LEDs in a Pincer Movement

There was quite a lot of interesting news this week, even if there was nothing ‘earth shattering’. We have Ken’s report on the SID Automotive show in MDM and at the end of the week, there was talk of a ‘crisis’ at Samsung Electronics. At the same time, the company forecast record profits for Q3. That’s some crisis!

At the end of August, I wrote an editorial headed ‘Samsung Seems Non-stick‘ that highlighted that the company seemed to be able to ‘glide past’ problems that could have mortally damaged other companies. However, it bet hard on OLED, which has worked incredibly well for smartphones, but was a failure in TV.

This week we have an interesting article on the roadmap proposed by the US Department of Energy for lighting. We don’t normally cover much in lighting, but this seemed interesting to me. The roadmaps suggest that LED lights will eventually get to around 330 lumens/W (although that might take 30 years), while OLED lighting starts to flatten off at somewhere around 180 lumens/W. Lighting designs, of course, do not have the colour filter that is an essential part of the TV OLEDs that LG is making and which reduce the efficacy of current OLEDs by 75% or so, so at the moment, TV panels are much worse.

These fundamentals often turn out to be important. For years, the OLED supporters based a lot of their case on the claimed simplicity of the structure of OLED compared to LCD (although OLEDs have turned out to be quite complicated, in practice).

In the same way, the core efficacy and efficient conversion of electrons to photons is a big attraction of LED. Like the ‘simplicity’ of OLED, the efficacy of LED is a key factor as it allows lower power consumption (saving cost) and higher brightness, which will be essential for HDR, eventually. Of course, LED also has all the traditional advantages of emissive displays in black level, viewing angles and response times.

I wouldn’t describe LED making as simple, but in electronics, scale sometimes beats simplicity. Chips and chip making are incredibly complex, but the sheer scale means low cost. As I have written before, LED is already a large scale business with around a billion per day being made. If TV displays can go LED, this will quickly go to trillions – there are 25 million LEDs in an ultraHD TV so for every 40 made, a billion LEDs will be used.

Like chips (I have seen estimates that there are around 400 million transistors made each year for every person on earth!), sheer scale might drive costs down to amazingly low levels. Of course, the internet of things will drive even the 400 million to even higher levels as everything gets intelligent.

It has long been a truism of the display industry that every technology and concept can find a home, somewhere. LED displays have a home at the moment in very large and outdoor displays and, Samsung hopes and plans, in premium cinema. They are also likely to appear in little displays, as well, given the work that is going on below the surface in microLEDs for wearables.

In that respect, LED might be a bit like DLP. DLP, initially, could not compete very well with the mainstream, so TI used a pincer movement, with projectors in large projectors and very small projectors, where the mainstream was not so competitive. Gradually, that business at the edges of the market allowed the volume and technology to develop to allow DLP to go mainstream effectively. LED could do the same.