CFLs... what's in your socket?

You're getting a momentary voltage drop on your lighting circuit. If all of your connections in the panel are tight and the wire to your compressor is correctly sized the only thing I could think of would be a new/ larger starting cap for your compressor or a ups on the light circuit.
 
Curtis_H said:
I put some retro fit leds in my 4ft fluorescent shop fixtures. With the kit you install new tombstones and do away with the old ballast. Now when my air compressor runs the lights seem to strobe really fast. I can run any thing else in the shop including welder, a/c, saws, etc with no issues. Any one have any idea how to fix this. The few fluorescent fixtures I have left the air compressor has no effect.
Click to expand...
Is the air compressor 110 or 220v? If 110 make sure they're not pulling off the same leg. It sounds like you have a voltage drop issue.
 
braxton357 said:
You're getting a momentary voltage drop on your lighting circuit. If all of your connections in the panel are tight and the wire to your compressor is correctly sized the only thing I could think of would be a new/ larger starting cap for your compressor or a ups on the light circuit.
Click to expand...
Didn't see this post before I replied :lol:
 
The compressor is 240volt. It may be a voltage drop. I’ll try and check that this weekend. Building is only fed with number 6 wire and is on a 60 amp breaker. That’s the way it was when I moved in. I haven’t really had a problem with it until the lights but may need to upgrade the wire feeding the building.
 
skyhighZJ said:
Completely untrained but a question. Are LED lights more susceptible to voltage fluctuations than cfl or fluorescent bulbs?
Click to expand...
It depends on the driver.
LEDs actually run at 3.3-5v DC. So all the LED lights that run in 110v sockets have a built in AC-DC transformer and driver. The susceptibility of that transformer to dropping out depends on how it's made and the bandwidth it can tolerate. The ones made to be dimmable will definitely be more robust.
 
a_kelley said:
LEDs can be modulated brightness by changing voltage input.
Click to expand...
Eeeeh, yes and no. Technically, yes, if you drop the DC voltage below the forward voltage of the LED (e.g. if it's a 3.7v LED, it will brighten between like 3.2.7 and 3.7). But since house AC is 110v AC, the necessary driver transformer is kind of like a buffer for AC power fluctuations, including decreases. So whether it dims, or just goes out altogether, depends on the behavior of that driver and its input vs output profile. A lot of them are just on or not, so when you try and dim the LED light (w/ a typical A/C dimmer) it could just go out altogether. You gotta have a dimmer made for LEDs, and an LED bulb made to be dimmed.

It would all be so much easier if houses and everything just worked on DC... but there's that distribution problem....
 
RatLabGuy said:
Eeeeh, yes and no. Technically, yes, if you drop the DC voltage below the forward voltage of the LED (e.g. if it's a 3.7v LED, it will brighten between like 3.2.7 and 3.7). But since house AC is 110v AC, the necessary driver transformer is kind of like a buffer for AC power fluctuations, including decreases. So whether it dims, or just goes out altogether, depends on the behavior of that driver and its input vs output profile. A lot of them are just on or not, so when you try and dim the LED light (w/ a typical A/C dimmer) it could just go out altogether. You gotta have a dimmer made for LEDs, and an LED bulb made to be dimmed.

It would all be so much easier if houses and everything just worked on DC... but there's that distribution problem....
Click to expand...
I've seen led tubes driven straight by 110, must be series strings of LEDs. No transformers. Ymmv.
 
a_kelley said:
I've seen led tubes driven straight by 110, must be series strings of LEDs. No transformers. Ymmv.
Click to expand...
There MUST be a driver or transformer somewhere. You just don't see it. Probably built ether into the base of the unit, or at every light. It's not physically possible for a light emitting diode to do it's thing at 110v AC.
Even the little strips of LEDs that are in Christmas lights, or use for under cabinet lights etc, they have a driver.
 
RatLabGuy said:
There MUST be a driver or transformer somewhere. You just don't see it. Probably built ether into the base of the unit, or at every light. It's not physically possible for a light emitting diode to do it's thing at 110v AC.
Even the little strips of LEDs that are in Christmas lights, or use for under cabinet lights etc, they have a driver.
Click to expand...
even if Vf of the series strings totals 110v (nominal)?
 
a_kelley said:
even if Vf of the series strings totals 110v (nominal)?
Click to expand...
Yes, b/c it's still alternating current. Has to be DC, otherwise the diodes would flicker really fast, only turning on briefly at one very specific spot on the sine wave.
 
RatLabGuy said:
Even the little strips of LEDs that are in Christmas lights, or use for under cabinet lights etc, they have a driver.
Click to expand...

Most LED Christmas/string lights don't even have a rectifier circuit. Some expensive ones do, but the rest flicker. You don't notice it unless they're the sole source of light for something fast-moving.
 
LED brightness is controlled by current, not voltage. If you're controlling brightness with voltage, there is something that is controlling the LED current based on that voltage.

