What is your experience RE: LED bulbs, relays and dimmers; and heat?

Discussion in 'Darkroom Equipment' started by David Brown, Jan 19, 2013.

  1. David Brown

    David Brown Subscriber

    Messages:
    3,597
    Joined:
    Feb 16, 2004
    Location:
    DFW, Texas
    Shooter:
    Multi Format
    This is a little esoteric, but I don't know where else to start looking, and it is for darkroom consumption.

    I'm hadn't been happy with the general work lights (white lights) in my darkroom, and am now on the 3rd iteration, which is two bare 150 watt light bulbs in porcelain ceiling fixtures. The light is now fine, but that's 300 watts and a surprising amount of heat! Since fluorescent is out (we've discussed this to death - suffice it to say I've seen the glow ...) I've thought about substituting LED bulbs, which would reduce the wattage consumption. However, LEDs with any comparison is actual light output to high wattage tungsten have heat sinks! Do they also put out heat? If they put out no more heat, or less, than the equivalent tungsten, then the point is moot. However, tungsten bulbs don't come with built in heat sinks! :blink:

    I've also considered replacing the tungsten bulbs in the safelights with LEDs, but am concerned about the heat AND the warnings with many LED bulbs to not use them with relays or dimmers. OK, dimmers is not an issue, but what about the relay circuits in enlarger timers? What is going to be damaged: the LED bulb, or the timer?

    OK, let's stay on topic. i.e., do LEDs produce heat like tungsten, and do they damage relay circuits? Does it make a difference if the relays are mechanical or electronic?

    Thanks!
     
  2. Chan Tran

    Chan Tran Member

    Messages:
    2,374
    Joined:
    May 10, 2006
    Location:
    Aurora, IL
    Shooter:
    35mm
    We can both do more research but as far as I know the larger wattage LED lights do generate significant heat but not as much as incandescant for the same light output.
    I can't see why they could damage the relay.
    For dimming you would need different dimming circuit I think.
     
  3. Lukas_87

    Lukas_87 Member

    Messages:
    114
    Joined:
    Jan 15, 2008
    Location:
    Prague, Czec
    Shooter:
    35mm
    there is a big difference in the light output per watt power input - typical light output efficacy for halogen tungsten bulb is about 17 lm/W for 100 W bulb and typical light output efficacy for LEDs is between 50-100 lm/W. that means if you need the same luminous flux you will generate less heat to sink with LEDs because they are more efficient.
    it's kind of a misunderstanding that bulb doesn't have heatsink since the bulb itself is a heatsink. however the LEDs need to be cooler (you will damage the devices if you let them to heat above 100 °C) so they need bigger heatsink to spread the heat more efficiently because there's lot lower tempreature gradient between the LED and room temperature. still they generally generate less heat.

    it's a good idea to use a red LEDs inside yout safelight.
     
  4. hgernhardt

    hgernhardt Member

    Messages:
    68
    Joined:
    Dec 24, 2012
    Location:
    Brockton, MA
    Shooter:
    Medium Format
    LEDs do produce heat, but not nearly as much as a related incandescent. The LED is a semiconductor which may be prone to a phenomenon called thermal runaway. The movement of electrons (and holes) through the PN junction of the diode results in heat. This heat increases the conductive efficiency of the junction. As a result, unshed heat can result in a cascade reaction which eventually destroys the junction.

    High-power semiconductor devices shed a lot of heat. Take a gander at the heat sink in your desktop computer. Look then at the size of the package in which the processor is contained. Consider that the processor itself is about a centimeter square within that packaging. LEDs are no different in this regard. The purpose of the heat sink is to cool the junction. I've touched the heat sink of an operating LED bulb, and found it to be warm to the touch. The wattage rating (real, not effective tungsten output) should give you an idea of how much heat the device will shed into the atmosphere.

