2
\$\begingroup\$

I have this UBERHAUS floor lamp, with a single-filament, two-terminal bulb. They have a rotary switch in the tube (stem) that turns clockwise to three positions: off -> dim -> bright -> off.

My question is, how does a rotary switch (which must be small because the tube is only 1" in diameter) dim a 2-terminal bulb? I tried searching the internet for answers, but I don't even know what you'd call one of these switches. Also, would this lamp damage a non-dimmable bulb?

UBERHAUS floor lamp

\$\endgroup\$
1
  • 1
    \$\begingroup\$ There must be a dimmer. Therefore un-dimmable bulbs must not be used, as they won't work with a dimmer, there might be damage to the un-dimmable bulb or to the dimmer. \$\endgroup\$
    – Justme
    Commented Sep 13, 2020 at 8:05

2 Answers 2

3
\$\begingroup\$

A dimmer module can be very small. They're efficient in that they don't waste much heat and so could be concealed in the lampstand quite easily.

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 1. (a) A crude diode dimmer. (b) A fixed phase-angle dimmer.

Figure 1a uses a diode to allow only positive half-cycles of the mains through the bulb. This reduces the voltage to half but would not be a recommended solution as it will probably cause flicker and draws DC from the mains which the utility companies don't like as if there is enough of that going on there would be a DC bias on their transformers.

Figure 1b shows a TRIAC dimmer with a fixed trigger point delay. On each half-cycle C1 is charged up by R1. When the voltage gets high enough to cause DI1, a DIAC, to break down (typically around 20 V) the TRIAC will turn on.

enter image description here

Figure 2. The upper trace shows the trigger delayed close to the end of the cycle. The resultant effective voltage is low. The lower trace shows the trigger close to the start of the cycle. This will result in close to full voltage. The relationship between phase angle delay and resultant RMS voltage is graphed on the right. Image source: Dimmers for LEDs.

My linked article has a little more on the topic.

\$\endgroup\$
2
  • \$\begingroup\$ Is there much chance that, for the crude diode dimmer circuit, the diode direction would be random and so there would be, averaged out over many households, no DC offset? \$\endgroup\$ Commented Sep 13, 2020 at 18:21
  • 1
    \$\begingroup\$ Sure. I considered that as I wrote my answer. In the lands of unpolarised plugs that would be automatic. In Ireland we use UK style plugs which are polarised so you'd be relying on the manufacturers to randomise the diode direction. I think the diode solution is unlikely due to flicker - only 50 or 60 pulses per second instead of 100 or 120. \$\endgroup\$
    – Transistor
    Commented Sep 13, 2020 at 18:25
1
\$\begingroup\$

Transistor has the answer; here's some background info and one solution.

That's a common question we get from novice electronics guys -- they think about how residential AC light dimming works, and figure

  • It can't be an inline resistance; that would make a lot of heat.
  • It can't be a variac; since those are huge.

Transistor (above) describes the triac and diode dimming.

Now, you say "non-dimmable bulb" -- the photo you have is of an incandescent bulb, which is nothing but a resistor with some weird properties, not least, light emitting lol). So dimming is inherent in all incandescent bulbs, and triac dimming is specifically designed to be a cheap solution to dim incandescents. It will play nice with diode dimming too.

My point: All this goes out the window when you start dealing with alternate lighting tech such as compact fluorescent or LED. Because heat is a concern, they use a switching power supply to make the correct current for the series string(s) of LED emitters. That is typically multi-voltage (like 90-264V, world voltages +/-10%) and dynamically adjusts to input voltage so fast that it can ride through triac dimming. It treats dimming like dirty power lol!

Dimming fluorescent is ...complicated... and triac dimming just gets in the way. LEDs can dim either by reducing current through the emitters or by PWM. Triac dimming doesn't do that, so the "dimmable" LED/CFLs have a circuit to reverse-engineer what the triac dimmer is trying to do, and then do that. Obviously that is extra complexity/cost.

I expect the "simple diode" dimming method is not one that these modern bulbs know how to emulate. Your best bet might be to fill the bowl with LED Christmas lights; those are typically wired in 2 series strings of LEDs, with each string facing the opposite direction, so each string lights up half the cycle. In your lamp, at the "dim" setting, one string would light.

\$\endgroup\$
2
  • \$\begingroup\$ Thank you! Even though I selected Transistor as the answer, your LED Christmas light trick also helped me to understand. \$\endgroup\$
    – AshleyZ
    Commented Sep 13, 2020 at 22:40
  • \$\begingroup\$ You well should select Transistor's answer; it's right on point and excellent, as always :) \$\endgroup\$ Commented Sep 14, 2020 at 1:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.