LED Radio Frequency Interference

Status
Not open for further replies.
G

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
160804-0918 EDT

Two nights ago I went out after dark in my Escape. Several days earlier I had put an Feit 10 W (so called 60 W equivalent) LED near my main panel. This was on at the time I left. I am usually tuned to 760 AM, WJR Detroit. I park my car in the breezeway all the time. This has a plaster ceiling with metal lath. Thus, somewhat of an RF shield.

Signal strength is less in the breezeway than in the driveway resulting from the metal lath shielding. WJR in years past was a clear channel station with a 50,000 W transmitter located possibly 40 miles away. I believe WJR is no longer a clear channel station and has to reduce power somewhat at night.

This discussion https://en.wikipedia.org/wiki/WJR seems to imply that WJR is still a clear channel station. But I still believe that signal strength is lower at night. I am at ground path range.

Some of the history mentioned is interesting. In 1936 my dad took a night class in radio broadcasting that was conducted at WJR. Close in time to when WJR became 50,000 W. My dad took me to a couple of the classes.

One was at the WJR studio and I remember the microphones from there and the room was fairly large. The other time was a trip to the WJR transmitter site. There I saw the very large vacuum tubes for the transmitter. These were long, had a copper (probably alloy) shell that would be the plate electrode (greatest heat generated at the plate), were water cooled, and I did not understand how they could use water cooling because most water is conductive, and as a small kid I thought all water was conductive. I believe they used distilled water.

The circuit could have been designed to have the plates grounded. There were two idential transmitters so that when one failed (tube burn out) they could immediately switch to the other. There was a speaker there and the audio quality was very good.

The antenna was about 7 or 8 hundred feet tall. One wavelength at 750 kHz is 1311 feet so some antenna tuning was done.

Also visited the WWJ transmitter and it was a lower power.

Back to noise. In the breezway I noticed a lot of static noise on WJR. Pulling out into the driveeway the noise reduced a little, and pulling under the primary line that runs along my street the noise increased a little, but not to the magnitude that existed in the breezeway. In retrospect the wires feeding the street light run below the primary lines and connect to the secondary of my pole transformer.

Today with WJR at daytime strength I ran a test with the LED and any other known noise makers in the house off. Other noise makers are phase shift dimmers and other Feit LEDs. There was no background static, in the breezeway or under the power lines. Then turned on the one single LED near the main panel and some static noise was present in the breezeway and under the power lines. Thus, this bulb is the source. Noise level is not as high as at night. Thus, I still believe WJR is lower power at night.

I had previously discussed LED RFI when I ran some bench tested. Not all LEDs are this bad.

.
 
160804-1141 EDT

A little more about the early days of radio.

See Chapter 9 of https://books.google.com/books?id=_...AC#v=onepage&q=wwi radio station ford&f=false
Scroll up a page or so to the start of Chapter 9.

Fred Black was the father of one of my schoolmates, and Bill Gassett the father of another. Charlie Thomas was head of the radio group that I worked in. Two men I worked with had been operators of WWI in earlier days. The area where I worked at this time, possibly 43 or 44, was the studio that was built in the Engineering Lab for WWI. In the 30s and 40s it was used as a recording studio and had two Fairchild 16" 33 RPM disk recorders. There was a small bench in one corner of the studio for radio repair, and this included a 3" scope and an audio oscillator. At that time I could hear as high as 21,000 Hz. In 1945 I was here and I remember listening to the reports of the atomic bombing of Japan. After the second bomb one just got the feeling that the war would end soon.

Charlie Thomas of Ford previously worked for Lee de Forest.

The physical station for WWI was converted to an amateur radio station in thre 1930s, and this had to be shut down at the start of World War II. It was never restarted at the end of the war.

See https://en.wikipedia.org/wiki/WWJ_(AM) for more on early radio.

.
 
is this just educational info?

perhaps a coupling that is creating Rf noise, or, perhaps the LED unit is amplifying noise that is creating interference waves on the freq you are tuned too? try with another radio, do you get same results? perhaps your radio tuner is sitting right on the edge of its SN ratio?

not sure how high the led lamp is, if you place some aluminum foil draped over the lamp does the noise go away?
 
