"High Lights" - high powered, CRI95 flowering boards made in Australia

Prawn Connery

Well-Known Member
A few people have been asking about these boards now that the cat's out of the bag. @Or_Gro has posted his diary here for anyone who's interested in seeing how they perform: https://www.rollitup.org/t/the-monumentally-epic-knockdown-dragout-take-no-prisoners-slapdown-aussie-high-light-vs-hlg-288.988144/

But for now, these are my plants (Mental Floss by Chimera - lovely smoke, BTW):
IMG_0967.JPG

I thought I'd better post a thread explaining what "High Lights" are and how (and why) they're different to most other boards on the market. These boards are designed and assembled in Australia and use Japanese-made Nichia and Korean Seoul Semiconductor UV white phosphor 3030 LEDs.

The boards were designed by growers for growers and are purely flowering boards - although we have been using them in veg and, apart from a bit of extra stretch, they have been performing well.

Basically they are a 4.5A @ 50.5V board that can be driven as high as 225W+ with matching heatsinks, but do not require heatsinks at 3A or below (around 150W). This is due to the large platform and LED spacing.

The boards measure 415mm x 205mm x 2mm (aluminium) and have 450 LEDs. When paired together, they form a perfect square of 415mm x 410mm, or 16" x 16" in the old money, with a combined 900 LEDs.

The LED arrangement in this configuration is 30 x 30 LEDs, and they come in two colours: one board has two types of high CRI LEDs (90 + 98 ) with added UVA 6500K white phosphor LEDs (CRI95) for a total 2950K, while the other features the two high CRI LEDs without UV for a nominal 2700K.

Both boards have a combined CRI95 - arguably one of the highest CRIs on the market - using the highest efficiency and top flux bins of their type. I'll provide more details on the actual LEDs in the next post.

One advantage of these boards is they can be run on most 48V-54V drivers commonly used in strip and HLG type builds, and do not require a second channel (driver) for the UV component. They also have a dedicated heatsink option, but can also be mounted on HLG type heatsinks.

Paired in a square configuration, they are designed as a true 600W HPS killer running at up to 400W (per pair) and covering a 3'x3' area.

They can also be paired to cover a 2'x2' grow area without heatsinks running at half that wattage. The boards in the first photo above are running at about 95W each for a combined 190W, or 210W at the wall.

Here is a photo of a two-colour board showing the layout. The black lines and circles denote "safe" areas that can be drilled out if need be. For example, if not using heatsinks, the middle 4mm mounting holes can be drilled out to provide a bit of air circulation, whilst the 4mm corner holes can be drilled out to 6mm to accommodate hangar hooks etc. Ostensibly the "safe" areas are designed as a guide for mounting screws, washers (optional, but not really required) and hanging hardware. The Molex wire connectors are rated up to 9A, which means you can safely wire two boards in parallel and run them at their maximum rating. More boards can be wired in parallel using Wagos or terminal blocks etc, and of course they can also be run in series.
IMG_0795.JPG

Here is a three-colour board. There are really only two "colours" - 2700K and 6500K - but there are three different types of LED (CRI 90@2700K, CRI98@2700K, CRI95@6500K), which is not at first obvious.
LED120W.jpg

Here is a two-colour board. The two-colour boards are design for those who are already running supplemental UV and require lots of red, or those who might want to use them to supplement their existing veg boards/strips (3500K-5000K etc) for flowering. They have an excellent red:far red ratio and plenty of 620-660nm spectra, as you will soon see.
IMG_0800.JPG

The spectra of each board is perhaps the most interesting thing about them, and I will go into further detail in the next post.

There were a number of reasons for designing and building these boards, so I will explain their philosophy.

Firstly, a lot of commercial growers have open rooms with stations that typically measure 3'x3' - which is the ideal footprint for a 600w HPS. Furthermore, many growers looking to switch from HPS to LED are already using 600W HPS lamps - the industry standard - but there are no real LED boards designed to truly replace a 600W HPS. Most LED boards are designed to fit in tents ranging in size from 2.5x2, to 4x2, to 4x4 etc.

We wanted to design a board that would truly be a 600W HPS killer, replacing it with better quality light, a much more even footprint, and enough power to produce the same yields with 2/3 the electricity.

There are plenty of strips and boards on the market that can be used for veg or dual-purpose veg and flowering, but there are not as many that serve as pure flowering boards, which again is the aim of replacing a 600 HPS.