    We don't use standard dimmers with some bulbs because of the nature of the power supply circuitry—it probably requires a certain line voltage in order to properly function. Low voltage conditions (i.e. standard wall dimmers) may not be appropriately handled by the power supplies. I would imagine that, for much the same reason, the voltage transient which occurs when relays switch may cause difficulty for some units.

    The worst case scenario is that the diode itself would be destroyed. Whatever the case, though, you should be able to find LED units which can handle dimmers and relays. As with anything else, you get what you pay for.

    Good luck!
     
  5. ic-racer

    ic-racer Member

    Messages:
    7,473
    Joined:
    Feb 25, 2007
    Location:
    Midwest USA
    Shooter:
    Multi Format
    No electronics to fry in an incandescent bulb. I'm surprised the LEDS have a heat sink. Maybe it is just for show as I'm sure the electronics are designed to fail in a year or so :smile:
    Personally, I'm stocking up on incandescent bulbs. Of course LEDS give off heat but the efficiency is superior to incandescent.
     
  6. Jim Rice

    Jim Rice Member

    Messages:
    227
    Joined:
    Jun 7, 2006
    Location:
    Jackson. MS,
    Shooter:
    Multi Format
    LEDs will dissipate as heat the fixed voltage drop (0.7 volt typically) via Ohm's law, less the energy produced as light. As a starting point, it is safe to assume that the entire load will dissipate as heat and add cooling as required.
     
  7. Jim Rice

    Jim Rice Member

    Messages:
    227
    Joined:
    Jun 7, 2006
    Location:
    Jackson. MS,
    Shooter:
    Multi Format
    And I would imagine that the biasing resistor is doing most of the heating.
     
  8. ME Super

    ME Super Member

    Messages:
    1,225
    Joined:
    Apr 17, 2011
    Location:
    Central Illinois, USA
    Shooter:
    35mm
    More than you probably wanted to know about biasing LEDs

    It's been a long time since I did any work with LEDs (back in the '90s when I was an Electrical Engineering student), but typically you have a power supply of X volts (let's say 6V for sake of argument). A typical red LED has a forward voltage drop of 1.7V. Typically you choose a current value of 50% of the LED's rated current (Imax). We'll say for sake of argument that this is 20mA.

    The voltage drop that will occur across the resistor is 6V - 1.7V, or 4.3V (power supply voltage - forward voltage drop of the LED). Ohm's law for resistors says R = V/I. R = 4.3 V / 20mA, which works out to 215 ohms. Pick the closest resistor you have and wire it in series with the LED.

    Since P=IE (also Ohm's law), the power dissipated by the LED (including light emitted) is 1.7V * 20mA = 0.03W. The power dissipated by the resistor is 0.086W. Note that in the case of an AC powered LED, there are other electronics involved as well (such as a transformer and rectifier circuit) which will likely produce more heat than the LED and resistor itself.
     
  9. DWThomas

    DWThomas Subscriber

    Messages:
    2,899
    Joined:
    Jun 13, 2006
    Location:
    SE Pennsylvania
    Shooter:
    Multi Format
    I haven't gotten into them yet, but it is quite possible that LED lamps use some sort of switching power supply which can be far more efficient than a dropping resistor. One of the reasons dimmers can be problematic is that some chop up the normal AC sine wave into weird chunks that play havoc with the internals of electronic gizmos down the line. A friend claims he got some CFLs that will operate on a dimmer -- but they were $20 apiece vs the usual buck or two.

    I wouldn't expect problems from relays; they are just a switch that uses a magnet instead of a finger to flip the contacts.
     
  10. polyglot

    polyglot Member

    Messages:
    3,472
    Joined:
    Jun 12, 2009
    Location:
    South Austra
    Shooter:
    Medium Format
    LEDs for illumination do not use series resistors, they use switching regulators, which are about 85% efficient. The power is being dissipated in the LED, and you should expect about 3 to 5x less heat to be emitted for the same level of light compared to tungsten. Tungsten bulbs run at a couple thousand degrees though, so they have no trouble dissipating the heat. LEDs must stay under 100C or the doping in the silicon will migrate and it will cease to be an LED.