160804-1355 EDT

FionaZuppa:

RFI noise comments are for education. You or someone else may run into noise problems and not realize that the noise could originate from as little as a single LED bulb at full power with no dimmer as part of the circuit.

See my posts #41 and beyond based on bench and scope tests at http://forums.mikeholt.com/showthrea...=174603&page=2

http://forums.mikeholt.com/showthread.php?t=178094&highlight=led+rfi+noise

I took the Feit bulbs that previously showed the high frequency oscillation and put them in the breezeway to find a place for them. This was weeks ago. They work fine from a perspective of providing light. These breezway lights are not normally on and therefore not a significant problem. I just happened need a bulb at the main panel and I put in a remaining Feit there. What really attracted my attentionn was the RFI below the primary lines from this single Feit near the main panel.

I don't need to solve a problem. I solved the problem by using the bulbs where they don't present much trouble, and for experimental purposes.

.
 
gar said:
160804-0918 EDT

Two nights ago I went out after dark in my Escape. Several days earlier I had put an Feit 10 W (so called 60 W equivalent) LED near my main panel. This was on at the time I left. I am usually tuned to 760 AM, WJR Detroit. I park my car in the breezeway all the time. This has a plaster ceiling with metal lath. Thus, somewhat of an RF shield.

Signal strength is less in the breezeway than in the driveway resulting from the metal lath shielding. WJR in years past was a clear channel station with a 50,000 W transmitter located possibly 40 miles away. I believe WJR is no longer a clear channel station and has to reduce power somewhat at night.

This discussion https://en.wikipedia.org/wiki/WJR seems to imply that WJR is still a clear channel station. But I still believe that signal strength is lower at night. I am at ground path range.

Some of the history mentioned is interesting. In 1936 my dad took a night class in radio broadcasting that was conducted at WJR. Close in time to when WJR became 50,000 W. My dad took me to a couple of the classes.

One was at the WJR studio and I remember the microphones from there and the room was fairly large. The other time was a trip to the WJR transmitter site. There I saw the very large vacuum tubes for the transmitter. These were long, had a copper (probably alloy) shell that would be the plate electrode (greatest heat generated at the plate), were water cooled, and I did not understand how they could use water cooling because most water is conductive, and as a small kid I thought all water was conductive. I believe they used distilled water.

The circuit could have been designed to have the plates grounded. There were two idential transmitters so that when one failed (tube burn out) they could immediately switch to the other. There was a speaker there and the audio quality was very good.

The antenna was about 7 or 8 hundred feet tall. One wavelength at 750 kHz is 1311 feet so some antenna tuning was done.

Also visited the WWJ transmitter and it was a lower power.

Back to noise. In the breezway I noticed a lot of static noise on WJR. Pulling out into the driveeway the noise reduced a little, and pulling under the primary line that runs along my street the noise increased a little, but not to the magnitude that existed in the breezeway. In retrospect the wires feeding the street light run below the primary lines and connect to the secondary of my pole transformer.

Today with WJR at daytime strength I ran a test with the LED and any other known noise makers in the house off. Other noise makers are phase shift dimmers and other Feit LEDs. There was no background static, in the breezeway or under the power lines. Then turned on the one single LED near the main panel and some static noise was present in the breezeway and under the power lines. Thus, this bulb is the source. Noise level is not as high as at night. Thus, I still believe WJR is lower power at night.

I had previously discussed LED RFI when I ran some bench tested. Not all LEDs are this bad.

.
Click to expand...

According to a quick on-line FCC check, WJR is still 50kw day and night, non-directional. The classic AM blowtorch. What has changed is the noise level across the radio bands thanks to all sorts of stuff like noisy LED's and CFL's. :)

The huge water-cooled transmitter that you saw has likely been replaced with a box the size of a large refrigerator full of MOSFET's.

It appears to be one really cool transmitter site. http://www.oldradio.com/archives/stations/detroit/wjrpix.htm

I have steered clear of Feit lights after some premature failures of their CFL's years ago. Since then, I've stuck with big-brand lamps with good success (GE, Philips, etc.)

