Proton BASIC Community

This just gets more confusing! I just dug out the "resistor" shown:

Hi Charlie,
It looks like you have 24 LEDs per 5cm and 2 resistors - is that correct?  The only thing I could think of was as per the attached but I'm not sure it answers what you're seeing.

Cheers,
David

Quote from: charliecoutas on Jun 22, 2025, 10:03 AMThis just gets more confusing! I just dug out the "resistor" shown:

What is the resistor ohm value, measure both ways.

David: Yes, 24 leds per 5cm which gives 480 leds per meter. So my seven meter run has 3360 leds!

Fanie: I can't, it's ssooo small. I attacked it with a scalpel. But it might be possible to shove two sharp voltmeter probes into the plastic goo. I'll try that and come back here if it works.

Charlie

It worked, what a good idea Fanie. The "thing", with no power applied measures 33 ohms. With 24V across the strip (same strip that I used last time), there was 0.95V across that thing. I am getting to accept that it is a resistor. 1V in 33 ohms is 30mA! It was difficult seeing what I was doing at 24V because it was SSSOOO bright. As you said, they don't seem to last long at 24V all the time.

Charlie

Charlie

So if there is only two resistors and the LEDs were arranged like my drawing it may mean there's only 15mA per string of 6 LEDs - reasonably conservative but then they are tiny LEDs.  Not sure if the plastic goo helps or hinders dissipation.

Sorry David, I missed your drawing. Yes, I think that's almost right, but:

Each 5cm section has 24 leds and 2 resistors (assuming they are resistors). I measured some more of the resistors and they are all 33 ohms. Which must means four chains of six leds. These chains are paired. So two chains of 6//6. Each double chain is driven by one 33 ohm resistor, which, as you said, gives 15mA per led.

The strange layout may be more obvious if you try to design the printed circuit strip!

Charlie

Quote from: charliecoutas on Jun 22, 2025, 01:28 PMIt was difficult seeing what I was doing at 24V because it was SSSOOO bright.

LOL
Even if you go into theater to be operated on, the bright parts get covered by a rag, to prevent the staff from getting blinded that is.  Cover up...

But then on the other hand, if a girl has a bikini on, we as gentlemen only look at the covered parts.

The drawback of a completed product like the LED strip is you don't get the LED specification or data sheet.  This will tell you what the ratings are and in many cases the expected lifespan at different temperatures.

Led's as laser diodes, can be switched on hard and exceed their max rating, but then only for a short duration.  This is how many LED torch aka flashlight get more light output than when run at it's max current rating.

If you PWM the LED's it will increase it's lifespan, since it is for decoration, it should last quite long, hopefully the next generation technology will be available by then.
Perhaps something like holographics, replacing displays and TV's.
You will still see the LED's shining, but no wires and small components

Fanie

"But then on the other hand, if a girl has a bikini on, we as gentlemen only look at the covered parts."

(I don't know how to do quotes properly). Yes, we gentlemen pay great attention to the covered parts, and imagine what lies underneath! (It's sometimes a disappointment but often not.)

I have designed a PWM driving a power MOSFET to dim the leds. I was surprised to see a huge inductive kickback when the MOSFET turned off, but a reverse diode across the load fixed that. It runs at 10KHZ.

I haven't tried it on the full 24V yet as the Chinese power supply only just arrived. I have a crash helmet and fire extinguisher ready for the first switch-on. I'll send a photo of the coving downlights in operation, or the Chinese psu on fire.

Charlie

LED's in your circuit cannot have any inductive kickback.  It must be something else.
Also try different frequencies, some components switch better on/off at other frequencies.

A few meters of cable will put ears on your pwm waveform.  We just don't think of it like a lumped inductance.

The "load" was a 10 ohm power resistor not the LEDs. It is a gold coloured metal finned type, 100W rating. I think it was probably that which gave some inductance, although as David said, I have also seem inductive kick-backs with a few meters of wire. It's testing time tomorrow, I am at the museum today.

Charlie

That sounds like a wirewound type which would have noticeable inductance.

John

Agreed. Switched on this morning, all working very nicely. The PWM works a treat. It will look much better when the sun has gone down; I'll post a photo tomorrow.

Charlie

All done and very happy with the results. For the record: run them at just below their rated voltage (12 or 24) and use PWM to dim them. I used 8 bit, 10KHZ. Thanks for all the help and ideas. The conservatory is next....

Charlie

Very nice Charlie but it raises some interesting observations.
Incandescent lamps have considerable thermal inertia so they truly average a PWM waveform.  LEDs on the other hand respond very fast so the LED strip is actually turning on at full supply brightness but for variable duty cycles.  This would seem to conflict with persistence of vision theory which would say that even after the removal of the light the retina is still seeing the original brightness but other effects come into play.
From Wikipedia-

The Talbot-Plateau law is an experimental observation related to the psychophysics of vision. If a light flickers so rapidly that it appears as continuous, then its perceived brightness will be determined by the relative periods of light and darkness: the longer the darkness, the weaker the light.[2]

The law was first reported in a 1830 article by the Belgian scientist Joseph Plateau.[3] This article stimulated the English photography pioneer Henry Fox Talbot to publish, in 1834, his own observations on this topic made back in the 1820s.[4] While both scientists followed each other's experiments, they maintained that they conceived the original idea independently.[5] In 1863, the experiments of A. Fick suggested that the Talbot-Plateau law does not hold for strong light intensities. This suggestion was later proven by O. Grünbaum in 1898 who demonstrated that flickering strong light appears brighter than its steady state value.[1]

I think it's saying that below a certain light threshold the eye sees the average intensity but if above that threshold the eye sees the peak intensity.
Just enjoy your lounge lights Charlie.

Cheers,
David

Crikey, I hadn't thought about that. It implies that the first pulse should be full power, but subsequent pulses could be lower power (voltage). Then just before the eye/brain persistence period expires, do another full power pulse. This would need PWM and voltage control.

What a nice experiment for somebody to get into!

Charlie