LED penetration

[Samjpikey]

I definitely agree that cheaper LEDs do just fine , I have them lol


Sent from my mobile telecommunications device

 

[Troglodyte]

I think we are at cross ends here. I am not saying anything against cheap (financial cost) LEDs. I am saying that you need to be comparing the same tech spec LEDs. ie in car terms, a 1litre car with another 1 ltr car, of the same technical spec.) Most LEDs are made in China these days. Some Chinese manufacturers comply with better EU build quality criteria than others. Some LED units are cheap because they reduce the LED criteria not all. So if we are comparing 1w with 1w then there is more of an agreement. However there are no MTBF (mean time between failures) figures on the Chinese LEDs. Difficult to get them for European builds too however and that is a point that too few consider. This might be because they are too new for the aquatic products being sold. However I think it would not "look good" on the sales literature.🙄 Hopefully in a couple of months, as manufacturers of LED light units get more confident in their product, they will let us know these figures. 🙂 😉

 

[GreenNeedle]

Troglodyte - I agree with what you are saying. I was just answering the OP original post and the follow ups that suggested it might be that 'Cree' or 'Bridgelux etc were fine and that maybe cheaper ones weren't.

That's the nature of the china cheapys though. The LEDs IME are reliable. I still have 1 setup that is 6 years old with no failures which was using small unhoused drivers on DC adaptors.

Drivers can be hit and miss though. I've had those gawdy metallic painted ones fail. The ones that are gold or purple etc. Had 5 of them fail on 2 setups, each lasting between 3-6 months before they started flashing the LEDs. I assume the flashing was down to them degrading and either soaking up too much power or not letting enough through as those drivers are the sort that are also adaptors and you just plug them direct into AC. Since then I have switched to Meanwell drivers and so far (6 months in) they are OK.

 

[Troglodyte]

SuperColey1,
Ah! I got you sorry for the confusion.
It would be real handy if you could remember the manufacturers names for the drivers. As this would be useful for my crib sheet on drivers.
Did they have any warranty on them at all? As you are quite within your rights to return the product for a refund/replacement/repair if it goes faulty within one year of purchase generally, in the UK. If I remember correctly. Which driver manufacturer have you had going for 6 years and driving what lighting configuration without failure?
Any model/manufacturer information appreciated.🙂 Thanks.

 

[Samjpikey]

Troglodyte have you got access to a par meter ??



Sent from my mobile telecommunications device

 

[Troglodyte]

I have, why do you ask?

 

[Samjpikey]

If I was to send you one of my bridgelux LEDs could you measure the par ??
Would that work ??
I would love to find out how powerful my lighting is 🙂
Cheers


Sent from my mobile telecommunications device

 

[GreenNeedle]

SuperColey1,
Ah! I got you sorry for the confusion.
It would be real handy if you could remember the manufacturers names for the drivers. As this would be useful for my crib sheet on drivers.
Did they have any warranty on them at all? As you are quite within your rights to return the product for a refund/replacement/repair if it goes faulty within one year of purchase generally, in the UK. If I remember correctly. Which driver manufacturer have you had going for 6 years and driving what lighting configuration without failure?
Any model/manufacturer information appreciated.🙂 Thanks.

TBH I tend not to bother with 'warranty' on low value stuff from china. Sometimes they send you another, sometimes they say they will but don't and sometimes....you get the gist. lol.

The drivers that all failed as you can imagine are unbranded and loads of different sellers on ebay. They are the ones that say they are water proof and can run x to y amount of LEDs. The ones I bought were for 10 to 18 x 3W. All of this type of driver/adaptor have an AC input. You just wire a plug to it. I then put them into a series of 16 LEDs and they worked fine straight away. after a while (between 3-6 months) each one would start to 'flash' the lights on/off like strobes. These types are easy to spot on ebay. They have ribbed metal housings which are always coloured in a gawdy gold, purple, silver colours like this:



If I removed one section of the series and just ran 12 then they worked fine again so I assume that they had degraded and could no longer supply enough power to the 16 but were fine with the 12.