Once the LED reaches forward conduction mode (when the supply voltage is greater then the forward voltage of the LED), the current needs to be limited, else the LED will pop like a fuse. That's all the Christmas bulbs usually use for current limiting; there are 30-whatever LEDs in series (the LED string itself performs the rectifier function) and a single resistor for the entire string to set the string current. That's a little reckless for a robust LED design (I would hope there is a reverse polarity diode as part of the LED string so the fragile LEDs are less stressed), but works for cheap consumer products with a very fixed supply voltage.

Because a simple resistor controls current based on the voltage across it, a variable voltage can also be used to do brightness control if things like supply voltage are simple (120VAC is as simple as it gets). Change the input voltage to the resistor, and the current through the resistor changes.
There are many LED drivers that do much more sophisticated things like constant current control, PWM brightness control, have switching power supply functions, etc., depending on what source your power is coming from and what you need your efficiency to be. Almost anything with any level of sophistication has some type of current sink driver or current controller to solve power problems or add product functionality.

LED replacement bulbs for your house generally have power supply and/or driver circuits built in, especially if they use only a high flux LEDs. With a string of only a few LEDs, you have a lot lower forward voltage than the 120VAC supply voltage, so you would generally use something like a buck converter or PFC controller to get into a proper range of DC voltage and current without wasting that big voltage delta as heat.
 
Last edited:
Fabrik8 said:
LED brightness is controlled by current, not voltage. If you're controlling brightness with voltage, there is something that is controlling the LED current based on that voltage.

Once the LED reaches forward conduction mode (when the supply voltage is greater then the forward voltage of the LED), the current needs to be limited, else the LED will pop like a fuse. That's all the Christmas bulbs usually use for current limiting; there are 30-whatever LEDs in series (the LED string itself performs the rectifier function) and a single resistor for the entire string to set the string current. That's a little reckless for a robust LED design (I would hope there is a reverse polarity diode as part of the LED string so the fragile LEDs are less stressed), but works for cheap consumer products with a very fixed supply voltage.

Because a simple resistor controls current based on the voltage across it, a variable voltage can also be used to do brightness control if things like supply voltage are simple (120VAC is as simple as it gets). Change the input voltage to the resistor, and the current through the resistor changes.
There are many LED drivers that do much more sophisticated things like constant current control, PWM brightness control, have switching power supply functions, etc., depending on what source your power is coming from and what you need your efficiency to be. Almost anything with any level of sophistication has some type of current sink driver or current controller to solve power problems or add product functionality.

LED replacement bulbs for your house generally have power supply and/or driver circuits built in, especially if they use only a high flux LEDs. With a string of only a few LEDs, you have a lot lower forward voltage than the 120VAC supply voltage, so you would generally use something like a buck converter or PFC controller to get into a proper range of DC voltage and current without wasting that big voltage delta as heat.
Click to expand...
yes yes. But I've seen 8ft tube with what looks to be 200+ individual LEDs.. and they flicker horridly, at least to my eyes. Yes, they should be current controlled, pfc, pwm, etc.. but "cheap" (mass produced, regardless of price tag), if they can get the hours of operation more simply, they will cut as many corners as possible..
 
a_kelley said:
yes yes. But I've seen 8ft tube with what looks to be 200+ individual LEDs.. and they flicker horridly, at least to my eyes. Yes, they should be current controlled, pfc, pwm, etc.. but "cheap" (mass produced, regardless of price tag), if they can get the hours of operation more simply, they will cut as many corners as possible..
Click to expand...

Not surprising for cheap LED tubes. Like I said, the bare minimum is a resistor per LED string and hopefully a diode for the entire device. If it's that crude, you're half-wave rectifying at 60Hz, which is likely the flicker freq you're seeing. No different than seeing the flicker of a 50-60Hz screen refresh rate on a TV or monitor.

200 LEDs just means there are multiple strings in parallel, so maybe 10+ parallel strings because white LEDs have a relatively high Vf.
 
Fabrik8 said:
If it's that crude, you're half-wave rectifying at 60Hz, which is likely the flicker freq you're seeing. No different than seeing the flicker of a 50-60Hz screen refresh rate on a TV or monitor.
Click to expand...
It's the same notion, but you're much more likely to notice it than on a TV/Monitor.
Any LCD based device - which is all of them since the mid 2000s - actually only has a "potential" change of state for any pixel at 60 Hz or whatever the refresh rate is. The pixel actually stays on for all 16.66 ms of the cycle, and continues staying on unless it is told to change by the video data feed. A monitor just showing white, constantly, doesn't (shouldn't) flicker. Now if the driver is clearing to black on every frame or something, yeah....
This is in contrast to old TVs/monitors that used a CRT, which only illuminates the pixel for <1ms, then has a decaying fluorescence. Seeing a slowed down time trace of a photodetector on each device is rather educational and entertaining. (this comes up in our lab science with human subject visual perception studies all the time).

But in the case of LEDs that are being AC cycled, it's only on for a portion of the cycle, b/c LEDs have a really fast response rate. So most of the time between strikes on the cycle it's not illuminated. The shitty temporal blur of the human visual system (photoreceptor wiring etc) masks the rest.

It's funny, I'll admit that in all my experience w/ measuring the temporal properties of different devices for use in perceptual research, I had never given any thought whether running LEDs on straight 60Hz AC would even be viable for constant light.
 
Back
Top