    Thermal power flow behaves a lot like current flow: the thermal power flow (P) is proportional to the temp difference (dT) and inversely proportional to the thermal resistance (theta). P = dT / theta

    dT(tungsten) = 3600 - 280 = 3320K
    P(tungsten) = 150W
    theta = 22 K/W

    dT(LED) = 90 - 30 = 60K
    P(LED) = 37.5
    theta = 1.6 K/W

    (A heatsink is a low-resistance thermal resistor. The larger surface area reduces the thermal resistance between the solid half and the air flowing past)

    That tells you that for a 37W LED to maintain its junction temperature at 90C in a 30C room (typical design point), the junction-to-ambient thermal resistance must be no more than 1.6 kelvin/watt. That's a non-trivial heatsink.

    The 150W tungsten bulb on the other hand will have a (designed) thermal resistance of 22 kelvin/watt so that it can reach the proper (3600K) temperature. The bulb size and thickness and gas mix are chosen in order to achieve that thermal resistance and therefore proper operating temperature.

    For your purposes, the 150W vs 37W comparison is what matters, it means about 4x less heat and power consumption for the same light.

    You can't use an LED on a dimmer unless it is specifically designed for that use and they will often not light at all (or just flicker very dimly) when switched by a solid state relay (triac), but they'll be completely fine with electromechanical relays. Yes, the problem is because they use a buck-mode switching regulator which has discontinuous input current, which means that if you run them from a triac, the triac switches off and you get no light.

    If you buy LEDs, make sure they have high CRI (colour quality) and a colour temp of either 5400K (daylight) or 3600K (tungsten orange). Do not buy anything with a 6000K or higher colour-temp; it will look disgustingly blue. If you print colour, the CRI of your light sources should be greater than 90 or you will have serious problems.
     
    Last edited by a moderator: Jan 20, 2013
  11. nicholai

    nicholai Member

    Messages:
    294
    Joined:
    Jul 12, 2012
    Location:
    Kolding, Den
    Shooter:
    Multi Format
    LED produce much more light with way less energy, and my experience tells i can easily touch them while they have been running for hours.
     
  12. Steve Smith

    Steve Smith Subscriber

    Messages:
    9,083
    Joined:
    May 3, 2006
    Location:
    Ryde, Isle o
    Shooter:
    Medium Format
    The company I work for makes some of the flat, flexible light panels shown on this company's website:

    http://designledproducts.com/

    One day I intend to try out something similar in my 5x4 enlarger.

    And they don't require heatsinks.


    Steve.
     
  13. bdial

    bdial Subscriber

    Messages:
    5,413
    Joined:
    Jan 2, 2005
    Location:
    NE U.S.
    Shooter:
    Multi Format
    I use LED's in my home office, the heat from them is significantly less than the tungsten bulbs they replaced, it's very noticeable in the summertime.
    The oldest one is about 4 years old now, and they are generally on for 15 - 18 hours/day * 7.

    They do have a slight afterglow, for maybe 15 seconds after the power is off.
     
  14. David Brown

    David Brown Subscriber

    Messages:
    3,597
    Joined:
    Feb 16, 2004
    Location:
    DFW, Texas
    Shooter:
    Multi Format
    Yes, if I understand correctly.

    Maybe, see below:

    Thanks, everyone. This has been interesting and informative. Allow me to narrow the follow-up question:

    Regarding the relays: I doubt much of anything would be damaged in an old TIme-O-Lite enlarger timer. It's basically the same as a light switch. However, what I would be concerned about is my RH Designs Stop Clock (APUG sponsor alert). I don't want to fry its electronics with the wrong type of bulb, if that's possible. I am going to email the company and ask specifically, but it would be interesting to hear actual field experience, if there is any.
     
