10watts for all 4, will it work

[ogden]

Anyway soon all LED bulbs will have no switching converters or bulk capacitors which are main contributors to price and failures too.
From inside they will look like this:



Not like this:



Such design with PF approaching 0.98 acts like resistive load and will have no problems to work from any waveform you drop at it. Just don't exceed voltage rating

[kalel]

If I have seen it right, it looks like a circuit with an integrated driver (I don't know what else to call it), similar to the video below. Don't those designs experience a lot of flicker as is?



There might be new designs that solve the flickering, or perhaps what you showing in the image is a different type of circuit all together.

[ogden]

perhaps what you showing in the image is a different type of circuit all together.

Different indeed.

Don't those designs experience a lot of flicker as is?

Exactly opposite.

They are based on AC-direct regulation IC which progressively switches on each of 4 LED strings until all are illuminated at the mains voltage’s peak. Note that for simplicity circuit below shows just 3 strings. This allows to "cover" with illumination nearly all the AC mains cycle. Resulting light is nearly flickerless.



Article here: https://www.electronicsweekly.com/news/why-not-direct-ac-drive-your-led-string-2016-04/

I already mentioned ON Semi FL77904 which is single chip 4-string IC requiring few small passives, bridge rectifier and that's it. Power factor better than 0.98(!). It is beauty if we can say so about engineering.

http://www.onsemi.com/PowerSolutions/product.do?id=FL77904

[Zero999]

PWM has a similar effect to phase control but it is not the same thing.

Indeed dimming is right word for AC mains phase control but principles to PWM is the same - as I already explained. So to say "Dimming is not PWM. Period." is not quite correct as well. 

Dimming is not the right word for AC mains phase control, since it can be used for other purposes than lighting, such as controlling the speed of motors and the power output of heaters.

PWM isn't normally used in AC circuits

It is. Modern PFC uses hi-freq PWM for AC mains power factor correction. Waveform looks like you are showing "the real PWM control of AC".

Not quite. Power factor controllers, PWM the DC side of the rectifier, not the AC side.

To stand the highest chance of success, use 100Hz or 120Hz, rather than 50Hz, because that's what the LED driver IC will see on the DC side of the rectifier.

You (again) ignored what I say. Reminder in bold: 50Hz bipolar square wave coming out of push-pull or full-bridge inverter after rectification will look like pulsing 100Hz.

Oh I see what you meant now. No I wouldn't  recommend doing that. Chopping the output of the inverter, would result in much higher frequency components, which might not play nicely with the LED driver, unless you're talking about something else.

It would be better to rectify the output of the inverter, then PWM that at 100Hz.

[ogden]

Chopping the output of the inverter, would result in much higher frequency components, which might not play nicely with the LED driver, unless you're talking about something else.

Again you assume inferior design. 

All UPSes with "simulated sine" output have what we call here "chopped inverter". Yet they somehow pass EMC regulations, computers are happy with such waveform, most likely LED bulbs will too. Chopping happens on primary side of the transformer, output have just quite simple (L)C filter.



Having proper snubbers on primary and some filtering on secondary winding of the transformer, output looks not that bad at all (from hi-freq ringing point of view):

https://www.hardwareinsights.com/database-of-ups-output-waveforms/

It would be better to rectify the output of the inverter, then PWM that at 100Hz.

Such waveform will cause much worse inrush current problems than modified sine wave, drooper-capacitor LED bulbs most likely will be very unhappy about such input too.

[edit] When you feed such modified sine into your "LTspice LED bulb", you will see that 10ohm resistor dissipates only ~1W at full power and ~1.1W when 50% dimmed. This BTW is just 2x exceeding "pure sine rating"

[Zero999]

Chopping the output of the inverter, would result in much higher frequency components, which might not play nicely with the LED driver, unless you're talking about something else.

Again you assume inferior design. 

All UPSes with "simulated sine" output have what we call here "chopped inverter". Yet they somehow pass EMC regulations, computers are happy with such waveform, most likely LED bulbs will too. Chopping happens on primary side of the transformer, output have just quite simple (L)C filter.



