Not All LED’s are the Same (Addendum)

by / Friday, 08 November 2013 / Published in Environmental Testing, Green Building Consulting
LED Dirty Electricity

Dirty Electricity Concerns?

LED Dirty Electricity

Dirty Electrcity Demo. Board

In the recent blog of September 10th, 2013, EMI and LEDs – Not all LEDs are the same! we described dirty electricity and examined several brands of CFLs and LEDs.  This blog tests several more brands of LEDs to determine which produces the least amount of LED dirty electricity .

It has often been said in the dirty electricity circles that LED lighting produces less dirty electricity than CFL lighting.  CFL lighting produces dirty electricity due to the fact that CFLs operate at high frequencies (typically between 10kHz to 40kHz) and need electronics to control the voltages and frequencies to make it operate.  LEDs, it is commonly thought, produce less dirty electricity since LEDs use DC current and the electronics needed to convert the 60Hz down to DC current is less complicated.

Our previous blog showed that this is not always the case.  Depending on the circuitry that a particular brand uses to convert AC to DC for the LED to operate, the amount of dirty electricity produced by an LED can be many times that of a typical CFL.

So which LEDs produce the least amount of dirty electricity?


In addition to the Ecosmart, Philips and CREE brands that were tested before, a couple more LED brands were put to the test.  Using a Stetzer meter and a Greenwave EMI meter, the dirty electric values of several LED manufacturers were measured.  Both meters were used since the methods of their quantification of dirty electricity is different.  These CFLs and LEDs were all chosen with the luminosity of near 800 lumens (the equivalent of a 60W incandescent bulb).

LED Dirty Electricity

Dirty Electricity-LEDs measured in mVolts

Using the Greenwave EMI meter on the LEDs and correcting for background dirty electricity levels: Ecosmart and CREE measured much better than Philips, though Ecosmart still produced more than 100mV of dirty electricity.  Sylvania and FEIT Electric were two more pricey LEDs that were added to the test.  While the FEIT Electric did perform better than the Ecosmart, the Sylvania measured a bit worse.  Overall the least amount of LED dirty electricity was still produced by the CREE brand LED.  The CREE brand added only 17mV of dirty electricity to the electrical circuit.


LED Dirty Electricity

Dirty Electricity-LEDs measured in Stetzers


LED Dirty Electricity


Using the Stetzer meter for the same test and correcting to a background level of LED dirty electricity: Ecosmart, CREE, FEIT Electric and Sylvania all again measured much better than Philips. Using the Stetzer meter, both the FEIT Electric and the Sylvania produced less than the Ecosmart.  The CREE brand LED still produced the least amount of dirty electricity of the brands tested.  Using the Stetzer meter, the CREE brand LED produced 39 ‘Stetzers’ of dirty electricity.  [Please note: we do not know the precise definition as to what a ‘Stetzer’ equates to.  The Stetzer Electric company has not openly defined it for others in the industry to understand and, therefore, is a questionable ill-defined unit.  However, it does allow for comparisons to be made between brands for this limited study.]


What do these measurements say to us?  On average, LEDs do produce less dirty electricity than CFLs, but be wary of the brand you are using!  If you are sensitive to dirty electricity or are concerned about dirty electricity, Philips brand LEDs should be avoided.

In this addendum research we explored LEDs from Ecosmart, Philips, Sylvania, FEIT Electric and CREE which were all available from Home Depot. If dirty electricity is a health concern in your home and you are interested in saving electricity and money in the long run, it is most recommended that you consider using CREE LEDs based on this limited research.  If CREE LEDs are not available, you might consider using FEIT Electric or Ecosmart.  Sylvania LEDs were are par with Ecosmart, however the Sylvania LED was the most expensive of the LEDs tested ($37 per bulb!).  Further research with different manufactures will be performed in the future to continually expand the selection of recommended brands.  A direct link to Home Depot’s CREE LED webpage can be found here.

12 Responses to “Not All LED’s are the Same (Addendum)”

  1. […] is our third blog about LEDs and dirty electricity. See EMF LED Blog 1 and Not All LEDs are The Same for Dirty Electricity Blog 2 for additional […]

  2. […] is our third blog about LEDs and dirty electricity. See EMF LED Blog 1 and Not All LEDs are The Same for Dirty Electricity Blog 2 for additional […]

  3. Mitigating and Eliminating Dirty Electricity – A Mom’s Guide (part 5) - Organic Housewife says :

    […] several compact fluorescent light bulbs (CFLs) around the house to LED.  However, according to this experiment, not all LEDs are created equal.  Cree brand LEDs added the least dirty electricity to […]

