"Full spectrum LEDs" on eBay, Amazon etc.

[enricosavazzi]

These days one can find plenty of "full spectrum LEDs", also described as "380-840 nm LEDs", on multiple online auction sites including Amazon and eBay. See for example a "full spectrum 380-840 nm" 10 W cob on Amazon (https://www.amazon.com/Chanzon-Spectrum-380nm-840nm-Components-Hydroponic/dp/B01DBZK3V6 ) and plenty of similarly described single LEDs from 3W and up.

 

The purpose of this heads-up is to warn that these devices are not the full-spectrum LEDs that several of us have been waiting for, to use as LED illumination sources emitting UV and VIS, and a little NIR as a bonus. I have not tested any of them, but from the information one can glean from the ads they are not what the name suggests, but 380 nm UV LEDs embedded in a resin mixed with a phosphor that emits in the red, deep-red and NIR range. Possibly there is also another phosphor in the mix, emitting in the range across the UV to VIS borderline (mostly purple and blue), but I cannot know for sure without testing these devices. In any case, their emission is not even remotely "full spectrum", but has a primary peak in the UV and possibly blue, and a broader and lower secondary peak in the red and deep red, possibly with a tail in the nearby part of NIR. There is most likely little or nothing in the rest of the VIS, as well as below 380 nm.

 

These devices are an extension of the well-known "LED grow lights" sold for over two decades, originally equipped with plenty of red power LEDs mixed with a few blue LEDs. The red light is well known to stimulate photosynthesis in virtually all plants, while the blue light and/or NUV supposedly stimulates flowering of some plants (exactly which plants I will not discuss here).

[Andrea B.]

Enrico, thank you for this warning. 

 

As Editor, I emboldened two phrases in your topic to help make the point "at a glance".

[colinbm]

I have seen on YouTube, probably Clive, that the white light cobs are a LW UV LED with a tri-phosphor coating.

 

[aphalo]

COB is an abreviattion for "chip on board", and refers to LEDs consisting of multiple LED chips mounted on a subtrate, either ceramic or metal based, that can efficiently transfer the heat away from the chips, allowed densily packing of the chips. SMD or "surface mounted device" are usually, but not always, based on a single LED chip.

 

White LEDs, COB or not, are most commonly blue LED chips emitting at wavelengths between 450 nm and 460 nm plus "phosphors" in a coating. Some of those with best colour rendition index (CRI) ratings are based on violet LED chips, emiting at 405 to 410 nm, instead of blue LEDs chips. I haven't seen yet any based on LED chips with peak emission at wavelengths < 400 nm, even for horticulture. As far as I know this is in part because some safety regulations make it difficult to sell such LEDs as white light sources, at least in EU.

 

Enrico's point is very true. In horticulture, the terms "full spectrum" LEDs and "broad spectrum" LEDs are used to describe LEDs that emit at least some light outside the red and blue peaks so common in the earliest LEDs or LED luminaires sold for plant cultivation. On-line sellers nowadays very easily describe almost any LED for use in horticulture as "full spectrum", and when described as emitting from 380 nm to 840 nm refers to the whole range considering what is just narely measurable.

 

I have bought and measured some of these LEDs as my field of research in is plant photobiology. There is, however a huge variation in the spectrum of LEDs sold for horticulture. The simplest are just arrays of blue (430-440 nm) and red (650-660 nm) LED chips with no phosphor coating. I include below some spectra of horticulture LEDs that do have phosphors including some high power COB LEDs and some lower power LEDs used to assemble LED modules. Most of them sourced from electronic component distributors like Farnell, DigiKey and Mouser. I have also bought some from AliExpress.

 

In the plots below, red labels are wavelengths in nanometres at the peaks and black ones wavelengths at 5% of the tallest peak. Spectra are normalised to the maximum emission and expressed as flux of photons rather than energy flux. None of these LEDs are COB devices. They have been measured with an Ocean Insight Maya 2000 Pro array spectrometer using the R packages in the R for photobiology suite.

 

First spectrum is an example of the LEDs Enrico described at the start of this topic, in this case just having a broader red peak than earlier blue + red LEDs. Clearly not what we would call "full spectrum" even if described as such by sellers. This LED is based on a LED chip with peak emission at 440-450 nm. (The peak with shortest wavelength, and usually also narrowest.)

 

 

This "Horti white" LED does not emit white light. It has an unusual spectrum as it is designed to be used in luminaires in combination with red LEDs emiting at 660 nm. This LED is based on a LED chip with peak emission at 430-440 nm.

 

 

This is a modern high efficiency LED type for horticulture from Nichia. It deviates compared to other horticulture LEDs in having the red emission peak at a wavelength shorter than 650-660 nm. This LED is based on a LED chip with peak emission at 440-450 nm.

 

 

A high CRI (> 95) white LED with colour temperature near 5000 K. This LED is based on a LED chip with peak emission at 405-410 nm. The technology involved is described by the manufacturer.

 

 

A high CRI (> 97) white LED with colour temperature near 5000 K. This LED is based on a LED chip with peak emission at about 420 nm. This works very well for visible light photography. A description is available.

 

 

 

 

 

[colinbm]

These could be interesting without the colour peaks ?
https://www.ledsmagazine.com/leds-ssl-design/packaged-leds/article/16700628/seoul-semiconductor-and-toshiba-materials-introduce-broadspectrum-led-technology-updated

[aphalo]

Yes, the spectrum in my reply above, in the plot labelled Seoul/Toshiba Sunlike, is for one of the LEDs described in the article linked yo by Colin, but this one I have is not as smooth as in the plots in the article. From the specifications, the warmer colours tend to have smoother spectra.  There is quite a bit of variation among different "bins" and individual types within a series. Here is the spectrum for a different LED from the Optisolis "series", in this case a 10W COB. Type Nichia NFCWL036B-V3, colour bin sm503, centred at 5000 K. This spectrum is smoother than for the Optisolis SMD above. While Nichia has recently released these COBs, I have not yet seen any COBs in the Sunlike series. There are high CRI COBs also from other brands, and even no-brand ones available from AliExpress sellers. A couple of these are on the way from China and I will measure them once they arrive.

 

.