The unhoused ones I used originally were unbranded. Just a 15mm x 20mm circuit board that you can (if wanted) add a dimmer to but I didn't. They worked fine on the first LED setup I used for 2 years. They were for 12 to 24V and needed 1.5V for themselves. therefore I think they could run 2-7 x 3W. I ran them in series of 3 with a 12v supply so the LEDs running circa 2.45W each. Those LEDs are still running with the Meanwell drivers I now use which are model No LPC 35-700. They run up to 48V so I run series of 12 on them.

These were the unhoused ones I used which according to my blog was from November 2008 😉 (ignore the bad diagram, yes I put the lines to the wrong places but it was early days. lol)


EDIT: Just looking at that picture I should say that those LED stars were the original LuxeonIII copies that china were making. They have the black housing round the die which was more of a square shape and didn't fit the lense holders that they were selling. Although they were and still are fine they are not as good as the ones they started selling a year later which had/have white housing round the die in a circular shape. They do fit the lense holders 🙂 and are noticeably brighter. Both performed better than flouros but the later ones (white housing) whilst probably not being much brighter in terms of PAR or Lux are more consistent in their K so they do look brighter.

 

[flygja]

Supercoley! Great to see you back. I knew you'd return at the mere whiff of LED conversation 😉

Chinese LEDs have pretty good reliability. I personally haven't tested them, but a friend of mine did at the same time we started playing with LEDs over planted tanks and his are still kicking while I've had to replace some of my Cree LEDs. I think my problem is heat dissipation, not enough of it that is, reducing the lifespan of the LEDs. I've replaced my Luxdrive buckpucks 5-6 times, but the chinese driver I got from dx.com has been going strong for more than a year. Go figure LOL

Trog is right, Chinese LEDs don't have specs attached to them so experimentation is in order. Some "high power" LEDs are configured as multiple lower-powered LEDs combined on a single package to produce the lumens associated with high power LEDs. Like 9x 1W LEDs on one package = 10W LED.

So here's a question I have - Does a single 3W LED driven at 3W have more/less/same penetration power as 3x 1W LEDs driven at 1W each? Assume all LEDs are exactly the same and produce the same lumens/PAR when driven at the same wattage.

 

[Troglodyte]

SuperColey1,
“No Warranty required” is understandable for your own requirements and testing needs. I would assume that you would see a unit that had a warranty as an improvement over one that did not, when you find a reliable driver that has one?
Just for general information,
The term waterproof is a generic statement made by manufacturers of all persuasions when they do not know (or want to hide the facts on.) whether their unit complies with the Ingress Protection Standards (IP). These standards set out what each requirement is to meet the various levels of Ingress Protection a unit has.
IP68 Is the highest level of protection available for equipment this means that the unit passed either IEC 60529 (Enclosures) or IEC 60598-1 (Luminaires)
IP67
IP67 Is the second highest level of protection available for equipment this means that the unit passed either IEC 60529 (Enclosures) or IEC 60598-1 (Luminaires)
IP64 Is the fourth highest level of protection available for equipment this means that the unit passed either IEC 60529 (Enclosures) or IEC 60598-1 (Luminaires)
I digress, sorry.

There are two types of drivers available current and voltage, the symptoms you describe with brightness may just be due to using different types of drivers. The best ones for consistent brightness are current driven ones. As voltage driven ones can vary the current load which means the current supplied to the LED is varied resulting in light variation. However the voltage driven ones have other perceived benefits. (Which I cannot remember off the top of my head, but think it was to do with control and cost/size efficiencies.)

Strobing, can be caused by a variety of issues which I will not go into here as it depends on the driver type and model. However I will say that LEDs normally flash on and off (switching) during their normal operation. However it is normally so fast that you cannot see it happening, hence where the power efficiencies come in.

