Hobbyist Forums banner

1 - 17 of 17 Posts

·
Registered
Joined
·
1,368 Posts
Discussion Starter #1
I'm preparing to build a Round2 1:350 Refit and want to light it. I have a good grip on what to do for the blinking and strobe and dish lighting effects but general window lighting has a rather problematic aspect.

I'm planning on a lot of "always on" leds (50-100) white leds for windows, spot lights and backlights etc. I'll also have other colors for shuttle bay etc.

I understand that I can use a lower wattage power supply if the leds are wired in series. That has the advantage of lower resistor count and therefore lower heat dissipation which is a good thing within a pretty much sealed styrene kit.

The problem or downside to this approach though is that if any single LED burns out the entire string of steady on leds in the lighting system would go out and I be left with an unrepairable disaster since there would be no getting at the single dead led inside the sealed model.

So the questions are:
1. How have you advanced lighting guys/gals approached this situation? What methods do you guys/gals use for high led count projects?

2. Would it be better to plan to use parallel leds? In that case if I lose an led then effect wise it would only mean that "nobody was using that cabin"?

3. If I use parallel leds how do I deal with keeping the model cool enough to prevent sagging or distortion?

4. Do I use branches of series leds and hope for the best?

5. If I were to go with series circuits is there a good type/source of "high reliability" (i.e. longer life" leds you would recommend?

6. Are there any sources of schematics that would help with wiring this kit?

Thanks in advance for any help.

Regard,
MattL
 

·
Modeler's Brand
Joined
·
2,376 Posts
I'm certainly not advanced. However my plan to tackle this is go parallel. At most, I might do 2 or 3 leds in series on parallel lines, so if any go out, it's only a couple,a dn they are symmetrical deaths (port/stbd).

However if critical lights go out, like the bridge or the spotlights, then parallel is the only way for those. And of course, leds do simply die over time...

As to heat. I am going to run a test in the near future where i put afew dozen individually resistored leds into a small, closed envio with a thermometer. My bet is that any ambient heat, at worst, will be a few degrees over room temp over a 24hr period.

Consider: How long at a time will you really keep the lights on? 10 mins, an hour? Factored over a year, one or two days powered up?

That kit in particular is thick ABS plastic with a melting point of couple hundred degrees. I don't buy the henny-penny panic stories of kits melting.

As to nacelle or neck droop, you don't need leds for this to happen. That thick plastic is heavy. However, it is also very well engineered. Despite this engineering, I will be securing those angles with brass rod and/or strips. (Or one can spend a $100+ on the metal frame that is available for this model.) The extra weight of the lighting will cause droop before any typical amount of heat will, short of a hair dryer applied for several minutes to a given part. Long hot summer days will cause more warping than hundreds of leds for minutes to an hour at a time over the same period. (Of course, I have no doubt that powering up at noon for several hours in the middle of a week-long heat wave will warp the model!)

If one is still worried about heat, take a tip from your cpu and put heat sinks on the resistors. How? Standard kitchen tin foil folded into a spike or fan tips. Move the heat away from the interior surface and into the interior space.


Here's a several links confirming the strength of ABS.

http://en.wikipedia.org/wiki/Acrylonitrile_butadiene_styrene
http://www.dynalabcorp.com/technical_info_abs.asp
http://www.polymerweb.com/_datash/abs1.html
http://www.parmaplast.no/gml/uk/technical/PPmatr.htm#ABS
http://www.machinist-materials.com/comparison_table_for_plastics.htm

So generally, the melting point of abs is just under 200°F. That's hot enough to burn your skin at a touch. That being said, large amounts of medium heat over long periods of time can/will subtly warp plastic. In which case, attend to the super structure as mentioned and minimize your worries.

However, don't bet the farm on my incompetence!

Plenty of people have stuffed a 100+ leds in this kit with no horror stories to date. However-however, I would like to hear of any horror stories as none have been posted here or any other forums in several years now.
 

·
Registered
Joined
·
6,776 Posts
where's renegade when you need him???

I'm with Model Man on using a combination of parallel and series where appropriate. If your power supply is 9V you're not going to do more than 3 or 4 LEDs in series anyway, so it's kind of self defining. I wouldn't use more than a 12V supply just on general principles (you'll put your eye out!).

If you're designing something for a commercial display where it's always on I'd stick a cooling fan in it. Something ripped out of your PC would do nicely.
 

·
Registered
Joined
·
789 Posts
If you're designing something for a commercial display where it's always on I'd stick a cooling fan in it. Something ripped out of your PC would do nicely.
Ok, this got me wondering about LEDs and temperature when run for long time....so I did a little experiment.

I took a 60 led string of Christmas lights (bright white color) and put them in a glass dome off an outdoor light fixture, then plugged them in.

