Non-dimmable LED lightbulbs are damaged by Triac dimmers?

[Flyback]

Supposing I have a mains LED lightbulb which is not meant to be dimmable.
Now suppose I accidentally connect a triac dimmer to it.
Will the constant "smashing" on and off of the mains by the triac switching repeatedly over-excite the LED bulb's EMI filter such that it fails prematurely?

 

[ronsimpson]

I do not think there will be failures.
I have read reports of non dimmable bulbs and they do not have a good relationship between percent of dim and amount of light. I believe between 100% and 45% there way no change in light.

 

[Flyback]

Thanks, BTW, we are bringing to market a new type of LED lightbulb dimmer which will sweep the world!
It is literally just a pushbutton switch by the side of your light switch.......when you press it, a mosfet in-line with the mains live will switch off at the mains zero crossing, and then back on at the next mains zero crossing............a simple circuit inside the led lightbulb will detect this missing half cycle, and dim the led light down....or back up again if it was already dimmed right down.
As you can tell....you just keep pushing the button till you like the dimness.
...This type of dimmer, since it switches the fet at the zero crossings, will not damage any type of lightbulb.......obviously it needs the simple "detector" cct in the led ligthbulb, but that'll be easy to do.....it could also be doe with flu's.

Forget those old triac dimmers which just don't work with all types of dimmable led bulb...and indeed, leading edge dimmers can reduce led bulb lifetime if accidentally left on with a non dimmable led bulb due to resonances in the emi filter of the bulb.

Not one single dimmable led lightbulb on earth can work well with absolutely all leading edge or trailing edge dimmers on the market......its time for us to say goodbye to trailing edge or leading edge dimmers

 

[ronsimpson]

I wish you good luck with your idea.
There are probably 50 ICs that claim to work with triac edge dimmers.
It sounds like all LED and CF bulbs need to be modified to work with your circuit.

 

[jpanhalt]

I do not think there will be failures.
I have read reports of non dimmable bulbs and they do not have a good relationship between percent of dim and amount of light. I believe between 100% and 45% there way no change in light.

If there is no change in light output between 100% and 45%power, why not run them at 45% and save the energy?

I searched on this exact question,"do dimmers damage non-dimmable LED bulbs. Lost of hits. Everyone of them that commented on triac-type dimmers , including manufacturers, say they can be damaged. Here's one example (https://www.integral-led.com/education/dimmable-and-non-dimmable-led-lamps) :

John

 

[Nigel Goodwin]

If there is no change in light output between 100% and 45%power, why not run them at 45% and save the energy?

If you think about it you'll realise how silly that question is!

The LED lamp has a regulated SM PSU, so as the incoming 'voltage' drops it draws more current to keep consumption (and light output) constant.

 

[Flyback]

..it wouldn't drop if it detected that a dimmer was there and thus intelligently reduced its current.

Anyway, no bulb would need modification to withstand my circuit of ost #3 above.....that is, our circuit would not damage any bulb.
.....and Even CFLs could be made much more nicely dimmable with our kind of circuit (just increase fsw when dimmed)

If there is no change in light output between 100% and 45%power, why not run them at 45% and save the energy?

...I think what was meant by this is that some dimmers, when you turn them that far, don't actually dim the led light at all.....

This just supports what I am saying......triac dimmers are not meant for led bulbs, they were meant for incandescants.......if dimming led bulbs then we need to change the dimmer

....and the powering PH series, even that will not handle every type of leading or trailing edge dimmer...and it wont give dimming down to zero with every dimmer type........also, the PH series uses bleeder circuitry and damping circuitry to handle being dimmed , which is a pain and not wanted..plus it makes for inefficiencies.

 

[ronsimpson]

If there is no change in light output between 100% and 45%power, why not run them at 45% and save the energy?

For LED and CF bulbs there are many different ways to design them. The early designs often used a full wave rectifier and filter cap off the power line. This type of filtering only draws power at the peak of the cycle. Because only 15% of the time the bulb is drawing current.....duty cycle dimming is not very effective. AND Many of the power supply ICs do not function for 100mS after the power is applied. This soft start or start delay gets triggered at low duty cycle settings.

It is very possible to have the power line off for 50% of the time and the bulb pull the same power as at 100%. (same brightness)

 

[tcmtech]

Thanks, BTW, we are bringing to market a new type of LED lightbulb dimmer which will sweep the world!

So how long until you start a thread here arguing about some almost irrelevant design in order to get us to design an actual functional working circuit for your job and company that you can present to your boss as your own design?

 

[ronsimpson]

I can't find the patent now but...........
There are some designs where the IC watches the power line on a separate pin. If a triac is detected....(the power line drops to zero volts when it normally should not be zero)....the IC shuts off the LED for the time the triac is off. The power filter gets it power as described in my last post. The LED is enabled/disabled along with the triac.

The problem is that the LED bulb needs a load on the power line other than its self. When a triac opens up the power line, with out a load, the voltage does not drop instantly. There is capacitance on the power line that keep the voltage from dropping. When the voltage does not drop to zero the IC can not detect the turn off of the triac and does not turn off the LED for part of the cycle. There have been a number of papers written discribing LED and CF bulbs need one incandescent bulb in parallel to dim correctly.

Disclaimer: All this is dependent on what type of dimmer and what type of rectifier filter in the bulb.