There is a generic niggle/thought in the back of my mind, that black based LEDs are possibly older varieties of LEDs using slightly older technology for production. This may be incorrect, but I believe they changed to white for the more recent builds for beneficial gains in the product. If anyone can keep me correct ? Please do so.😉

I appreciate that you obviously enjoy making the units and investigating the possibilities of what can be done, possibly for a fraction of the cost. So keep your experimentation going and keep us informed of them, as it is very interesting to hear what you are finding.

 

[Troglodyte]

Flygja,
I appreciate your input to the topic, however, I would not agree with such a general statement of
“Chinese LEDs have pretty good reliability.”

I appreciate it may have been a throw away comment and I do not wish to upset anyone, however if they were so reliable why are so many vendors having issues with LED lighting as described on this forum? Many of the light manufacturers source their LEDs from China. There is such a glut for new cheaper LEDs that many vendors in China want a piece of the market. However the problems are that as well as the good Chinese manufacturers that follow CE build standards to sell into Europe, there are also the bad vendors that do not comply to build specification in the bid to make the LEDs more cost effective. Some of these companies are creating confusion by marketing their LEDs as the same as the prior, which they are not.
This is why I personally cannot agree with such a generic comment.🙁
Your question
Does a single 3W LED driven at 3W have more/less/same penetration power as 3x 1W LEDs driven at 1W each? Assume all LEDs are exactly the same and produce the same lumens/PAR when driven at the same wattage.

This depends on several factors but the generic answer is no.
A 3W LED normally has the capability to be driven at a variety of currents. (ie 350ma/700ma/1A). 3 x 1W cannot normally be driven at the higher currents. If it is it will generally burn out.
So it is unlikely that any 1W LED, no matter how many, will be able to produce the same light as one driven at a higher current.(although there is a limit beyond which, no matter how much current used, the light output will not change.)
They certainly will not penetrate anywhere near as much depth of water as a 3W/10W. The deeper the tank the higher wattage LEDs will be required.
However a shallow tank less than 15” and a low PAR level desired at substrate level will be met by multiple low wattage LEDs. Hence the variation in the aquarium LED lighting market using air and water PAR measurement. Water PAR measurements are always superior to Air PAR measurements.
Hope this helps?

 

[GreenNeedle]

My experimentation tends to veer towards what Flygia says. I have zero problems with the LEDs. Even those black ones. Not had a single failure with them. Like I say the only problems I have had have been with drivers.

I wouldn't call it experimentation anymore. The first one was but since then I just use what the same as back then (albeit with the newer style LEDs) and do exactly the same thing. Something I tried (was the first time I used a soldering iron 🙂 ) I could go into higher power wattages and less LEDs but with my planted tank head on I am vehement about the need for a nice spread of light so 16 x 3W spread out is better in my eyes than 5 x 10W or 1 x 50W. I like to use a grid formation rather than squeeze things into a slimline retail'esque' sleek unit. I think that is where they fail although of course it is pandering to consumer preference in terms of aesthetics.

Flygia, on that first build I was following what others had been saying r.e. heat issues and built fans into the box to cool them. That only lasted for about a month or so before I turned the fans off. I have always used heatsinking and 'touch wood' that seems to have been enough. I never take the 'star is a heatsink' route. I always use thermal adhesive paste onto heatsink and that seems to do the job fine. That has been for the first 2 builds which were enclosed in boxes and the current one which is open.

Glad I got interested in this thread as it spurred me on to carry on re-doing my lounge and when that is finished I can set up my opti-white with the new luminaire which has been standing unused in the lounge since April last year (much to the wife's unrivalled joy.....oooops)

 

[Samjpikey]

.

 

[flygja]

I appreciate your input to the topic, however, I would not agree with such a general statement of “Chinese LEDs have pretty good reliability.” I appreciate it may have been a throw away comment and I do not wish to upset anyone, however if they were so reliable why are so many vendors having issues with LED lighting as described on this forum? Many of the light manufacturers source their LEDs from China. There is such a glut for new cheaper LEDs that many vendors in China want a piece of the market. However the problems are that as well as the good Chinese manufacturers that follow CE build standards to sell into Europe, there are also the bad vendors that do not comply to build specification in the bid to make the LEDs more cost effective. Some of these companies are creating confusion by marketing their LEDs as the same as the prior, which they are not. This is why I personally cannot agree with such a generic comment.

