left over from cannibalising other projects.
In any event, i dont think its bad advice to tell someone to use up their headroom. In my case I can buy 3 extra hlg-185h-c700b's for the money I saved by not doing it your way. If the drivers fail I can swap them in 5 minutes and get a replacement sent to me.
I'm going to keep doing it my way for every project here on out, but to each their own!
In any event, i dont think its bad advice to tell someone to use up their headroom. In my case I can buy 3 extra hlg-185h-c700b's for the money I saved by not doing it your way. If the drivers fail I can swap them in 5 minutes and get a replacement sent to me.
I'm going to keep doing it my way for every project here on out, but to each their own!
Spuzzum
Well-Known Member
Fair enough. I'd rather not come home to find the garden's been dark for 10 hours, only to have to fix it after the fact.
Talk to any electrician.. 20% headroom is important. It's not fantasy. If we have the cash to pay $50 a bulb, then a few extra dollars per driver isn't that hard to come up with. I know more than most how hard it is to pay for this shit.. I'm on disability.. I only get $906 a month, living in the most expensive city in Canada. When I bought my Cree XPE/XPG's in 2011, I made monthly purchases.. a little bit each cheque. It took me nearly a year to pay for it all, but I did it. I could have gone with ChinaCrap, but I wanted the quality. I bucked up.
Spuzzum
Well-Known Member
Reliability Basics
Derating for Electronic Components
Derating is a technique usually employed in electrical power and electronic devices, wherein the devices are operated at less than their rated maximum power dissipation, taking into account the case/body temperature, the ambient temperature and the type of cooling mechanism used. In this article, we will briefly explain the theoretical background of derating and how it is applied.
Derating increases the margin of safety between part design limits and applied stresses, thereby providing extra protection for the part. By applying derating in an electrical or electronic component, its degradation rate is reduced. The reliability and life expectancy are improved.
Intuitively, if a component or system is operated under its design limit, it will be more reliable than if it is operated at or above the design limit. Theoretically, the benefit of derating can be explained using load-strength interference theory.
Load-Strength Interference
Usually, failure happens when the applied load exceeds the strength. Load and strength should be considered in a general way. For electronic parts, "load" might refer to voltage, power or an internal stress such as junction temperature. "Strength" might refer to any resisting physical properties.
Electronic components of a given type are not identical. They have strength variability. This variability results from the differences between raw materials and between manufacturing processes. Even for components made from the same materials and by the same processes, differences still exist due to noise factors such as microscopic material defects or variations within a single manufacturing process. Therefore, the strength of a component is considered to be a random variable. The load applied to electronic parts, such as power, temperature or humidity, is also a random variable. Thus, statistical distributions are usually used to describe the load and strength.
http://www.weibull.com/hotwire/issue92/relbasics92.htm
If I remember correctly, it also has to do with the ac power coming from the main circuit... it's not completely smooth.. it spikes here and there. If your driver is maxed out when a spike occurs, this can damage the unit. No headroom to accommodate the spikes.
Safety first.
Derating for Electronic Components
Derating is a technique usually employed in electrical power and electronic devices, wherein the devices are operated at less than their rated maximum power dissipation, taking into account the case/body temperature, the ambient temperature and the type of cooling mechanism used. In this article, we will briefly explain the theoretical background of derating and how it is applied.
Derating increases the margin of safety between part design limits and applied stresses, thereby providing extra protection for the part. By applying derating in an electrical or electronic component, its degradation rate is reduced. The reliability and life expectancy are improved.
Intuitively, if a component or system is operated under its design limit, it will be more reliable than if it is operated at or above the design limit. Theoretically, the benefit of derating can be explained using load-strength interference theory.
Load-Strength Interference
Usually, failure happens when the applied load exceeds the strength. Load and strength should be considered in a general way. For electronic parts, "load" might refer to voltage, power or an internal stress such as junction temperature. "Strength" might refer to any resisting physical properties.
Electronic components of a given type are not identical. They have strength variability. This variability results from the differences between raw materials and between manufacturing processes. Even for components made from the same materials and by the same processes, differences still exist due to noise factors such as microscopic material defects or variations within a single manufacturing process. Therefore, the strength of a component is considered to be a random variable. The load applied to electronic parts, such as power, temperature or humidity, is also a random variable. Thus, statistical distributions are usually used to describe the load and strength.
http://www.weibull.com/hotwire/issue92/relbasics92.htm
If I remember correctly, it also has to do with the ac power coming from the main circuit... it's not completely smooth.. it spikes here and there. If your driver is maxed out when a spike occurs, this can damage the unit. No headroom to accommodate the spikes.
Safety first.
Spuzzum
Well-Known Member
lol... you have bigger balls then me then.
Like I say.. an electronics guru helped me design my drivers back in 2011. The main point he stressed was that 20% headroom. Even when it comes to wiring.. you don't use a part that's going to be maxed out. You use something bigger.. for the safety measure.
Changing a driver is no big deal. Letting your light schedule get messed up.. even just for 1 day... is not something most growers would want to endure. No.. it doesn't mean automatic hermies... but why risk it? I'd rather pay a few extra bucks and have that safety net. Peace of mind.
cheers.
BM9AGS
Well-Known Member
Do you know the capacity that hid lights like gavita are designed at off top of your head? I've not heard of a failed meanwell.....other than human induced problemslol... you have bigger balls then me then.
Like I say.. an electronics guru helped me design my drivers back in 2011. The main point he stressed was that 20% headroom. Even when it comes to wiring.. you don't use a part that's going to be maxed out. You use something bigger.. for the safety measure.
