3.19 umol/j

LED Teknik has just done a new round of efficiency and spectrum testing for us and we were surprised to see the 1A test report come back as 3.19 umol/j. Without blowing our own trumpet (OK, we are!) that's a pretty good efficiency figure for a "true" PAR+ full spectrum light. Especially considering there is 1% UVA and over 10% Far Red in there.

This is our Gen3 spectrum, which is 2900-3000K. The CCT changes as the current increases. Overall efficiency at each current setting is:

1A 3.19 umol/j
2A 3.01umol/j
3A 2.89 umol/j
4A 2.79 umol/j
5A 2.72 umol/j

System efficiency will be lower than LED efficiency and each of those figures falls to around 2.6-2.8 umol/j once connected to a driver that is 88-94% efficient. The lower the driver setting the more efficient the LEDs are but the less efficient the driver.

The most interesting thing about these tests was there has very little (less than 0.5%) difference between the uncoated and coated (waterproofed) panels. We don't use a typical silicon coating on our LED panels so we can apply it much thinner than 100% silicon.

We've been a bit quiet lately working on a few new things but I thought some people might be interested to know what we've been up to.

Grow Lights Australia said:

LED Teknik has just done a new round of efficiency and spectrum testing for us and we were surprised to see the 1A test report come back as 3.19 umol/j. Without blowing our own trumpet (OK, we are!) that's a pretty good efficiency figure for a "true" PAR+ full spectrum light. Especially considering there is 1% UVA and over 10% Far Red in there.

This is our Gen3 spectrum, which is 2900-3000K. The CCT changes as the current increases. Overall efficiency at each current setting is:

1A 3.19 umol/j
2A 3.01umol/j
3A 2.89 umol/j
4A 2.79 umol/j
5A 2.72 umol/j

System efficiency will be lower than LED efficiency and each of those figures falls to around 2.6-2.8 umol/j once connected to a driver that is 88-94% efficient. The lower the driver setting the more efficient the LEDs are but the less efficient the driver.

The most interesting thing about these tests was there has very little (less than 0.5%) difference between the uncoated and coated (waterproofed) panels. We don't use a typical silicon coating on our LED panels so we can apply it much thinner than 100% silicon.

We've been a bit quiet lately working on a few new things but I thought some people might be interested to know what we've been up to.
View attachment 5325509

Click to expand...

Very nice. Tweaked new spectrum with a bit more of everything. How does the numbers scale onto higher power levels? And are these available yet?
And oh please tell us about any new things youre up to

Rocket Soul said:

Very nice. Tweaked new spectrum with a bit more of everything. How does the numbers scale onto higher power levels? And are these available yet?
And oh please tell us about any new things youre up to

Click to expand...

1A 3.19 umol/j
2A 3.01umol/j
3A 2.89 umol/j
4A 2.79 umol/j
5A 2.72 umol/j

We've been working on these a long time as the first prototypes weren't going to work on the pick and place machine as they were too narrow (40mm). These new ones are 480mm x 50mm and are 2.87 umol/j at 50W, which is operating current. They are much higher efficiency at lower currents.

These are designed to mount directly to a T-slot system. We have the T-slot in stock but it is taking time to get it machined. We'll be building some smaller bar lights with these for veg, clone and lower-light species such as leafy greens and ornamental shade plants. They are designed for vertical horticulture, as the High Light panels are too powerful for confined spaces.

The strips are hard-wired and water-proof coated and have almost exactly the same spectrum as the High Light Gen3 boards, however these ones have a little more blue and slightly higher CCT at 3200K. We've stopped making the 3100K and 3400K High Light boards and now make just the boards in 2900-3000K while the first strips are 3200K. We may make another version of the strip that is lower, but 3200K is a very good alround colour temperature. In fact, anything around 3000K seems to work across the board. We don't see much stretch due to the addition of UVA.

Note the CRI is 94.5 on these. I'm sure you guys will appreciate how hard it is to get this sort of efficiency from a low CCT, high CRI LED fixture that also includes UVA and Far Red. I don't think there is much else like it on the market. We've seen a few Chinese companies try to copy our spectrum, but they always cheap out on the LEDs.

Rocket Soul said:

Very nice. Tweaked new spectrum with a bit more of everything. How does the numbers scale onto higher power levels? And are these available yet?
And oh please tell us about any new things youre up to

Click to expand...

Forgot to answer. Yes they are available now. All the new lights ship with the Gen3 spectrum.