I'm fighting RFI right now from ballasts at one of my client radio stations interfering with their off-air monitors. They seem to have at least one ballast go noisy every month...CFL's and T-8's.
 
grich said:
According to a quick on-line FCC check, WJR is still 50kw day and night, non-directional. The classic AM blowtorch. What has changed is the noise level across the radio bands thanks to all sorts of stuff like noisy LED's and CFL's. :)
Click to expand...

When I was a kid back in the 50's and 60's I used to lay awake at night prowling the AM band listening to far away stations and logging their locations and call letters in a notebook. At night in the late 60's we used to listen to KAAY in Little Rock clear as a bell in south Louisiana on our car radios. You can't do that sort of thing anymore, at least not with the kind of equipment I had access to back then. I'm listening to a baseball game right now being broadcast by an AM station just across town from me and it's pretty noisy.
 
160804-2128 EDT

FionaZuppa:

For the particular Feit bulb I am presently discussing no need for an FFT. On the bench previously there was a moderately good sine wave in the middle of the half cycle with a frequency in the general range of 800 to 1100 kHz. A broadcast receiver was a crude spectrum analyzer. More than adequate for the purpose.


grivch:

I have setup my Panasonic all band receiver. Not quite battery powered at the moment. Have a DC power supply with the output somewhat isolated from the input AC line. With all my noise makers off the S meter reads 9 on WJR 760. Just off channel S meter reads 1. Tomorrow I will check during the day. I need to get a new battery. Not using the AC power cord greatly reduced the residual noise level. I am not sure that battery operation will reduce the residual noise much.


ggunn:

I put a Corcom filter on the AC input to the Panasonic and this had little affect on the noise. Back in the 40s thru 50s car radios used a tuned radio frequency amplifier in front of the mixer. This reduced image interference and some noise. Tuning was by inductance rather than capacitance. This made pushbutton radios feasible. In the mid 30s tuning was with a variable capacitor. Testing on the pushbutton mechanism was for 50,000 cycles and this could be done over one weekend.

Testing of automotive voltage regulators was for about 1000 hours. The generator was run in the lab with an AC motor to drive the generator. I don'y remember what kind of a load we had, probably resistive. The voltage regulator contacts were the problem because these oscillated many times per second. All sorts of different contact materials and combinations were evaluated.

We had a screen room where the interior was free of external RFI. So the only receiver noise was the thermal noise from the input RF stage.

.
 
160805-0820 EDT

Experiments last night and this morning.

The receiver power source is the same regulated power supply with 12 V DC output that provides some isolation from the AC line.

With none of my noise makers on (dimmers and Feit bulbs) WJR was an S = 9, and just off channel S = 1. Somewhat later when I thought all my noise makers were off the off channel noise rose to somewhat above S = 9, and on channel, WJR, S dropped to 9 when the WJR had a sort of capture effect, but I could not understand the audio.

About 4 in the morning I tried again. Same result.

After the sun came up and street lights were off the same result. All my noise makers are off. I went to the basement and found my Feit at the main panel on. So it turned itself on because I had not adequately unscrewed it. Turned the Feit off and I was back to the good measurements of last evening of S 9 and 1.

New experiment. Both my test receiver and a dimmer are on the second floor sub-panel, but opposite phases of the power company center tapped transformer. With the dimmer on at my normal minimum light position the off channel S reading went to about 2 from 1.

.
 
160805-0924 EDT

From the signal strength measurements it is clear that WJR is still a 50,000 W clear channel station.

Did any of you live near WLW when it was a 500,000 W station and bed springs or other objects would talk to you?
Also there was a story of one amateur with a 1 kW station that would light his neighbor's back porch light when his beam antenna was pointed in that direction.

Some reading:
https://en.wikipedia.org/wiki/Skywave
http://www.todayifoundout.com/index.php/2015/01/radio-signals-travel-night-day/
 
Last edited:
160805-0952 EDT

Another reference:
http://www.globalsecurity.org/military/library/policy/army/accp/ss0002/le1.htm
See section 3.

Another reflection mode occurs under temperature inversion conditions. One very foggy night, possibly 1947, I was operating on 2 meters with a 16 element beam and 25 W power input. Heard a signal from Erie PA, and was able to establish contact between Dearborn and Erie.