  15. Jim Rice

    Jim Rice Member

    Messages:
    227
    Joined:
    Jun 7, 2006
    Location:
    Jackson. MS,
    Shooter:
    Multi Format
    If the LEDs are within the current rating of the timer it will be fine.
     
  16. polyglot

    polyglot Member

    Messages:
    3,472
    Joined:
    Jun 12, 2009
    Location:
    South Austra
    Shooter:
    Medium Format
    Does the Stop Clock make a clicking noise (could be extremely quiet) when it turns on and off or is it completely silent? If it clicks, it has a mechanical relay and it will be fine. If it's silent, it's probably a triac.

    While a triac can in theory be damaged by attaching an LED switching supply to it, that's highly unlikely and I would not expect RHD to cheap out on the protection circuitry. Surviving a globe-blow event on an enlarger with transformer is far more stressful to a triac than an LED supply. More likely though is that the LED just doesn't light, not that any damage occurs.

    If you have the ultimate in paranoia then you can buy an electromechanical mains relay, i.e. one with a 240V (or 115V) coil and mechanical contacts, e.g. ebay 390510321612 or similar. Let the timer switch the relay on, which will switch your load on - that will work, guaranteed, and it will not damage the timer or the LEDs. I don't reckon it's necessary, but wait to hear back from RHD before taking my advice on their product that I actually know nothing of the specific internals :wink:
     
  17. David Brown

    David Brown Subscriber

    Messages:
    3,597
    Joined:
    Feb 16, 2004
    Location:
    DFW, Texas
    Shooter:
    Multi Format
    Waiting patiently ... :cool:
     
  18. polyglot

    polyglot Member

    Messages:
    3,472
    Joined:
    Jun 12, 2009
    Location:
    South Austra
    Shooter:
    Medium Format
    Does it make a little click or tink noise?
     
  19. Steve Smith

    Steve Smith Subscriber

    Messages:
    9,083
    Joined:
    May 3, 2006
    Location:
    Ryde, Isle o
    Shooter:
    Medium Format
    I'm fairly sure it's a relay. In a thread on another forum, I raised the issue of using a solid state relay module for a timer and Richard of RH Designs stated that he uses mechanical relays.

    It was a few years ago and I don't know if that is the case with all of his products but I suspect that it is.


    Steve.
     
  20. polyglot

    polyglot Member

    Messages:
    3,472
    Joined:
    Jun 12, 2009
    Location:
    South Austra
    Shooter:
    Medium Format
    Presuming that to be true, there will be zero issues using a Stop Clock with an LED enlarger and/or safelights.
     
  21. David Brown

    David Brown Subscriber

    Messages:
    3,597
    Joined:
    Feb 16, 2004
    Location:
    DFW, Texas
    Shooter:
    Multi Format
    Mr. Ross has likewise replied to me with essentially the same information; i.e., electromechanical relays, not solid state ones. He assures me it is no different than turning a light switch on and off.
     
  22. RH Designs

    RH Designs Advertiser Advertiser

    Messages:
    657
    Joined:
    Apr 9, 2003
    Location:
    Yorkshire Da
    Shooter:
    Multi Format
    Just to add that the very first batch of Analyser Pro/StopClock Professional did use triacs but we reverted to electromechanical relays as some enlargers didn't like the triacs very much - specifically the LPL 5x4 which has its own relay. The leakage current through the triacs was sufficient to keep the LPL's relay closed when switched off - obviously it was a very sensitive relay! If you can hear a "click" when the timer switches the enlarger on or off, then it's NOT a triac model.
     
  23. David Brown

    David Brown Subscriber

    Messages:
    3,597
    Joined:
    Feb 16, 2004
    Location:
    DFW, Texas
    Shooter:
    Multi Format
    I was going to post Richard Ross' email back to me with his permission (which he granted), but his post above is even better!

    Thanks to all that contributed. All questions answered. :cool: