4
\$\begingroup\$

I have seen similar questions to this one but none of them were asked in the range of seconds only minutes.

I just finished building a incubator for chicken eggs and I am using an incandescent light bulb as source of heat. The thing is I used a pretty sensible digital termometer an the light bulb turns on, last ~20 seconds on and then turns off again. I'm worried this will cause any damage to the light bulb or the relay that acts as a switch.

The incubator works well but I dont want to replace parts every week.

\$\endgroup\$
  • \$\begingroup\$ See If I flick the light switch on and off will it damage the light?. Will it not bother the chickens? \$\endgroup\$ – Transistor Feb 6 '16 at 22:43
  • 1
    \$\begingroup\$ Why not use a much higher duty cycle, so the lamp doesn't have time to cool down? Ideally, you would combine PWM with a preheat (so the lamp is never actually off). \$\endgroup\$ – uint128_t Feb 6 '16 at 22:52
  • \$\begingroup\$ Incandescents have a larger tolerance to intermittent use than other bulb types (led and CFL). You can use a snubber or other soft start to extend it's life as well as the relay \$\endgroup\$ – crasic Feb 6 '16 at 23:00
  • 1
    \$\begingroup\$ @crasic while led & cfl bulbs may have electronics that can wear prematurely from repeated switching, the light-emitting elements themselves ('bulbs'), at least in the case of led lamps, are far more tolerant of the stress. "Cold cathods" fluorescents will be highly tolerant too, but fluorescents with preheat contain incandescent cathodes that can damage the same as any 'regular' incadescent bulb. \$\endgroup\$ – Robherc KV5ROB Feb 7 '16 at 0:15
  • \$\begingroup\$ @transistor The chicks are (supposed to be) removed from the incubator as soon as (someone notices that) they've hatched. The way I read the question, it looks like they're using an out-of-coop incubator with the eggs in it (probably a hobby-farm sized thing; larger operations often devote entire rooms to incubate dozens of racks @ hundreds of eggs each - disgusting). With that said, I guess a legitimate argument could be made that we don't know when chick embryos first sense light, so that could bother them. ;) \$\endgroup\$ – Robherc KV5ROB Feb 7 '16 at 6:04
2
\$\begingroup\$

I'm not sure this is so mich an "original" answer as a combination/explanation/redinement of the other 2 already posted answers, but I'll try to help a bit.

  1. As transistor pointed out in the question comments, the 'bulb'-life impact of rapid switching on several varieties of lights is discussed in some detail in If I flick the light switch on and off will it damage the light?

  2. Autistic, Rothloup, and 2 posters in the other thread are all right about 'dimming' the incandescent greatly improving bulb life.

    • I would recommend using an inductor in SMPS buck-converter arrangement as likely being the most efficient (be careful though, the efficiency of that one inductor could surprise you, incandescents have a very non-linear response & tend to respond well to under-voltage drive).
    • While the buck converter could arguably be run on from non-rectified ac current, I would personally recommend using rectified & conditioned/smoothed DC (for predictable performance), and a pwm management that never cycles slower that 40KHz (so you don't have to hear it 'sing').
  3. You should definitely replace the relay with solid-state components.

    • At the least, a triac or optoisolator/'solid state relay' can switch your load faster than a mec. relay & love longer.
    • Better by far would be using a bridge rectifier & a capacitor to convert your input to smoothed/conditioned DC, then use a power MOSFET (if you get a 'logic-level drive' one, it won't even need any external buffering transistors) to run the aforementioned buck converter.
  4. Putting it all together, if you PWM a buck converter, with solid-state switchgear, to keep the light somewhere between 'just warm' and 'hot, but barely glowing,' you should be able to maximize your energy efficiency, component lifetime, and the precision/stability of your incubator's temperature control.

\$\endgroup\$
1
\$\begingroup\$

If the issue is that your thermometer is too sensitive, then it sounds like you need to add some hysteresis to the switch point.

If you are running the light off of a programmable microcontroller like an arduino or something, you can read the temperature and turn the light on a few degrees below your set point, and off a few degrees above.

You can also use pwm (maybe with an in line inductor to smooth the current) to turn it on at less than full power.

Combining these techniques should allow you to control your temperature, on time, and duty cycle as you wish.

\$\endgroup\$
  • \$\begingroup\$ @SeanBoddy comment moved to answer \$\endgroup\$ – Robherc KV5ROB Feb 7 '16 at 6:47
1
\$\begingroup\$

The life of the Bulb is shortened by repeated turning on and off.
The relay life is also shortened by the inrush current of a cold filament.

You must keep the repeat rate low because the relay has a finite number of operations before mechanical failure.

  • If you did something with a light dimmer lamp life would be very good and there would be no electromechanical relay to worry about.
  • If you used a solid state relay, a high repeat frequency could be chosen so the bulb doesnt get stone cold.

This is all hindsight so here are some things that can be done easily on what you have already built:

  • Use a ntc (Negative Temperature Coefficient) thermister to limit inrush current helping bulb and relay life
  • Wire larger bulbs in series which reduces inrush and will be fine because its the heat you want.
  • Preheat the lamps by allowing a small current to pass through the relay conntacts.
    (Preheating is nothing new, it has been done for years on car indicator lamps.)

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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