Having proper snubbers on primary and some filtering on secondary winding of the transformer, output looks not that bad at all (from hi-freq ringing point of view):

https://www.hardwareinsights.com/database-of-ups-output-waveforms/

Now you're making an assumption too: that output of the original poster's inverter is a modified sine wave. You've also forgotten that the original poster doesn't want to or cannot change the inverter.

Unfortunately we have to make assumptions, in the absence of all the information. The original poster has not posted a schematic or an oscillogram of the output waveform of their inverter, so we don't know the full story.

It would be better to rectify the output of the inverter, then PWM that at 100Hz.

Such waveform will cause much worse inrush current problems than modified sine wave, drooper-capacitor LED bulbs most likely will be very unhappy about such input too.

DC PWM will  be no worse or no better than a modified sine wave, for the circuit I simulated in LTSpice. Yes a capacitive dropper won't work from PWM DC because, it will be blocked by the capacitor.

[edit] When you feed such modified sine into your "LTspice LED bulb", you will see that 10ohm resistor dissipates only ~1W at full power and ~1.1W when 50% dimmed. This BTW is just 2x exceeding "pure sine rating"

Yes, still not good though and could lead to reduced reliability.

[ogden]

Now you're making an assumption too: that output of the original poster's inverter is a modified sine wave. You've also forgotten that the original poster doesn't want to or cannot change the inverter.

You'v forgotten that you discussing to me here, not original author. I do not share authors plans, provided mine in return. Obviously when I say something about inverters and so on, I am considering my view, not original posters plans. That's it. Tired of reminding. Bye.

[Zero999]

Now you're making an assumption too: that output of the original poster's inverter is a modified sine wave. You've also forgotten that the original poster doesn't want to or cannot change the inverter.

You'v forgotten that you discussing to me here, not original author. I do not share authors plans, provided mine in return. Obviously when I say something about inverters and so on, I am considering my view, not original posters plans. That's it. Tired of reminding. Bye.

And you've forgotten that this is not your thread! By barging in to some one else's thread, you've derailed it and created pages of useless information, which is totally irrelevant to the original poster's question, which is not helpful.

If you wanted to discuss your inverter/LED lamp design, then you should have created your own thread, rather than derailing someone else's!

I apologise to the original poster for playing my part in this. I should have recognised this and stopped responding to you, long ago.

[mikerj]

Indeed dimming is right word for AC mains phase control but principles to PWM is the same - as I already explained. So to say "Dimming is not PWM. Period." is not quite correct as well. 

Not a single person has stated such nonsense as "Dimming is not PWM. Period."  If you have to resort to making things up then your argument is a lost cause.

As for you PWM AC waveform doubling the power dissipation in the 10R resistor, why do you think this is acceptable?  Do you believe that mass produced LED lamps are all sufficiently over-engineered that you can ignore this?

[ogden]

By barging in to some one else's thread, you've derailed it and created pages of useless information, which is totally irrelevant to the original poster's question, which is not helpful.

So you say that someone who offers alternate solution for consideration, actually creates "pages of useless information, which is totally irrelevant to the original poster's question". With emphasis on "totally irrelevant".

If you wanted to discuss your inverter/LED lamp design, then you should have created your own thread, rather than derailing someone else's!

Are you serious? How many times did you offer something that differ from original poster's plans? - Shall we count? Or maybe you did not offer anything useful, just denied everything you did not agree to?

[ogden]

Not a single person has stated such nonsense as "Dimming is not PWM. Period."  If you have to resort to making things up then your argument is a lost cause.

Well, Hero999 stated: "It supports TRIAC dimmers, not PWM! They are not the same thing!". In my opinion saying "Dimming and PWM are not the same thing!" (pay attention to exclamation sign mark) is the same as "Dimming is not PWM".

As for you PWM AC waveform doubling the power dissipation in the 10R resistor, why do you think this is acceptable?

I did agree that this is not acceptable. Perhaps you missed that (as usually happens here in this forum thread):

Checked your LTspice design and agree - "dimmer incompatible" LED bulb which is based on constant current IC, most likely will fail due to overheat.