I agree with you that chinese made electronics in general may not meet all the FCC/CE/C-tick/BSMI/VCCI requirements in other countries. And there is a huge amount of brands and models and bins and lot codes of LEDs on offer. My statement is mine alone, and its based on my experiences and those of people I know who have used them. Your mileage may vary. Of course if you've got the extra resouces, go for something branded with proper datasheets - Osram, Cree, Seoul Semiconductor, Luxeon, Lumileds, etc. You get what you pay for, but sometimes there are gems waiting to be uncovered. For e.g - using Akadama/cat litter instead of expensive designer soils. Sink scrubbies as filter media vs established branded ceramic/sintered glass. EI vs designer ferts. All down to experimentation.

A 3W LED normally has the capability to be driven at a variety of currents. (ie 350ma/700ma/1A). 3 x 1W cannot normally be driven at the higher currents. If it is it will generally burn out. So it is unlikely that any 1W LED, no matter how many, will be able to produce the same light as one driven at a higher current.(although there is a limit beyond which, no matter how much current used, the light output will not change.)

Let me rephrase my question. Assume you have a 3W LED, when driven at 1A will produce 2100 lumens. You also have a 1W LED, when driven at 300mA, will produce 700 lumens. The drive currents are assumed to be at maximum. So if you pit a 3W LED and 3 of the 1W LEDs together, they all make the same lumens - 2100 at max drive current. So which one will have greater penetration. Also assume that the 1W LEDs are placed closed enough together to act as a single light source. Both 3W and 1W LEDs have the same spread. Do you get my question now?

I don't know about the black-base vs white-base LEDs. Isn't it just soldermask colour selection?

Flygia, on that first build I was following what others had been saying r.e. heat issues and built fans into the box to cool them. That only lasted for about a month or so before I turned the fans off. I have always used heatsinking and 'touch wood' that seems to have been enough. I never take the 'star is a heatsink' route. I always use thermal adhesive paste onto heatsink and that seems to do the job fine. That has been for the first 2 builds which were enclosed in boxes and the current one which is open.

Maybe its better in the UK but I live in the tropics where temperatures regularly reach 32-34'C in the afternoons. The star isn't really a heatsink, its a heat spreader 😀

 

[GreenNeedle]

Let me rephrase my question. Assume you have a 3W LED, when driven at 1A will produce 2100 lumens. You also have a 1W LED, when driven at 300mA, will produce 700 lumens. The drive currents are assumed to be at maximum. So if you pit a 3W LED and 3 of the 1W LEDs together, they all make the same lumens - 2100 at max drive current. So which one will have greater penetration. Also assume that the 1W LEDs are placed closed enough together to act as a single light source. Both 3W and 1W LEDs have the same spread. Do you get my question now?

This is subjective really. A single point source will shoot further because you have the whole '2100' lumens in one point source where the others are apart. That is if it was a single vs 3. I would suggest however if you were only doing a 9W version for a Nano which we would assume is pretty shallow (under 30cm) then I prefer to use 9 x 1W than 3 x 3W purely to spread the light out better. As you start using more LEDs like 30W+ I think it cancels the spread issue out so its more sensible to wire 10 LEDs @ 3W than wire up 30 x 1W plus it would likely mean that if you were going to use 30W you are using a bigger larger tank.

Your scenario would seem to be the approach that LED household bulbs are using. My whole house is LED now and all these bulbs tend to have multiple LEDs in a single housing. however from simple maths I would assume that my 8W bulbs that have 5 LEDs in them are actually underpowering 3W LEDs??? Either way they are pretty good. My kitchen light which was made for 5 x 40W GU10 halogens (200W aargh) now has 5 x 6W GU10 LEDs in it and is brighter plus they are daylight and not the slightly yellow warm white of halogens. Plus 30W is much better than 200W for a room where the light is on a lot of time. I should add that all these LED bulbs in my house are also china generics, plain packaging, no datasheets but work perfectly so I can't tell you a brand or anything. I don't use 'energy saving' lamps anymore because they use too much energy which is very expensive now in the UK.

