The Good, Bad And Ugly Truth About Led Tech

  • Thread starter jumpincactus
  • Start date
  • Tagged users None


Premium Member
LED tech is predicted to hit 4.19 bil by 2019 with good reason as more and more mfgr's are marketing to both small and large scale growers especially in the burgeoning cannabis market.

I can remember as the LED craze hit cannabis growers with the marketing promises of less heat and better yields that legacy HID lighting is slowly becoming a thing of the past. However we/they still have a way to go yet to truly bring out the full potential of this exciting tech. One of the remaining challenges is thermal management. Heat and LED's do not play well together. Here is a great article on some of the facts concerning LED's and the heat they produce and what needs to be done to manage the thermal load so the alleged touted 40,000 hr life span of the bulbs can be realized.

With innovations in LED grow lighting, society is on the cusp of a revolution in horticultural production. But, says John Cafferkey, marketing manager of Cambridge Nanotherm, the success of this revolution may hinge on finding ways to improve the thermal efficiency of LED modules…

It’s an incredible time for those in the horticultural industry right now. Arguably, few people outside of the industry realise what a fundamental transformation is currently taking place in terms of how our produce is grown, and some of the far-reaching implications this could have.

For many years high-intensity discharge (HID) lighting has been used as a substitute to natural sunlight. Where it is economical to run them – a critical factor – HID light sources have allowed farmers to grow crops indoors in a controlled environment and to a more flexible schedule (out of season and often year-round).


While HIDs have been the mainstay of many horticulturalists, they are now rapidly being replaced by LED technology. A recent report from RnR Market Research suggests that the LED grow light industry will be worth a whopping $4.19bn by 2022, with a CAGR of 9.63 per cent between 2016 and 2022.

Compared to HIDs, LEDs turn on and off quickly, making them simpler to fit into automated grow cycles. They can also be developed to emit very specific wavelengths of light so they can be ‘tuned’ for different types of crop and for different stages in a plant’s growth cycle.

However, most significantly, LEDs are considerably more cost-effective and cooler running than HIDs. It is these last two properties that reveal the true potential of LED grow lighting.

First, the massively reduced running cost of LEDs opens grow lighting up to a much wider range of crops, including many of those that would have previously been regarded as uneconomical because of their lighting requirements or cheap sale price. In other words, LED grow lighting is making the transition from price-premium produce (for example cannabis) to everyday agriculture.

Secondly, cool-running devices mean that farmers can grow crops in much tighter formations without the danger of heat damage to the plants. This enables some exciting applications, for example in areas such as vertical farming, where produce is stacked tightly in vertical formation, and can be grown close to source, with all of the economies this entails.

Waste heat is still an issue for LED designs

However, while LEDs run much cooler than HIDs, around 60 per cent of the power that goes into them is still outputted as waste heat. This waste heat can pose a serious threat to the lifespan of LEDs.

LEDs are covered by an encapsulant lens to protect them from the atmosphere and diffuse the light they generate. Unfortunately, this lens is also thermally insulating. Thus, the heat generated by LEDs cannot escape via radiation, as it would with a HID lamp. Instead heat must be conducted through the bottom of the LED design.

Should heat fail to be conducted away from the LED effectively the lifetime of the LEDs will be reduced, often catastrophically. LEDs that fail don’t just incur a replacement cost, there’s also the additional maintenance costs involved in constantly replacing them.

What’s more, removing this heat from LEDs is becoming a more pressing problem. For reasons of economy designers are looking to cram more higher powered LEDs (1-5W) into smaller spaces as a means of reducing the materials-cost of their designs. Individual LEDs running at a higher power density in this way present a greater challenge for thermal management systems.

To understand how thermal management can continue to be improved, we need to look at how LED modules are constructed to be thermally efficient.

Fundamental elements of a thermally-efficient LED design

Let’s assume that our example design uses ‘packaged’ LEDs. These are individual LED die that are ‘packaged’ on a thermally-efficient submount with their own encapsulant (so that the whole ‘LED package’ can later be easily mounted onto a PCB). Generally, the submount used underneath the die will be a ceramic. While expensive, ceramics act as an excellent transporter of heat (and heat spreader). This is referred to as ‘Level 1’ (L1) packaging.