[ogden]

By barging in to some one else's thread, you've derailed it and created pages of useless information, which is totally irrelevant to the original poster's question, which is not helpful.

So you say that someone who offers alternate solution for consideration, actually creates "pages of useless information, which is totally irrelevant to the original poster's question".

Actually I did good service to thread - created special attention to square wave inverters which are not compatible with some LED bulbs. I doubt that 5$ inverter mentioned by original poster, is pure sine wave one. Disregarding this fact you did offer your ring oscillator. No further comments...

[dave_j_fan]

If you wanted to discuss your inverter/LED lamp design, then you should have created your own thread, rather than derailing someone else's!

I apologise to the original poster for playing my part in this. I should have recognised this and stopped responding to you, long ago.

its ok Hero999, not your fault
I request the other poster  to  keep discussion to target the plan of mux on the AC line .
Let us not go deep into LED drivers and million inverter design. Then it will be like making a chair for 2yo to 90yo with a size that fits all .
We are engineers , seek to make viable products not science explorer .


-some good news here!!! -
 The plan works by unloading the inverter 
 a small inverter that used to flicker after 15watts now happily taking 25W , pretty amazing bright lights .
 i am too lazy to make lux meter  but  compared lightof 1 bulb with MUX and without MUX .

 Did it save power ?
 Not much,9watter was pulling  800mA  DC
 25watts load was showing 600mA / per 5watts  drain

 did the lamps over heat ?
 NO
   
 i wish someone could fund me for a patent! 

 Many experts cant achieve to a solution because they lose focus ,  so it is better the discussion can limited to reach some  solution
  these are the  parameters for a BETA of the AC MUX
=============
- Approx square wave ( a little different in my case  ) inverter , take frequency 10khz
-target is to get atleast 4 bulbs on 10w inverter
-all bulbs are SMPS ( no future tech and , cap dropper)
===================================

There are 2 ways
-simulation model
-real one

Second one will save time initially

 

[Zero999]

By barging in to some one else's thread, you've derailed it and created pages of useless information, which is totally irrelevant to the original poster's question, which is not helpful.

So you say that someone who offers alternate solution for consideration, actually creates "pages of useless information, which is totally irrelevant to the original poster's question". With emphasis on "totally irrelevant".

If you wanted to discuss your inverter/LED lamp design, then you should have created your own thread, rather than derailing someone else's!

Are you serious? How many times did you offer something that differ from original poster's plans? - Shall we count? Or maybe you did not offer anything useful, just denied everything you did not agree to?

You didn't contribute any schematics, simulations or ideas to the thread. I posted a simulation which demonstrated what happens to the input stage of a mains LED, when driven with various waveforms. All you did was derail the thread with nonsense about it being fine to overstress parts, dimmable lamps and PWM and phase control being the same thing.

If you wanted to discuss your inverter/LED lamp design, then you should have created your own thread, rather than derailing someone else's!

I apologise to the original poster for playing my part in this. I should have recognised this and stopped responding to you, long ago.

its ok Hero999, not your fault
I request the other poster  to  keep discussion to target the plan of mux on the AC line .
Let us not go deep into LED drivers and million inverter design. Then it will be like making a chair for 2yo to 90yo with a size that fits all .
We are engineers , seek to make viable products not science explorer .


-some good news here!!! -
 The plan works by unloading the inverter 
 a small inverter that used to flicker after 15watts now happily taking 25W , pretty amazing bright lights .
 i am too lazy to make lux meter  but  compared lightof 1 bulb with MUX and without MUX .

 Did it save power ?
 Not much,9watter was pulling  800mA  DC
 25watts load was showing 600mA / per 5watts  drain

 did the lamps over heat ?
 NO
   
 i wish someone could fund me for a patent! 