However in this scenario it is cheaper to buy a 9 x 1W Beamworks luminaire for £30 than make your own...........which is what I have done 🙂 The high power beamworks units are much better than their predecessors which were using massive amounts of low power LEDs and their only failing are the lenses which are too narrow a beam for their application. They are supposedly 'replaceable' but they are much shallower than the ones you buy on the net so I removed the lenses and the 2mm sheet that held them and replaced them with a small 2mm clear acrylic sheet...no lenses:

It's plenty bright enough over this tank and being honest (shh, don't tell anyone) I find it too powerful for this 30 litre and only use the 9 for photos. It has a switch on the end that toggles 3LEDs/Off/9LEDs and I leave the switch on night mode that it is only using the central 3 normally and that is much better. Still bright to the eyes but more suited in terms of light to the non CO2 tank it is over.

I don't know about the black-base vs white-base LEDs. Isn't it just soldermask colour selection?

The only differences I notice:
The ring around the LED that the die sits on is black instead of white.
The 'ring' on the black is squared and on the white is circular.
The die on the white ones is consistently crystal clear where the one on the black ones was not so consistent. Some where crystal clear but some yellowing.
I think they either changed the actual material of the die or the manufacturing process from the black one to the white one because of the above and also that I received 2 black ones from my original order of 40 where the die was still soft and cloudy. Like a clear silicon that is cured.
The markings +/- on the black ones was pretty poorly 'printed' often hard to tell the difference between the + and - markings where the newer white ones are very clearly 'printed'.

I assume from the above it was simply that the early ones were crudely made and as they got experienced they got better equipment / processes and were learning as they made them. I have had no failures once wired from either those first black ones or the white ones other than mistakes I made when wiring them. lol. First soldering job I did and of course wasn't the best. melted a few LEDs whilst learning on the job.

Maybe its better in the UK but I live in the tropics where temperatures regularly reach 32-34'C in the afternoons. The star isn't really a heatsink, its a heat spreader 😀

I would guess it's more the humidity than the temperature. I wouldn't think an ambient difference of 10-15C difference between my home and yours would make much difference when these get too hot too touch. The heatsinks are very important though 🙂

 

[flygja]

Your scenario would seem to be the approach that LED household bulbs are using. My whole house is LED now and all these bulbs tend to have multiple LEDs in a single housing. however from simple maths I would assume that my 8W bulbs that have 5 LEDs in them are actually underpowering 3W LEDs???

This is pretty normal in order to meet the lifespan targets of LEDs. More to do with heat dissipation I believe, but I could be wrong.

The reason why I asked the penetration-per-watt question is because I'm thinking of re-doing the LED fixture on my 300L. It currently has 10x 10W Cree LEDs over it. I'm not running it a full power, just enough to get 50 umol/s over the subtrate. I've been wondering if its better to mix 3W and 10W LEDs (different drivers, more wiring) or just stick with all 10W LEDs (easier wiring, lesser spread). No optics on this but the C-shaped aluminium channel I'm using tends to narrow the beam slightly.

LEDs for nanos and shallow tanks are easier as they don't have to be high powered, i.e. don't need to be more than 1W per LED. Its when things get deeper and you get areas blocked off by hardscape where I worry about penetration and spread.

 

[GreenNeedle]

Indeed. Personally as you are using 10 x 10W LEDs I would stick with that. I think you have to be really confident with your work to do more wiring. That or seperate it into several series rather than one string. Fault finding is easier that way.

 

[dmachado]

I have a 300L aquarium, 60 cm tall, and a 12 led DIY hood, underdriven 10w leds at about 30%-35% current. I fitted glass lenses to all leds, and at 80-90W the light is unbelievable.

I never tried it at 120W, I run it at about 35W now.

My leds & lenses can penetrate more than 2ft of water, I can guarantee you that.