The next stage of heat transport must be handled at the board level, ‘Level 2’ (L2). Standard PCBs cannot cope with the thermal demands of horticultural lighting, as the fibreglass/epoxy composites of which they are made is a poor thermal conductor. For this reason, ‘metal clad’ PCBs (MCPCBs) are typically used for such High Power LED designs.

How is an MCPCB constructed? There are a few structural elements that remain broadly similar across all MCPCBs. First you have a copper circuit layer, then a dielectric layer which separates the circuit above from the metal beneath, preventing the two from shorting each other out. Lastly, you have the metal board itself, usually aluminium, but sometimes copper, which transports heat effectively to the system-level assembly (such as a heat sink).

Dielectrics pose a challenge for effective thermal management in MCPCBs

As you’d expect, the copper circuit and aluminium board layers in MCPCBs are both excellent thermal conductors. The same cannot be said for dielectric layers. As the industry has discovered, finding a material that is electrically insulating (up to a suitable breakdown voltage) yet has high thermal conductivity is a real challenge.


The most commonly used approach has been to create a dielectric with a layer of epoxy resin adulterated with thermally conductive ceramic grains to increase the overall thermal conductivity. However, only so much ceramic can be added before the epoxy becomes brittle and starts to crack. And even the best such material on the market can no longer match the thermal demands of horticultural lighting systems.

To enable the continued evolution of horticultural lighting applications many designers today require boards with a level of thermal conductivity higher than ceramic filled epoxy can provide. It’s clear that a new approach is needed.

Nanoceramics: A new way of enabling effective thermal management in MCPCBs

A Cambridge-based nanotech company has developed a new way of addressing this ‘dielectric problem’, with an entirely unique approach to constructing thermally efficient dielectrics.

The company uses a patented electro-chemical oxidation (ECO) process to transform the aluminium surface of a MCPCB into an extremely thin layer (between 10–30µm) of alumina (Al2O3).

Whilst alumina isn’t technically the best material in terms of thermal efficiency, the extraordinary thinness of this dielectric layer renders this fact moot. As heat only has to travel through a few microns of material the overall thermal performance of the stack (the circuit, nanoceramic dielectric and the aluminium board) comes in at 115W/mK – head and shoulders above what can be achieved using alternative approaches.

Ultimately, horticultural lighting designers want their products to be cooler, smaller, brighter and more cost effective. They also want to ensure that they last for their stated lifetime. All of these factors are critical to the continued development of the horticultural lighting sector, and the new wave of farming techniques they enable. And as this market continues to blossom you can expect to see nanoceramics at the core of horticultural LED designs.


Premium Member
Here is another paper on the same topic,

This is a bit more for the nerds among us that have a solid grasp on the science and tech surrounding this technology.


Premium Member
Overcoming the Thermal Challenges of High-Power LEDs
Sep 21, 2015 9:32 AM | by Editorial Staff | NO COMMENTS
Dr Giles Humpston, Field Applications Manager
Cambridge Nanotherm

The LED industry is heating up – both literally and metaphorically. LEDs are rapidly increasing in power whilst being crammed into ever smaller modules and arrays. Packaging approaches such as Chip-on-Board LED (COB LED) are packing ever increasing numbers of bare LED die directly onto a PCB and covering with a phosphor to create the ubiquitous ‘fried egg’ device. COB LED devices are predicated to make up around 25 percent of the packaged high-power LED market by 2019 so are very much on the ascendant. There is, however, a problem. The power density of COB LEDs is making thermal management a critical issue.

Failure to keep the junction temperature of LEDs within their safe operating temperature rapidly leads to dimming, deterioration of the quality of the light emitted and ultimately to catastrophic failure. Not ideal when the key selling point of LEDs is their longevity.

Existing thermal management solutions present different problems. Either they can’t offer the required thermal performance or their cost is prohibitive. The industry needs an alternative.

Conventional Approaches

The thermal power that needs dissipating for COB LEDs can be significant – tens of W/cm2. As such the key characteristics required for thermal substrates are excellent thermal conductivity and sufficient electrical isolation to prevent a short. This limits the possible choice of materials to aluminium nitride (AIN), alumina (Al2O3) or metal clad PCBs (MCPCB).