 Many experts cant achieve to a solution because they lose focus ,  so it is better the discussion can limited to reach some  solution
  these are the  parameters for a BETA of the AC MUX
=============
- Approx square wave ( a little different in my case  ) inverter , take frequency 10khz
-target is to get atleast 4 bulbs on 10w inverter
-all bulbs are SMPS ( no future tech and , cap dropper)
===================================

There are 2 ways
-simulation model
-real one

Second one will save time initially

You can't rely on your eyes to make quantitative measurements. A lux meter is the only way to perform objective tests.

I suggest doing long term trials. Just because they're SMPS, it doesn't mean their reliability won't be affected.

Have you looked at the output current and voltage waveform of your inverter with an oscilloscope?

[ogden]

You didn't contribute any schematics, simulations or ideas to the thread.

I did. Very early in this thread I said: "I would run all the LED bulbs in parallel - to save on wiring in many installations and create 10W PWM-dimmable AC square wave inverter with input current limiter. Current limiter as safeguard against inverter overload and output short. "

I posted a simulation which demonstrated what happens to the input stage of a mains LED, when driven with various waveforms.

You posted simulation which uses inverter of infinite power just to prove your uneducated point, ignoring fact that original poster mentioned 10W (in the topic of this thread). If we take in account limited power of the inverter which can sustain let's say 1A max output current, then "parts overstress" picture changes dramatically and inrush resistor dissipates just 155mW average during 1sec of simulation which is few times less than nominal which you put at 500mW. Simulation attached.

All you did was derail the thread with nonsense about it being fine to overstress parts, dimmable lamps and PWM and phase control being the same thing.

It is actually you who are spilling nonsense all over this thread by literally attacking me with sentences like this, providing misleading "facts" using simulations that proves nothing more than how incompetent or careless you are. I do not recall saying "it is fine to overstress". I said - if everything is designed properly then nothing will blow up meaning there will be no parts overstress.

[ogden]

For those who still believes that square wave is much worse for LED bulb than sine, I offer to consider that bulbs shall be designed to withstand power-on at any phase angle of AC mains including max peak voltage, otherwise they will blow-up nearly every power-on. Best and worst case dissipation of inrush resistor results for AC mains sine attached. Best case obviously - to show that it gives result we already know, 0.5W @ 1sec.

Worst case is much more interesting.. because it's 2.49W is pretty close to "LED blown-up by square wave" simulation result of 2.54W 

[ogden]

I request the other poster  to  keep discussion to target the plan of mux on the AC line .

Sorry, I had to correct misleading information Hero999 provided here - regarding LED bulbs on square wave supply.

AC line? Do you have "pure sine wave" inverter? Please show it's output waveform on the scope so we all know what you have.

[dave_j_fan]

orry, I had to correct misleading information Hero999 provided here - regarding LED bulbs on square wave supply.

AC line? Do you have "pure sine wave" inverter? Please show it's output waveform on the scope so we all know what you have.

no issue . I hope to know additional possibilities .

Q.Will the driver less LED help me to cut power usage

[ogden]

Q.Will the driver less LED help me to cut power usage

Most likely not. If driver-less bulbs would be universally more effective- they would dominate market, but this is not the case. All the hi-end LED bulbs are equipped with some kind of driver. Anyway you shall not rely on assumption. Better search internet for research/reviews which are comparing consumed power versus light output.

[dave_j_fan]

Most likely not. If driver-less bulbs would be universally more effective- they would dominate market, but this is not the case. All the hi-end LED bulbs are equipped with some kind of driver. Anyway you shall not rely on assumption. Better search internet for research/reviews which are comparing consumed power versus light output.

i was asking  if i use AC MUX

[ogden]

Most likely not. If driver-less bulbs would be universally more effective- they would dominate market, but this is not the case. All the hi-end LED bulbs are equipped with some kind of driver. Anyway you shall not rely on assumption. Better search internet for research/reviews which are comparing consumed power versus light output.

i was asking  if i use AC MUX

AC mux will unlikely change anything regarding various LED bulb efficiency. Frequency of supplied AC voltage instead can change a lot. If you use bulbs at much higher frequency - you are on your own. You shall get precision LUX meter and do your own efficiency measurements for various types of led bulbs.