Grow Lights Australia said:

LED Teknik has just done a new round of efficiency and spectrum testing for us and we were surprised to see the 1A test report come back as 3.19 umol/j. Without blowing our own trumpet (OK, we are!) that's a pretty good efficiency figure for a "true" PAR+ full spectrum light. Especially considering there is 1% UVA and over 10% Far Red in there.

This is our Gen3 spectrum, which is 2900-3000K. The CCT changes as the current increases. Overall efficiency at each current setting is:

1A 3.19 umol/j
2A 3.01umol/j
3A 2.89 umol/j
4A 2.79 umol/j
5A 2.72 umol/j

System efficiency will be lower than LED efficiency and each of those figures falls to around 2.6-2.8 umol/j once connected to a driver that is 88-94% efficient. The lower the driver setting the more efficient the LEDs are but the less efficient the driver.

The most interesting thing about these tests was there has very little (less than 0.5%) difference between the uncoated and coated (waterproofed) panels. We don't use a typical silicon coating on our LED panels so we can apply it much thinner than 100% silicon.

We've been a bit quiet lately working on a few new things but I thought some people might be interested to know what we've been up to.
View attachment 5325509

Click to expand...

I remember getting red/green/blue/uv/far red levels in some old tests of Teknik, is it possible to add this info here?

Also: iirc on the old highlight strips they were litterally 1/5 of a 420 board. Is this still true or have diode count gone up slightly?

Even more also: for these efficiency numbers, whats the spectral measuring range? 400-700nm? Or 380-780nm? Or other?

I must say they all look really good, the strips dont show up yet on your web, whats the pricing like?

Hey Rocket, Teknik has a new goniometer and Viso Systems is still tweaking the software, so he can't export the full reports yet. That is why I posted a photo of the screen. Once I get the full reports I will post them up.

The strips will be going up on the website very soon! We are just figuring out the pricing. They will be somewhere around AUD$60-$70 each including tax (with bulk discounts) and they can be driven at up to 70W each on a heatsink (standard T-slot) although nominally they will be 50-60W strips. So about $1 a watt. Not cheap I know but they have a water-proof coating and are pre-wired with a very robust 20AWG Australian-made insulated wire that is chemical and heat resistant (up to 130C). They are on a 1.6mm alloy PCB and as you know include UVA and Far Red. Some strips will have one set of wires and others will be double ended so that you can use a Wago connector or solder to connect each strip. Each connection will be a parallel connection so that when you wire 2 or 3 or 4 or more strips together you will have a 48V fixture which will actually be 43-47V depending on current.

The strips are more efficient than the boards when you compare the relative currents. We had to change the design of the original 480x40mm strip because it was too narrow for the pick and place machine, so we had to widen it to 50mm and decided to add another row of LEDs. They are now closer to 1/4 of a High Light board. That means four strips will easily replace a board at the same current and at a similar efficiency.

The efficiency is measured across the entire spectrum so includes roughly 390-780nm. The goniometer measures practically every photon. If you measure one of our boards with a typical PAR meter, it will under-read by about 10% because it only reads 400-700nm and we have more than 10% of out light outside that range.

Very nice mate. Thanks for posting these up. I know there were a few issues with some of the TE connectors and this new mod should solve that, but you've been good about the warranty repairs and replacements so far – even though I know it has cost you money. Sometimes you just have to work through things.

If I didn't buy HLG, I would have gotten these lights! I Love GLA's products keep up the great work, if I ever upsize I'm getting some boards from you!

Great work guys, absolutely love it!

Grow Lights Australia said:

1A 3.19 umol/j
2A 3.01umol/j
3A 2.89 umol/j
4A 2.79 umol/j
5A 2.72 umol/j

We've been working on these a long time as the first prototypes weren't going to work on the pick and place machine as they were too narrow (40mm). These new ones are 480mm x 50mm and are 2.87 umol/j at 50W, which is operating current. They are much higher efficiency at lower currents.

These are designed to mount directly to a T-slot system. We have the T-slot in stock but it is taking time to get it machined. We'll be building some smaller bar lights with these for veg, clone and lower-light species such as leafy greens and ornamental shade plants. They are designed for vertical horticulture, as the High Light panels are too powerful for confined spaces.

The strips are hard-wired and water-proof coated and have almost exactly the same spectrum as the High Light Gen3 boards, however these ones have a little more blue and slightly higher CCT at 3200K. We've stopped making the 3100K and 3400K High Light boards and now make just the boards in 2900-3000K while the first strips are 3200K. We may make another version of the strip that is lower, but 3200K is a very good alround colour temperature. In fact, anything around 3000K seems to work across the board. We don't see much stretch due to the addition of UVA.