After 3 days of constant use I pulled out the trusty digital instant read lab thermometer and got the following results

Room temp at the time of reading: 70.8 degrees f.
Temp inside of the glass globe with LEDs: 91.8 degrees f.

So the LED strand raised the temp by 21 degrees.

To be fair the resister pod on the strand was in the globe as well and I'm assuming that is where most of the heat came from.
So if the resistors could be moved away from the leds and placed outside of a model then heat wouldn't be an issue???
 

·
Registered
Joined
·
6,776 Posts
Ok, this got me wondering about LEDs and temperature when run for long time....so I did a little experiment.

I took a 60 led string of Christmas lights (bright white color) and put them in a glass dome off an outdoor light fixture, then plugged them in.

After 3 days of constant use I pulled out the trusty digital instant read lab thermometer and got the following results

Room temp at the time of reading: 70.8 degrees f.
Temp inside of the glass globe with LEDs: 91.8 degrees f.

So the LED strand raised the temp by 21 degrees.

To be fair the resister pod on the strand was in the globe as well and I'm assuming that is where most of the heat came from.
So if the resistors could be moved away from the leds and placed outside of a model then heat wouldn't be an issue???
Heat is generated by the LED; not all the energy goes into creating light, and even if it did some would be converted back to heat, at least what doesn't radiate out of the model. I would guess the LEDs depending on the circuit design would be the source of most heat.

Most model displays are "momentary on." I wouldn't worry about heat in this case. If you are leaving it on constantly and it's an enterprise with loads of LEDs and some awful stress points I'd consider cooling it with a small DC fan (computer fans are quiet and typically 12V dc but should run on 9 even more quietly).

kudos for experimenting :thumbsup:
 

·
Registered
Joined
·
3,987 Posts
Regarding the heat, there are circuits that would switch off the lights when the environment reaches a certain temperature, no?
 

·
Modeler's Brand
Joined
·
2,376 Posts
Technically, we're not at the experimental stage yet.

Spockr posed a question. Based on minor experience, I posited an hypothesis. SWhite made a commendable observation, the results of which, on their own, are inapplicable. However, it's an interesting result that warrants further study.

Questions SW's observation do not answer:

:If run another 3 days, would the temp have risen by another 22 degrees? If only one day, would the temp have been 1/3rd? Was the increase incremental over 72 hours or did it peak early and stabilize? If so, when?

: What was the room temperature at the initial setup and was the time of day the same and done in the same location? A reading at noon versus midnight will inherently produce vastly different numbers, as well as being in direct sunlight/ perm shadow for any or all part of the duration. An initial temp of, say 78°, versus the 71° at completion is a major factor.

: Have the thermal properties of glass (created specifically for exterior use) versus injection molded plastic (of whatever standard variety) for model kits been factored out? Any form of glass dissipates heat at a radically different value than plastic and is not a suitable comparison on face value of one reading.

: The very physical structure itself of xmas lights versus hand wired leds will create different amounts of heat as well. And the physical amount of free space and air movement inside a glass sphere stuffed with xmas lights versus laying out individual leds in a disbursed structure such as any given model kit will alter the heat levels and dissipation as well.

------------------------------
Not trying to be difficult, but the scientifc method has hundreds of years of validity behind it. With multiple readings, further hypothesis can be made. With testing, we have a theory. If that theory holds up in all circumstances, we have a law.

For a 'proper' experiment I propose:

Multiple readings of ambient room temperature simultaneous to object temperature at least twice per day, but 3x would be better.

Out of direct sunlight and/or permanent shadow; and un-moved over the time frame.

Housed inside a styrene, or whatever variant of plastic desired, housing.

------------------------------
This will still not be fully accurate, but given the ability to graph data points into a data set will give us the best idea of what's really going on and how it might apply to models in general.

Also to note is that 90° is a common summer temp for any given place in the world.

As I said initially, I would not rec. running the lights for several hours at noon in the middle of week-long heatwave. A few minutes/ hours at a time though is not, generally speaking, going to harm the model, as SW's observation suggests...

22° over 72hrs = 1° every 3.72hrs! 180 leds = +1° every 1hr. In this light, these numbers are nothing to even bother considering and suggest, without further experimentation, that even the slightest concern over heat is baseless and unwarranted.

The most important question to ask is: how long a time are you really gonna run the lights for? 3 or more days straight through? Or minutes to a few hours at a time every few days or weeks?


And of course, you could get a thermal switch to shut the warp core down when the engines go critical. :thumbsup:
 

·
Premium Member
Joined
·
2,614 Posts
I think I need to point out that unless you are driving your LED's at a higher current then they are rated, which is bad for there lifespan, or you are using 1 watt or larger LEDs, Heat will not be an issue......:rolleyes: I've been using LEDs in my builds for more then 30 years and heat is the last thing I worry about. I tend to worry more about failure, so I age all the LEDs I use in my builds. I will put them in a proto board like this to burn them in for at least a week.
 