So that was the first aim. And the first problem: a single board would be much too big to replace a 600W HPS, so we designed the ideal footprint and then cut it in half to produce a board that could be mounted in pairs (16"x16" square configuration), or in other configurations depending on grow area. The boards are 1.5 times the size of HLG's QB boards, so can also be used in tents (one board per 2'x2').

The next aim was to produce a board that could be run soft without a heatsink, or run hard with a heatsink - giving the grower the option to run them anywhere from 75W to 225W per board, depending on their budget and grow style.

To do this, dedicated heatsinks were required, and these were modelled on HLG Slate 2 clones. Mainly because they were a cheap existing design, but also because those with HLG type heatsinks could use them with these boards.

We trialled two heatsinks and went with the heatsink on the left, as it was a little bit lighter and testing showed it was more than adequate to cool the boards. Incidentally, the heatsinks measure 444mm x 206mm and so are equal to one half of a HLG Slate 2 triple heatsink (891mm x 205mm) - cut one in half, and you have two High Light heatsinks. The High Light heatsinks are, however, a slightly heavier design at 1.2kg each (compared with 1kg for a similar sized Slate 2 design). They are obviously pre-drilled for the boards and have 5mm (3/16") mounting holes on each corner.
IMG_0687.JPG

If you are going to spend money on LEDs, then you may want the option to drive them hard, right?

Well, our philosophy was there are always new LEDs around the corner, so if you're a commercial grower, you can afford to run these hard for a couple of seasons and then replace them if you so wish. Or just keep running them. They are a true 225W board, so you can get the most out of them if required. Just bear in mind they won't be as efficient at 225W as they will be at 95W . . .

Flexibility, as you can see, was the third requirement. The ability to use existing drivers was factored in. These boards will drop about 48V at half power, rising to about 50-50.5V at full power. Most 48V "A" type HLG drivers will supply the 50.5V required at full power, while 48V "B" type drivers can supply half current. If you want to drive the boards at full current with a "B" type driver, then 54V will fit the bill.

Finally, the boards needed to be a robust design with good light spread and heat dissipation: and that is why 2mm aluminium PCBs were chosen with an even spread of LEDs to produce the most even footprint and heat markers.

By now, you're probably wondering about the spectra and choice of LEDs . . .
 

Prawn Connery

Well-Known Member
Why high CRI?

Well, for several reasons. The first is that most high CRI 3030 LEDs have a better spectra, with more red and blue at either end of the scale compared to CRI70 or CRI80 LEDs. The better quality light also helps in the grow room, as having "true colour" allows growers to identify nutrient deficiencies and other issues, instead of having them masked by washed-out colours.

But perhaps the most interesting reason to use them is that high CRI 3030 LEDs emit a lot of red in the 620-660nm range, as well as far red in the 730+ range - negating the need for supplemental red and far red LEDs.

Supplemental red and far LEDs not only add to cost, but in many cases require a separate driver (more cost). The same goes for supplemental UV - the UV component often runs on a separate channel.

But what if we could combine all these things on one board powered by one common driver? Pardon the pun, but that was the "driving" force behind the High Light design.

Here is a 2950K High Light (blue line) compared to a typical 3000K Samsung LM561C strip build (orange line). The first thing you notice is the colour shift towards the red end of the spectrum, as well as the reduced green and blue, and added UVA (400-420nm). There is also slightly more cyan, for a more even curve.

This is the difference between a typical CRI80 and a CRI95 board.
original_spectrum_normalized-1.png

Here is the UV-enhanced High Light spectrum (three-LED board):
Production Board.png

And here is the non-UV board (two-LED):
Screen Shot 2019-04-03 at 13.57.00.png

These spectra look similar to typical CRI80 boards with supplemental 620, 660 and 730nm LEDs - except this is one board with one channel (driver).

The choice of LEDs came down to what was on offer and what could be combined easily on to a board without ballast resistors (less efficiency) and without multiple channels.

Nichia make arguably the best LEDs on the market. We could debate Nichia vs Samsung all day long - and indeed, Nichia lab tests have shown that Samsung are not averse to inflating their Flux figures (but Nichia would say that, right?).

But the bottom line is, Nichia's latest 757 series is at least the equal (or better) than Samsung's LM301B series.

Nichia also makes an "Optisolis" LED that has a very high CRI rating - amongst the highest in the industry - and it was this that drew us to include them in the design.