Changing a driver is no big deal. Letting your light schedule get messed up.. even just for 1 day... is not something most growers would want to endure. No.. it doesn't mean automatic hermies... but why risk it? I'd rather pay a few extra bucks and have that safety net. Peace of mind.
cheers.
Spuzzum
Well-Known Member
I have no idea of hids, other than the old style models... 1000W used 1140W total. Or something like that.
As for Mean Wells dying.. they do, and they have. The reef builder community uses these drivers.. although they do max out the voltage, they drop the current to the string. 1400mA drivers get turned down to 1000mA or so. And yes.. there's been a few dead driver posts over the years.
I even mentioned that growmau5 had said he had a driver blow on him.. or die, anyway. He didn't say why it dies.. just that he opened it up afterwords, but couldn't see why. It was in Greengene's tech talk series on youtube... episode 3 if I remember correctly. They're 60+ minutes each.. I didn't feel like searching through it again.
If it's not in that video, it'll be in episode #2. Pretty sure it was #3 though.
ttystikk
Well-Known Member
Sometimes that headroom is essential. An example is the power circuit for main lighting, so when building my systems I'm mindful to leave some headroom. For my 5400W rack, I use a 30A circuit at 240V.
I'm not an electronics expert here or anywhere else, so I know that I don't have the expertise to weigh in on the debate about whether my drivers need extra headroom. I'd like to see hard numbers on relative durability under constant loads.
Spuzzum
Well-Known Member
Exactly. A licensed electrician will make sure your breaker box has breakers bigger than the load you'll actually need. Not just so the circuit won't trip.. but for safety reasons. It just makes sense. The way I look at it... would the fixture pass an electrical inspection? I'm licensed, but I'm subject to inspections. Using a diy panel is bad enough.. even worse if it doesn't pass inspections.
I'm watching that diy tech talk #3 again.. I'll find the spot growmau5's driver died on him. I can't even remember if Greengene or robincnn mentioned if they've had a driver die on them as well or not.
ttystikk
Well-Known Member
I give 20% headroom on my power circuits, but that's rather a different load than a COB LED chip needs. I have a lot of these drivers and if there was a problem with running them near 100%, I'm quite sure I'd have seen it by now.
Spuzzum
Well-Known Member
Power usage is power usage. It doesn't matter if it's the main breaker, or an individual power source. And as I said.. growmau5's lost a driver, and a quick google shows a thread on this very site, where 2 36v CXB3590's were driven by a Mean Well ELG-100-C1400B... maxed out at the 72v/1400mA... his driver died within 2 months.
https://www.rollitup.org/t/led-drive-elg-100-c1400-dead-wrong-setup.902504/
It happens. I'd rather prevent it beforehand, rather than deal with it after the fact.
BOBBY_G
Well-Known Member
No, a licensed electrician would select breakers based on the conductor size they are designed to protect
Spuzzum
Well-Known Member
Fair enough. But a licensed electrician would also give themselves that 20% headroom when selecting a power supply.
BM9AGS
Well-Known Member
So.....off topic of all this electrical mumbojumbo.... If I wanted to build a 3500k 70% efficient cob light.... What would y'all choose?
For a 5'x5' area. Max 750ppfd. Cooled on computer waterblocks..... I was thinking to buy like 30 of Robincnn clu0 58. That's the latest cob right?
I say 750ppfd of cobs and supplement it with 200w of Osram ssl80 12 LED chips of 660nm. For flower. For veg there would be 200w Osram ssl80 450nm... Cobs dimmable of course because for veg with the 440nm only 400 watts would be needed.....based off my experience.
For a 5'x5' area. Max 750ppfd. Cooled on computer waterblocks..... I was thinking to buy like 30 of Robincnn clu0 58. That's the latest cob right?
I say 750ppfd of cobs and supplement it with 200w of Osram ssl80 12 LED chips of 660nm. For flower. For veg there would be 200w Osram ssl80 450nm... Cobs dimmable of course because for veg with the 440nm only 400 watts would be needed.....based off my experience.
Last edited:
REALSTYLES
Well-Known Member
It was because he over ran it. It maxed out at 72v and he had 144 so of coarse there was failure. It's all good do yo lol although your wrong that guy made a newbie and paid for it a lot of guys have. I've had PM from people telling me they wired it in reverse power in to the out and visa versa. I've been running my HLG not the ELG for a year and a half 100% no problems. I've made 16 lights and in the process of making a 500w CXA/2 3590 Studio 5700k BD 36v and using a HLG-240H-C1400A with reflectors
Spuzzum
Well-Known Member
36v * 2 = 72v. The supply was 72v @ 1400mA. He maxed it out, he didn't over run it. In fact, the 3590's use less than 36v @ 1400mA. They use 34.84v... x2 comes to 69.68v.
REALSTYLES
Well-Known Member
my bad thought it was 72v lol but still he was over at that current thus dead driver I just went back and reread the whole thread all 6 post lol as for the PCT it's not that accurate it has discrepancies just ask the led guru's I'll just give Growmau or Supra a call and ask.
Spuzzum
Well-Known Member
The driver was rated from 35-72v @ 1400mA. 1/2 the driver most use for their 4 COB strings. So.. if you think he over ran his driver, then so's everybody else using their 144v drivers @ 1400mA.
As for asking guru's.. I had an electrical engineer help me with my own diy drivers... someone who actually has his ticket. I used to mod some hack sites back in the d!shnet days, and he was a resident hardware specialist. I'd trust his word over a diy guy any day.
If you don't believe me.. ask any of your local electronics shops.. they usually hire people who know what the fuck they're talking about.