  4. Gerry Wootton says :

    The hazmat issues with CFLs have been somewhat overblown: in well regulated jurisdictions there are upper limits on the amount of mercury but in less regulated jurisdictions much higher content can be found although this is still a small quantity. Arguments about this generally ignore the number of toxic materials in incandecent lamps including lead and chromium compounds and mercury residuals which are toxic (the complete list runs to some 9-12 substances); this is where lifetime is a factor since the total amount of material placed in the waste stream is inversely proportional to lifetime. In the US (if you study the regulations) one must treat incandescent lamps as hazardous waste in 100+ quantity; however, as with CFLs, it is good practice not to dispose of them in general waste. In Europe where hazmat regulation is more pragmatic, the whitening coating on the inside of the glass is not allowed as both the material and it’s binder are hazardous materials considered to be non-essential.
    Yes virginia, in the US the 2.5 micrograms of mercury in a typical CFL pass the toxicity test because the test is based on percentage of total weight of the product rather than the amount or concentration of the material itself – this is in line with pollution standards that deal with concentration rather than absolute and/or cumulative amount (dilute don’t pollute?!). The same is true for incandescent bulbs except for the lamp base materials which can in some cases (depending on construction) fail the test.

  5. Gerry Wootton says :

    There’s several things acting here which can only be sorted out by examining radiated EMI and conducted harmonics. This is mainly a function of ballast design but the plasma discharge in fluorescent lamps can also produce EMI – my basement got a lot quieter when I got rid of all the 4′ linear fluorescent lamps. The front end of the ballast can be a simple rectifier bridge which has a poor power factor and produces a lot of conducted harmonics but it can be a more sophisticated e.g. polyphase rectifier stack which has a better PF or a switch mode inverter which can provide a very good PF; beyond that the ballast can incorporate PF correction. The current regulator portion of the ballast generally runs a high-frequency PWM circuit so the combination of circuit layout and input filtering will determine how much or how little high frequency signal couples to the line through the base of the lamp. The same problem exists with simple so-called wall-wart power supplies, which are already being regulated out in some jurisdictions; I eliminated a lot of noise in my keyboard amps by replacing the original keyboard wall-wart, a simple rectifier type, with a PF correcting switch mode version. The configuration of the emitters also has some effect on effectiveness as an antenna radiating RFI – lamps that use a light guide to distribute light rather than simple placement of the emitters (e.g. so-called pineapple arrangement) are advantageous. For both CFLs and LEDs there is a lot of variation possible in the ballast design and circuit layout which will affect the amount of electrical noise produced. Another issue is the lamp socket: the traditional A base fully encloses the base with metal and should, if properly wired, be connected to the designated (neutral) conductor which should provide a moderately good ground path however in more modern sockets this may be aluminum which is a poorer RF shield and in some more recently produced bases, the full socket is replaced with a plastic thread and a contact which could be even worse or at least relies entirely on the metal in the lamp base.

  6. Michael Holm says :

    I am very surprised to see that the Philips did measure so badly here.. I tested my Philips LED bulb against others mostly no name bulbs and it is MUCH lower on all radiation or poisoning on the household electrical wiring and in the air than the other brands. (sorry, this is not my primary language)

  7. David says :

    Another data point: I own several Sylvania 75W replacement, 14W bulbs labeled “14A19/DIM/0/827” and “Intertek: 4004755”, bought in 2013.

    Unlike the Sylvania I referred to in my previous post, that one added only about 10-15 GS when turned on.

    One thing is clear: you cannot generalize by brand.

    I’d recommend to everyone who is concerned about EMFs to invest into a GS or Greenwave meter. Buy LED bulbs, test, return if necessary, repeat.

  8. David says :

    I found another decent LED bulb: the UtiliTech Pro 60W replacement (8W) Warm White, bought from Lowe’s. The model # is LA15DM/CL/500/LED.

    It added about 10GS units when I plugged it in – it’s hard to say precisely because of constant fluctuations.

    Also tested – and then promptly returned for a refund: the Sylvania 60W replacement (8.5W) A19 bulb, also from Lowe’s. It added about 600 GS units.

  9. Alex Stadtner says :

    Agreed. Lots more testing is required. And using a standardized testing method and unit of measurement is key. Dirty electricity is a big mystery for most electronics manufacturers, and we’re a long way from having devices with mandatory and uniform reporting on their labels. Glad you’re interested and doing your own testing. Let us know what else you learn.

  10. David says :

    Based on your findings, I bought a CREE soft white LED 100W replacement bulb (BA21-16027OMF-12DE26-1U100) at Home Depot. I was hoping that it would very favorably compare to my existing (2013 bought) Philips #9290002097 LED bulbs (the yellow L-prize bulb). I found the opposite.

    On two outlets that had baseline reading of 200 GS units, switching on the Philips made no discernible difference, whereas switching on the CREE increasing the reading on my GS meter to about 800.

    Perhaps CREE changed the transformer inside their bulbs, or your readings were never representative of the entire company to begin with. More frequent testing is definitely needed.

  11. Sten says :

    Thanks for the test – could you perhaps include IKEA LED bulbs in your next test?

    BR /Sten

  12. Hey. Thanks for this! We are slowly changing some of the lights to our house to LED for energy and heat savings (from incandescents). We had actually formerly switched to CFLs, but then switched back to incandescent lighting after learning about health hazards from CFLs.

Leave a Reply