Note the CRI is 94.5 on these. I'm sure you guys will appreciate how hard it is to get this sort of efficiency from a low CCT, high CRI LED fixture that also includes UVA and Far Red. I don't think there is much else like it on the market. We've seen a few Chinese companies try to copy our spectrum, but they always cheap out on the LEDs.

View attachment 5325781

View attachment 5325778

View attachment 5325780

View attachment 5325777

View attachment 5325779

Click to expand...

Regarding the strips: can you run strips in parallel with the boards?

Regarding sealing the strips, does that plastic layer protect the strips if we had to burn sulfur?

Rocket Soul said:

Regarding the strips: can you run strips in parallel with the boards?

Regarding sealing the strips, does that plastic layer protect the strips if we had to burn sulfur?

Click to expand...

Hi mate, sorry for the delay. Yes you can wire the strips in parallel. I'll ask GLA to run some tests on them to check the voltage, but they are similar enough that we have seen them work without any apparent current hogging. In theory, the voltages should be the same, although the boards will be fractionally more due to the extra circuits on them. When wiring in parallel, you actually have the strips in parallel with the boards, which themselves have 10x parallel circuits on them, but we use wide 2oz copper trace, so voltage drop across the boards is minimal.

Regards the sulfur burning, I would like to give you a straight and honest answer, but we haven't tried it. What I can say is that, under normal growing conditions we have seen original High Light boards with no conformal coating run for 4-5 years with no adverse effects, so the Nichia diodes do appear to have some protection in-built. The other thing I can say is we don't use a 100% sillicon conformal coating, but rather a modified coating that has a small amount of silicon (for heat cycling flex) and acrylic, that sprays on very thin but offers better optical perfomance and protection from gassing. Acrylic is not as good as epoxy potting, but it is much better than silicon, as it is no-where near as permeable. We don't get mold on our conformal coating like most other lights do, for example.

If you'd like to do some trials for us, I'm sure we could work something out and get some strips to you.

Any interest in coming out with a light that features a green spectrum like mammoth lighting has done with the newest Samsung Evo diodes?

Sickofitall420247 said:

Any interest in coming out with a light that features a green spectrum like mammoth lighting has done with the newest Samsung Evo diodes?

Click to expand...

Not really. Those Evo chips are efficient, but they also have a huge cyan gap. It is true that green and far red penetrate canopies, but for slightly different reasons. Green light is mostly reflected by leaves – it gets reflected into the lower canopy by the plant itself – although it also penetrates into deeper chloroplasts, like far red light. Far red light mainly penetrates the leaf and drives photomorphogenic response, which can make leaves bigger (more efficient photosynthesis), accelerate flowering and contribute to the Emerson Effect (which is less effective as PPFD increases).

Although green is photosynthetically more efficient than other colours, it is not as readily absorbed as blue or red, and hence a much lower percentage of it is absorbed. So it penetrates, but does not drive photosynthesis as well as other colours, which is why a rich red light produces much bigger yields than a rich green light.

The trick is to balance everything. A small amount of blue light – and violet/UVA – is important for quality, as it drives secondary metabolite production (cannabinoids, terpenes), while far red has proven effects in accelerating flowering times and increasing yields through more efficient photosynthesis. There is also the question of quantum effiency and how many photons of each colur you can produce with the same amount of energy.

The proof is in the pudding, so to speak. We feel there is enough green already in a balanced spectrum and that adding more has no real benefit, because you are potentially sacrificing an ideal spectrum in the top canopy to penetrate a bit further with green light that has less photosynthetic value. There are other ways to get light penetration, and far red – which also penetrates the canopy – has greater effects.

Have a look at the spectra below. Old-school growers (such as myself) have always, always produced better yields under HPS than MH. HPS has a lot of amber, red and some green light (but also more far red light). MH has a lot more blue and green (and UV). The old saying "HPS for yield, MH for quality" has been true for 40+ years of indoor cannabis growing.

Try putting a plant under a red light, a green light and a blue light and see which one grows best.

Seems like the mint white spectrum is the most complete at least when you look at it under the McCree spectrum.

Sickofitall420247 said:

Seems like the mint white spectrum is the most complete at least when you look at it under the McCree spectrum.

Click to expand...

Just to clarify. The mint white diodes are a specialty LED, not a complete spectrum.

Here's a graph where you can see the Samsung EVO mint white diode in blue and the EVO 5000K in black. The mint white spectrum is different, but it is not really complete.