Here is a Nichia Optisolis 3000K CRI98 spectrum graph for reference (we use 2700K, which have less blue and more red). Nice curve, eh?
Screen Shot 2018-11-01 at 13.56.25.png

The choice of a UV-based white phosphor LED was an interesting one. UV is really the missing component in most LED builds compared to HIDs (MH and even HPS has a small amount of UV). We wanted to incorporate some UV and/or near UV into the design, and wanted to do this without using single-colour UV LEDs, which are expensive, don't last as long, and are more difficult to put on a single channel board.

UVA is the spectrum that helps bring out terpene and, to a lesser extent, THC production in plants.

Yuji LED pioneered the UV-based white phosphor SMD LED, and we looked here first, but couldn't find anything that suited (for the price). It was then we turned to Seoul Semiconductor (SSC).

Let me first explain what a "UV white phosphor" LED is. Most LEDs start with a "blue pump" - single blue spectrum in the 450-460nm range - which is then scattered using red and green phosphors to create all the colours of the rainbow.

A UV-based white phosphor LED uses a "purple pump" - 400-420nm single light source - which is scattered using red, green and blue phosphors.

Why UV? Well, it creates a more even spectrum. It also adds near UV and UVA to the spectral curve. And finally, very high CRI (95-98 ) LEDs can be built this way as you are introducing an extra "colour" (phosphor) to the mix.

Here is what a Seoul Semiconductor 6500K LED spectrum looks like. These are used on our boards.
Screen Shot 2019-04-30 at 16.16.35.png
Notice is has four spectral "peaks" compared to two for typical LEDs.

For the eagle-eyed among you, you may have noticed a slight discrepancy in some of the graphs - that's because SSC changed the design of their LED between our test board and production board, moving the "purple" pump from 400-405nm to 410-415nm. This meant there was slightly less UVA, but the new SSC design was a lot more efficient, as we have seen in our production boards (which use the latest SSC higher efficiency LED).

Suffice it to say, the production boards ended up with slightly less UVA than the test boards, but are about 1% more efficient. The SSC LEDs make up about 10% of total spectral output.

What we have ended up with is a high-powered, high-efficiency, high-CRI (hence "High Light) board that extends the spectral curve from one end (UVA) to the other (Infrared), whilst providing a very strong red output to increase flowering yields.

The trade-off is these boards will not be as efficient as typical CRI80 boards due to the types of phosphor used, however they are not far behind and actually have higher YPFD outputs. YPFD measures light from 380-780nm instead of the usual PPFD measurement of 400-700nm.
 
Last edited:

Prawn Connery

Well-Known Member
The proof is in the pudding, as they say, and so of course all the claims in the world are just that - claims - until put to good use. To that end, these boards are being scientifically tested as I write, and we will have accurate umol/j figures for them soon. No speculation, just real figures.

Some of these boards have already gone out to test growers - some of whom are using supplemental UV and some of whom aren't. Preliminary PAR mapping has been promising with 1000 PPFD at 24" using eight boards in a 4'x4' with even coverage throughout the canopy using a total of 670W.

You can see the results here: https://www.rollitup.org/t/the-monumentally-epic-knockdown-dragout-take-no-prisoners-slapdown-aussie-high-light-vs-hlg-288.988144/page-4

In a 3'x3' open configuration (below), two boards running at 345W total is giving a fairly even 950 PPFD at 18" without the aid of reflective walls (and tested with no overlapping light from the next station).

In addition to @Or_Gro's head-to-head against HLG's QB288 + red supplement, and QB96 boards, we're also running them against HLG's QB324 V1 (mixed CRI 80/90 3000K LEDs).

In this configuration, you can see two High Lights running at 345W (360W at the wall) vs two HLG QB324 V1s at 400W (420W) at the wall for the same PPFD figures at 18" - however, the HLGs obviously don't have the same even footprint, as the boards are smaller.
IMG_0827.JPG

High Light vs HLG QB324
IMG_0868.JPG

The boards and heatsinks are easily connected to each other using existing holes and T-bar or L-bar extruded aluminium with 3/16" screws. Push-in connectors are designed for use with solid wire 18AWG.
IMG_0888.JPG

The layout of the UV LEDs is restricted by the circuitry (the LEDs are not typically connected in series, but are grouped together in parallel/series strings), but testing has shown a very even spread with no spectral change from one end of the board to the other at 12".
IMG_0824.JPG

I guess one question that hasn't been answered yet is, are these available and how much?

There was an initial production run of 75x 3-colour and 21x 2-colour boards and about a third of these went to existing growers who were part of the research and development. Boards have gone out to the US, Germany and locally here in Australia.

