Probably less than 50 or 60Hz AC.

Just a point of interest - 500Hz was a common strobe rate for multiplexed LED displays.

I'm not sure what you're trying to say here. That I should use 50-60Hz PWM? With or without LP filter? Are you talking about running just LEDs like this?

 

1. It is nuts to run your leds at 1khz, unless if course you have a pair of golden eyes.

TV manufacturers used to brag about their 100hz or 120hz TVs until no one could notice a difference between them and their 50hz 60hz counterparts.

2. It is nuts to put a lpf before your lpf. A. It woukd nt work, and B. It is not necessary given your eyes persistence of vision.

Start simple and if that doesn't work, try more complex solutions.

 

Right, because then it becomes linear dimming, duh.
But that may present another problem. 1kHz PWM for LEDs is fine, but ideally I'd like to be able to use a single source to dim the bulbs and LEDs. 1kHz is audible and without an LP filter will likely cause ringing/buzzing/hum in the filament no?

The thermal time constant of the filament in a 12 V lamp is an eternity compared to the PWM period, the lamps will be happy and quiet.

Audible noise is usually a product of magnetic or piezoelectric effects, neither of which is a problem here.
On rare occasions I have had ceramic capacitors whine audibly in PWM dimmers, when exposed to pulsing voltages...

 

The thermal time constant of the filament in a 12 V lamp is an eternity compared to the PWM period, the lamps will be happy and quiet.

Audible noise is usually a product of magnetic or piezoelectric effects, neither of which is a problem here.
On rare occasions I have had ceramic capacitors whine audibly in PWM dimmers, when exposed to pulsing voltages...

Excellent information, thank you.


So with all of your help this is where I am now:
- Combined bulb and LED dimming through a P-FET
- Current limit resistor for each LED
- No RC/LP or filtering
- PWM 500-1kHz

One last question:
In all of my reading for this dimming I came across an article suggesting a small capacitor (100nF) to limit ringing, and a large capacitor (load in amps X 1000uF approx, in my case load maximum is expected to be no more than 5A so 4700uF). Thoughts on this?

 

1. It is nuts to run your leds at 1khz, unless if course you have a pair of golden eyes.

TV manufacturers used to brag about their 100hz or 120hz TVs until no one could notice a difference between them and their 50hz 60hz counterparts.

.

I've heard American visitors to the UK complain about flickering lights and TV pictures - what you're used to can have a big impact on what you perceive.

 

Right, because then it becomes linear dimming, duh.
But that may present another problem. 1kHz PWM for LEDs is fine, but ideally I'd like to be able to use a single source to dim the bulbs and LEDs. 1kHz is audible and without an LP filter will likely cause ringing/buzzing/hum in the filament no?

That is only true for Pwm motors and transformers. An LED does not have any magnetostriction or voice coils that can hum. You will not hear the electrons flow through the copper or the LED.

 

That is only true for Pwm motors and transformers. An LED does not have any magnetostriction or voice coils that can hum. You will not hear the electrons flow through the copper or the LED.

I totally agree about that in regards to LEDs, but the same is not true of bulbs. Their filaments are inductive loads, and have been known to hum when driven by PWM, which I am trying to avoid.

 

I totally agree about that in regards to LEDs, but the same is not true of bulbs. Their filaments are inductive loads, and have been known to hum when driven by PWM, which I am trying to avoid.

Get the bulbs with straight rather than micro-coiled filaments.

 

I may not have that much flexibility, as such I would like to design for the possibility of not having that option.

 

I may not have that much flexibility, as such I would like to design for the possibility of not having that option.

I can't hear any chorus of75 watt bulbs in my house. All running at 50% duty cycle (should be loudest output). Some circuits have up to 12 X 75 watt bulbs all in phase with each other. From your concern, my living room should sound like a Bose test room at the local electronic store. Not happening here.

 

That is only true for Pwm motors and transformers. An LED does not have any magnetostriction or voice coils that can hum. You will not hear the electrons flow through the copper or the LED.

A strange anecdote from years ago - someone was making adjustments to a kit built radio, for some reason the speaker got disconnected, but the audio could still be heard. It turned out the metal cases on the transistors were ringing. AFAICR: the author stated that the transistors were of Italian origin - I don't remember ever hearing a satisfactory explanation for this strange event.

 

A strange anecdote from years ago - someone was making adjustments to a kit built radio, for some reason the speaker got disconnected, but the audio could still be heard. It turned out the metal cases on the transistors were ringing. AFAICR: the author stated that the transistors were of Italian origin - I don't remember ever hearing a satisfactory explanation for this strange event.

I could see a slug tuned RF coil with a metal can making an audible noise but a transistor is an odd one.

Also, "transistors of Italian origin" rings a bell. As high-school ended, a friend finally saved enough cash to buy a "high end" amplifier to start his system. At the last minute, the salesman at the audio store (The Sound of Music) talked him into an all-in-one unit (I think it was Blauplunkt) because he was moving into a dorm in a few months. His dad came home and opened it up to check out the "German quality" - only to find Italian-made SGS (Società Generale Semiconduttori) transistors. He managed to return it the next day and get the receiver he originally planned to get. Interestingly, that was in a south Minneapolis suburb and the audio store, Sound of Music, eventually changed the name of the local chain to BestBuy - the HQ is still in the south Minneapolis suburb.

 

I could see a slug tuned RF coil with a metal can making an audible noise but a transistor is an odd one.

Also, "transistors of Italian origin" rings a bell. As high-school ended, a friend finally saved enough cash to buy a "high end" amplifier to start his system. At the last minute, the salesman at the audio store (The Sound of Music) talked him into an all-in-one unit (I think it was Blauplunkt) because he was moving into a dorm in a few months. His dad came home and opened it up to check out the "German quality" - only to find Italian-made SGS (Società Generale Semiconduttori) transistors. He managed to return it the next day and get the receiver he originally planned to get. Interestingly, that was in a south Minneapolis suburb and the audio store, Sound of Music, eventually changed the name of the local chain to BestBuy - the HQ is still in the south Minneapolis suburb.

SGS has become a highly respected semiconductor manufacturer, last seen trading under the brand; ST after combining with the French Thomson company - who acquired the American RCA BTW. Not sure if that's just the TV and domestic side of the business though.

 

but the same is not true of bulbs.

If you experiment a little, you may find how "regular" most people's eyes are, regardless of their claims otherwise.

 

SGS has become a highly respected semiconductor manufacturer, last seen trading under the brand; ST after combining with the French Thomson company - who acquired the American RCA BTW. Not sure if that's just the TV and domestic side of the business though.

It was in their doldrum days - after fairchild spun them off (Fairchild SGS in Europe until early 1970s). The Italian government wouldn't let them fail so they operated the company as SGS until acquired / merged with Thompson.

My friends dad was not an engineer - just a WWII vet, pilot for Northwest, and Ham guy. He had some preconceived notions about Italian Engineering (and German Cooking).