I think you were referring to the spectrum graph posted,"Mint White Series with Emerald Green Canna Spec". This is the graph from a Mammoth light which uses a mix of diodes.

Sickofitall420247 said:

Seems like the mint white spectrum is the most complete at least when you look at it under the McCree spectrum.

Click to expand...

How do you figure that? I don't see that big green spike in the McCree Curve, but then I don't particularly want to get into a pissing match over Mammoth lighting because there is more to growing flowering plants than looking at the McCree Curve.

Do you know much about the McCree Curve? Because if you do, then you'll know it is simply a measure of photosynthetic efficiency at low light (150 umol/j), for an average of 22 common crops (not including cannabis), and does not take photomorphology into account.

Circadian rythem is controlled by cyan, secondary metabolites are affected by UVA, viuolet and blue, and shade avoidance and Emerson Effect are affected by the red to far red ratio. That's just some of what's going on.

The true test of a grow light is how it performs on your chosen crop in your particular environmental conditions – and that will change depending on the crop, light levels and other environmental factors.



FYI this is our spectrum against the McCree Curve.

Just to add to the above, in relation to horticultural LED efficiency, you also need to look at the McCree Action spectrum, which is an overlay of the photosynthetic efficiency curve compared to the energy required to produce each photon – which is what we are doing when we convert electricity to light.

The blue line is what we are more interested in when designing LED lights, but again, it does not tell the full story.

Rocket Soul said:

Regarding the strips: can you run strips in parallel with the boards?

Regarding sealing the strips, does that plastic layer protect the strips if we had to burn sulfur?

Click to expand...

Hi Rocket Soul I finally got the message to answer this from PC! It looks like he has answered to bit about burning sulfur but I took some photos of the voltages with one strip parallel wired with one board and I can confirm you can definitely wire them in parallel. I also took a photo of the strip at the other end and there was no voltage drop at all. So PC is correct that our strips and boards are over-engineered at the circuit level.

The strip and board I used were uncoated so they have not been electrically isolated. You can't do this on the conformal coated strips and boards.

There is a small difference in each circuit, as the strips have 4x 3030 LEDs in parallel for each circuit (there are two circuits on the strips) and the boards have 3x 3030 LEDs in each circuit (there are 10x circuits on each board). The ratios of LEDs are the same, so same ratio of UVA, Far Red, Red etc.

Strip at 42.90V (low current)


Other end of strip at 42.90V


Board at 42.89V

Grow Lights Australia said:

Hi Rocket Soul I finally got the message to answer this from PC! It looks like he has answered to bit about burning sulfur but I took some photos of the voltages with one strip parallel wired with one board and I can confirm you can definitely wire them in parallel. I also took a photo of the strip at the other end and there was no voltage drop at all. So PC is correct that our strips and boards are over-engineered at the circuit level.

The strip and board I used were uncoated so they have not been electrically isolated. You can't do this on the conformal coated strips and boards.

There is a small difference in each circuit, as the strips have 4x 3030 LEDs in parallel for each circuit (there are two circuits on the strips) and the boards have 3x 3030 LEDs in each circuit (there are 10x circuits on each board). The ratios of LEDs are the same, so same ratio of UVA, Far Red, Red etc.

Strip at 42.90V (low current)
View attachment 5337312

Other end of strip at 42.90V
View attachment 5337311

Board at 42.89V
View attachment 5337313

Click to expand...

This is great, it really helps with spread and spaces that dont follow standard foot measurements.

Prawn Connery said:

How do you figure that? I don't see that big green spike in the McCree Curve, but then I don't particularly want to get into a pissing match over Mammoth lighting because there is more to growing flowering plants than looking at the McCree Curve.

Do you know much about the McCree Curve? Because if you do, then you'll know it is simply a measure of photosynthetic efficiency at low light (150 umol/j), for an average of 22 common crops (not including cannabis), and does not take photomorphology into account.

Circadian rythem is controlled by cyan, secondary metabolites are affected by UVA, viuolet and blue, and shade avoidance and Emerson Effect are affected by the red to far red ratio. That's just some of what's going on.

The true test of a grow light is how it performs on your chosen crop in your particular environmental conditions – and that will change depending on the crop, light levels and other environmental factors.

View attachment 5337280

FYI this is our spectrum against the McCree Curve.
View attachment 5337307

Click to expand...

Another thing to remember is that the whole study was made under monocromatic conditions; they only tested one nm value at a time, never the efficiency of the whole spectrum put together.