In the early 1950s one morning I turned on the TV and saw a TV broadcast from Texas. This would have been channel 2 I believe. Our channel 2 was not broadcasting at that time of day. In the early days of TV broadcasts 24 hours per day was non-existent. Same for the early days of radio. Also in the early radio days recorded music was played on an orthophonic player into a carbon button microphone.

.
 
gar said:
I put a Corcom filter on the AC input to the Panasonic and this had little affect on the noise. Back in the 40s thru 50s car radios used a tuned radio frequency amplifier in front of the mixer. This reduced image interference and some noise. Tuning was by inductance rather than capacitance. This made pushbutton radios feasible. In the mid 30s tuning was with a variable capacitor. Testing on the pushbutton mechanism was for 50,000 cycles and this could be done over one weekend..
Click to expand...

Why would tuning have to be by variable induction to make the pushbuttons work? The tube phonograh/radio combo I had as a kid that I took apart and invented the electric guitar (that's another story) definitely was tuned by variable capacitance; it had two sets of interleaved semicircular metal plates - one moveable and one stationary - with an air gap between them, and turning the tuning knob and shaft varied the area that the plates overlapped to change the capacitance. The old car radios I took apart had pulleys and strings that the buttons operated that turned a shaft. Pulling a knob out disconnected the button from the string and pushing it in clamped it. Why couldn't that shaft be to a variable cap?
 
ggunn said:
it had two sets of interleaved semicircular metal plates - one moveable and one stationary - with an air gap between them, and turning the tuning knob and shaft varied the area that the plates overlapped to change the capacitance. The old car radios I took apart had pulleys and strings that the buttons operated that turned a shaft.
Click to expand...
we are being dated by this stuff that kids today would have no idea

nos-antique-radio-variable-air-capacitor-ham-tuning-condenser-365-mmfd-section-3c6d943559d971753495e67504a7f52a.jpg
 
160805-1938 EDT

FionaZuppa:

Very nice photograph. Was this for a superhet? I can't see whether the left capacitor has smaller plates.



ggunn:

A rotary air capacitor has mass, inertia, friction, and a typical rotation of 180 deg. Further there is automotive vibration and space requirements. I believe that typical pushbutton motion was about 1/4". This small pushbutton motion had to tune from one end of the band to the other. A difficult job to combine with a rotary capacitor.

I went looking for my 1952 car radio and did not find it. It used powdered metal rods to adjust inductance for tuning with a unique mechanical mechanism.

A 1935 or so car radio I played with used an air variable capacitor wiyh a flexible shaft from a steering column mounted control box. No pushbutton selection. Some time between 35 and 45 the shift to inductive tuning was made.



Zenith was a major manufacturer of radios for Ford in the 1940s. At Zenith at the end of the assembly line the troubleshooter used his eyes, ears, brains, a screwdriver, and fingers as his troubleshooting instruments. High voltage DC might be present in the range of 300 or more volts. A screwdriver was a good high voltage voltmeter, and also a signal generator.

.
 
160805-2049 EDT

Back to the noise problem.

I now have battery power to the Panasonic and no conductive connection to the AC power system. WJR is still at S = 9. Before sundown and Feit off the off channel background was about S = 0.5 .

It is now darker and off channel is up to S = 1.5 or 2 from distant transmitters.

Turning on Feit and the background rises to S = 2 to 3. There is no interference to the audio on WJR. Thus, the primary problem from Feit is conductive, and probably from the hot wire. I believe the Panasonic has a a tuned RF amplifier and better selectivity than my car radio. I don't know how car radio front ends are made these days.

.
 
160807-1653 EDT

FionaZuppa:

My Rigol scope reads about 0.1 ohm from a BNC shell to the power cord EGC. One of my Tektronix scopes reads around 3 ohms from BNC shell to EGC.

,
 
gar said:
160805-1938 EDT

FionaZuppa:

Very nice photograph. Was this for a superhet? I can't see whether the left capacitor has smaller plates.
Click to expand...
just an online pic. i had my days of yanking out old car radios for the latest/greatest in digital stuff, then i would take apart the old crud just to see what was inside, most of them had a axial tunable cap right on the control knob, and explains why when you tuned to 99.1 the slide indicator was at 100, darn cables were worn out/stretched.
 
Status
Not open for further replies.
Top