Yes, there are boards and heatsinks available. But at the moment, this isn't a commercial venture - the boards were specifically designed and built for local growers who wanted them (myself included) with sales of any left-over boards going to offset the cost of development and production. If there's further interest, there may be other runs.

If you are in the US, then the bad news is shipping will be the killer. If you are in Oz, NZ or possibly Europe, then I'm happy to receive PM enquiries and I can get you a price on boards and shipping (they will be cheaper than other offerings from OS).

As I said, this isn't really a commercial venture - we just built a run of boards that we wanted for ourselves, and there are some left over for anyone else who is interested in trying them. I just thought it was time I explained what they were as other growers are starting to showcase them at RIU.

If anyone has any questions, I'm more than happy to answer. Just bear in mind I am not an electrical engineer - I am a grower - and I may need to refer in-depth technical questions to our engineer. Everything else I can pretty much answer.
 
Last edited:

Frank Cannon

Well-Known Member
Here's a wee teaser of his "HIGHRED" boards mounted to a custom designed heatsink to accommodate 2ft UV Arcadias in the recess to keep everything flush mounted. Heatsink not required in my application (100w per board) but it makes mounting everything easy....

20190430_155324.jpg

These will be replacing 6 x QB96s running at 1050w in my 4 1/2 x 4 1/2.
Sorry for butting in cuz, now carry on with the lesson pls:clap:

FC
 

Prawn Connery

Well-Known Member
Here's a wee teaser of his "HIGHRED" boards mounted to a custom designed heatsink to accommodate 2ft UV Arcadias in the recess to keep everything flush mounted. Heatsink not required in my application (100w per board) but it makes mounting everything easy....

View attachment 4325468

These will be replacing 6 x QB96s running at 1050w in my 4 1/2 x 4 1/2.
Sorry for butting in cuz, now carry on with the lesson pls:clap:

FC
Yes, that was one of the aims - producing a board that could run hard (up to 150W) without heatsinks. As FC can attest, the boards are very robust and will stand up to a lot of punishment.
 

Prawn Connery

Well-Known Member
Two reasons: the 5000K-6500K Optisolis have a borderline UV pump (420nm) vs purple pump (410nm) and are max rated at 100mA vs 150mA for the SSC LEDs. That's the main reason. In fact, the SSC LEDs when we first tested them had a 400nm purple pump, which is the main reason we chose them, as we really wanted to extend the spectrum into the UV range and the Optisolis were 20nm higher at the time. Currently, the Optisolis is still 10nm higher than the latest SSC revised LEDs (420nm vs 410nm).
 

Prawn Connery

Well-Known Member
Take your time mate. The main reason we haven't done extensive PAR mapping is because once we get the data back from photometric testing, we'll be able to use an IES file to map any layout we want on a computer - a lot faster and just as accurate (if not more) than physically mapping everything. The preliminary mapping we've done has mainly been to match drivers to certain areas and uses. As you've already found out, they don't need to be driven anywhere near their maximum rating to get good numbers.

Is it possible for you to post up spectral graphs of your QB288 + QB18 red/far red supplements. From what I remember, the combined spectrum was very similar to the High Lights.
 

Or_Gro

Well-Known Member
Take your time mate. The main reason we haven't done extensive PAR mapping is because once we get the data back from photometric testing, we'll be able to use an IES file to map any layout we want on a computer - a lot faster and just as accurate (if not more) than physically mapping everything. The preliminary mapping we've done has mainly been to match drivers to certain areas and uses. As you've already found out, they don't need to be driven anywhere near their maximum rating to get good numbers.

Is it possible for you to post up spectral graphs of your QB288 + QB18 red/far red supplements. From what I remember, the combined spectrum was very similar to the High Lights.
8x288 plus 4x35 (35 is same bd as 18 )
C376C150-6880-4026-ADDE-DBAC640107BA.jpeg
 

Prawn Connery

Well-Known Member
Production Board.png

Thanks mate. The reason I asked is because it's striking how similar they are considering one is a single board (High Light) and the other is a regular CRI80 board with supplemental red and far red strips.

It's a good example of the way the high CRI colour shift already incorporates most of the red and far red growers are adding to their CRI80 boards. This is exactly what we were trying to achieve with our boards using one board, one driver, and some added UVA.

Also note how the High Light stretches all the way into the infrared region. This is partly why 400-700nm PPFD figures don't tell the whole story with these boards. The spectral curve covers a lot more ground and should trigger more photoreceptors in the plants.
 
Last edited:
Top