 

[shokjok]

I can't see this "design" as being feasible or profitable. Triacs won't work with LEDs unless they are strung to work at line voltage.

 

[ronsimpson]

I can't see this "design" as being feasible or profitable. Triacs won't work with LEDs unless they are strung to work at line voltage.

That's what we are talking about. LED bulbs.

 

[MrAl]

Thanks, BTW, we are bringing to market a new type of LED lightbulb dimmer which will sweep the world!
It is literally just a pushbutton switch by the side of your light switch.......when you press it, a mosfet in-line with the mains live will switch off at the mains zero crossing, and then back on at the next mains zero crossing............a simple circuit inside the led lightbulb will detect this missing half cycle, and dim the led light down....or back up again if it was already dimmed right down.
As you can tell....you just keep pushing the button till you like the dimness.
...This type of dimmer, since it switches the fet at the zero crossings, will not damage any type of lightbulb.......obviously it needs the simple "detector" cct in the led ligthbulb, but that'll be easy to do.....it could also be doe with flu's.

Forget those old triac dimmers which just don't work with all types of dimmable led bulb...and indeed, leading edge dimmers can reduce led bulb lifetime if accidentally left on with a non dimmable led bulb due to resonances in the emi filter of the bulb.

Not one single dimmable led lightbulb on earth can work well with absolutely all leading edge or trailing edge dimmers on the market......its time for us to say goodbye to trailing edge or leading edge dimmers

Hi,

Are you saying that you are building a missing half cycle dimmer, or is that just the communications signal? What is actually going to time the LED itself?

If you say higher frequency PWM i will be happy.

 

[Flyback]

that's just the comms signal.
Sorry I don't understand "time the LED".
The leds will get dimmed by adjusting the fed back signal, which delivers led current information

 

[ronsimpson]

From what you said above; it appears there is no up or down command. I think there is just a change command. It could be you can remove a + or a - 1/2 cycle to send UP or DOWN.

 

[MrAl]

that's just the comms signal.
Sorry I don't understand "time the LED".
The leds will get dimmed by adjusting the fed back signal, which delivers led current information

Hi,

"dim" the LED...

 

[ronsimpson]

The leds will get dimmed by adjusting the fed back signal, which delivers led current information

Many white LED dimmers do not reduce the current because the color of white changes with current.
Many white LED dimmers PWM the LED. To get 50% light, turn off the LED current 1/2 the time. Keep the current at full current (when on). This way the color does not change.

 

[Flyback]

PWM dimming is a good idea.......I am going to think of a solution with that now....but here is the analog dimming solution.(attached)
This is, quite simply, going to sweep the world, and you saw it here first....theres no simpler way to do it.
And its obvious to everybody that its madness to keep manufacturing triac dimmers , which are meant for incandescent bulbs, when nobody is using incandescants any more.

Seriously, is anybody going to defend the use of triac dimmers with led lightbulbs? Its madness.....half the time you don't know what the holding current is with any triac, as theyre so slack in tolerance.

We all know that not one single led lightbulb on the earth can dim without flicker, and down to zero, with every single triac dimmer on the market.
Also, non dimmable led bulbs are damaged by use with triac dimmers...due to emi filter resonance.

 

[ronsimpson]

The LT3799 has a triac dimmer function. See page 10 of the data sheet. It functions like I said. When the triac is open the LED current is zero.

The LT3799 does not need two or one opto isolators. You don't need to monitor the LED current form the secondary side. The LT3799 can see the LED current from the primary side. (with out an isolator)

Your micro can feed a voltage into one of the control pins and do the same thing. Change the LED current by the control pin voltage.

The Vin_sense pin monitors the line voltage and knows if the triac dimmer is open or closed. It will also know if there is a missing 1/2 cycle. I think you micro can set on the LT3799's ground, monitor the line like the Vin_sense pin does (with out opto) and the micro can adjust the Crtl voltage thus the LED current.

I think you can remove all the opto isolators, op-amps, Q1, V2, v3, v5, v6, v7 etc.

 

[MrAl]

PWM dimming is a good idea.......I am going to think of a solution with that now....but here is the analog dimming solution.(attached)
This is, quite simply, going to sweep the world, and you saw it here first....theres no simpler way to do it.
And its obvious to everybody that its madness to keep manufacturing triac dimmers , which are meant for incandescent bulbs, when nobody is using incandescants any more.

Seriously, is anybody going to defend the use of triac dimmers with led lightbulbs? Its madness.....half the time you don't know what the holding current is with any triac, as theyre so slack in tolerance.

We all know that not one single led lightbulb on the earth can dim without flicker, and down to zero, with every single triac dimmer on the market.
Also, non dimmable led bulbs are damaged by use with triac dimmers...due to emi filter resonance.

Hi,

That looks like a step down switching regulator. You dont need to add PWM then because it already has that.

Opto's are often used for feedback in these kinds of designs. That's so that there is no interpretation of what the current is, it's an actual measurement of the current being controlled, and that means less attention to what the losses are in the transformer and associated circuitry and any variations that might occur. Feedback is almost always better than feedforward.
Opto degradation over long time periods however is an important design criterion that has to be carefully mitigated.

I dont think triac dimmers are that bad really but yes you should only use an LED bulb made for use with triac dimmers if you want to use one. The reason triac dimmers are so popular is because they can be made quite cheap with few components. Triac holding current is only an issue with certain designs.