On paper AIN fits the bill perfectly – it has both high thermal performance and excellent dielectric strength. There are a number of factors that hold it back though. A complex manufacturing process requiring carbonthermal reduction of aluminium oxide or direct nitridation of aluminium, together with the extremely high temperatures involved, make it very expensive. The size of tile is limited to around 10 cm by 10 cm and it’s also brittle; limiting the yield and increasing production cost. Add to this the requirement for specialist processors and you end up with a very uneconomical product.

The alternative ceramic option, Al2O3 , is far cheaper than AIN but there’s a reason for this – it’s thermal performance is far lower. This discounts it from many high-power applications.

This leads us to the MCPCB option. Aluminium itself is a compelling option – it’s a great thermal conductor, cheap, readily available and extremely robust – but what distinguishes the extensive range of MCPCBs is the construction of the dielectric layer.

The standard approach of adding an epoxy to the surface to produce a dielectric layer has a huge impact on the thermal conductivity, resulting in a thermal performance below that of Al2O3. Attempts to anodise the surface of aluminium to combine the thermal conductivity of aluminium with the dielectric properties of a ceramic have been repeatedly unsuccessful over the years. The anodising process leaves gaps in the dielectric layer which can create electrical short circuits.

A New Approach
Cambridge Nanotherm has a unique approach using a patented electrochemical process for creating a dielectric on the surface of aluminium. This process produces a composite substrate with dielectric and thermal properties comparable to AlN but far more cost effective.

The Nanotherm process converts the surface of a sheet of aluminium into a dielectric nanoceramic layer, with the crystals formed as small as 30 nanometres. As a conversion process is used to form the nanoceramic from aluminium it creates a robust bond between the dielectric and the base plate, resulting in a dense, uniform layer of ceramic which acts as a perfect dielectric.

The nanoceramic dielectric layer is usually between 10 and 30 microns thick, depending on the breakdown voltage required. This makes the thermal path between the LED chip and the aluminium as small as is feasibly possible, resulting in extremely high overall thermal conductivity and low thermal resistance.

There are two approaches to attaching the copper wiring trace to the nanoceramic. The most cost effective solution is to add a laminated copper surface – Nanotherm LC. For the most thermally demanding applications the copper layer is sputtered or plated on – Nanotherm DM, a fully inorganic MCPCB.

The Result
Nanotherm LC has a composite thermal conductivity of 115 W/mK (measured by the laser flash technique) and Nanotherm DM 152 W/mK. This outperforms Al2O3 and other MCPCB materials and rivals AIN, making it ideal for the most aggressive thermal challenges.

In the LED industry, cost is always a factor. With an industry desperately trying to find ways to bring the cost of LED products down to a point that the mass market are comfortable with any saving is good. Not only do nanoceramic substrates offer a significant cost advantage over ceramics, they can also be processed through standard PCB manufacturing facilities avoiding the costs associated with specialist manufacturing processes. Nanoceramics open up an entirely new class of thermal management aimed squarely at the high power LED market.

Dr. Giles Humpston is a metallurgist by profession and has a doctorate in alloy phase equilibria. He is a cited inventor on more than 250 patents and has co-authored over 150 papers as well as several text books. Dr Humpston currently works as the Field Applications Manager for Cambridge Nanotherm on thermal substrate technologies.



Good Post +1, A competitive Lighting Market is always good news for Us - And it is Us in Real Time who are the Sponsor's of this Tech, I have used HID lighting 20+ years, and when it came to making a 1st Purchase, the question on my mind was "How much do I want to spend on a Tech in its infancy" money was no object, I have the coin but consider myself a smart buyer when it comes to the new shiny objects. I purchased 2x MH TSW 2000 and 2 more MH TSW 3000 a year or so later, I almost went for HLG's, but after seeing how well the tsw 2000's performed for my scenario I didnt "buy-in" to the craze, knowing LED's will get better and better as the research gains speed in this Competitve Market. I am comforted in the many articles now being released that I made a good decision, when the peer reviews come out and the Tech catches up(and my lights wear down) I'll jump in with a bigger investment. thanks M8

(just to note, the less expensive MH's compared to my HID's are not quite there yet but very usable with good results, without all the heat)

Latest posts

Top Bottom