·
Modeler's Brand
Joined
·
2,376 Posts
A full week? Nice idea. I've been shooting for 30hrs based solely on the notion that computers have a 24 burn in time. A component, if it is going to fail, is most likely to fail immediately or a very, very long time away.

Of course, lights are most likely to fail on the on/off moment. Bulbs at least. I don't know about leds.

Hopefully the idea that heat is a true concern (except under the most exceptional of exceptional circumstances) can be put to bed.
 

·
Premium Member
Joined
·
2,614 Posts
A full week? Nice idea. I've been shooting for 30hrs based solely on the notion that computers have a 24 burn in time. A component, if it is going to fail, is most likely to fail immediately or a very, very long time away.

Of course, lights are most likely to fail on the on/off moment. Bulbs at least. I don't know about leds.

Hopefully the idea that heat is a true concern (except under the most exceptional of exceptional circumstances) can be put to bed.
You are right that most solid state components, if faulty, will fail in 24 to 48 hours, but when putting something in a build that might not come apart easly,
7 days works for me.......;) as far as incandescent lamps, they fail when they are turned on due to the fact that the filament vibrates violently due to the inrush of current heating the filament to white hot in a fraction of a second,
LEDs have no filament.:wave:
 

·
Modeler's Brand
Joined
·
2,376 Posts
Most excellent words, sir!
 

·
Registered
Joined
·
6,776 Posts
At the point it causes styrene to sag/distort.

Sorry, being facetious, but without scientific experimentation, the only thing we have to go on is intuition and the fact that heat is produced. Mine tells me that for seconds or minutes don't worry. Hours and days are an unknown, but if I were building a display that's permanently on I'd probably over-engineer and put a cooling fan in it. Of course that would be the first thing to fail, so I'd get a speed sensing fan and build a safety so if it stopped spinning the lights went out. Then of course all my electronics would be in modular bays for easy replacement at warp speed.
 

·
Registered
Joined
·
432 Posts
At the point it causes styrene to sag/distort.

Sorry, being facetious, but without scientific experimentation, the only thing we have to go on is intuition and the fact that heat is produced. Mine tells me that for seconds or minutes don't worry. Hours and days are an unknown, but if I were building a display that's permanently on I'd probably over-engineer and put a cooling fan in it. Of course that would be the first thing to fail, so I'd get a speed sensing fan and build a safety so if it stopped spinning the lights went out. Then of course all my electronics would be in modular bays for easy replacement at warp speed.
I have done alot of hard testing of electronic hardware in models over the years, Steve 244 has made a good point! Its all in the applacation of use. Randy
 

·
Registered
Joined
·
789 Posts
I can answer some of these...

Questions SW's observation do not answer:

:If run another 3 days, would the temp have risen by another 22 degrees? If only one day, would the temp have been 1/3rd? Was the increase incremental over 72 hours or did it peak early and stabilize? If so, when?
The temp did not rise as the days went on it stayed the same. The globe was warm on the 2nd day- not much to go on but better than nothing.

: What was the room temperature at the initial setup and was the time of day the same and done in the same location? A reading at noon versus midnight will inherently produce vastly different numbers, as well as being in direct sunlight/ perm shadow for any or all part of the duration. An initial temp of, say 78°, versus the 71° at completion is a major factor.
The room was a bathroom that didn't spent the day bathed in direct sunlight, the temp in that room holds around 70 degrees which only changes when someone takes a shower.

The globe sat in the corner away from drafts, and was turned so the opening was covered while the leds were on.

The reading was taken around noon. After my original post I turned the globe on its side and pulled the small resistor pack out, and let it run for another day with out seeing much change in the globes temp.

: Have the thermal properties of glass (created specifically for exterior use) versus injection molded plastic (of whatever standard variety) for model kits been factored out? Any form of glass dissipates heat at a radically different value than plastic and is not a suitable comparison on face value of one reading.
Ahhhh... No!

: The very physical structure itself of xmas lights versus hand wired leds will create different amounts of heat as well. And the physical amount of free space and air movement inside a glass sphere stuffed with xmas lights versus laying out individual leds in a disbursed structure such as any given model kit will alter the heat levels and dissipation as well.
True! But it answered the question I had which is do leds generate heat. The answer is yes.

Given that no one here is sticking a 110v 60 led strand into a model it's safe to say as you pointed out that most all models will survive what little heat an led generates.

If the heat problem is that critical then steps can be taken to lessen its impact such as vents, or moving the resistors outside of the model if possible.

Any other experiments will have to wait as my leds and globe have been confiscated by the powers that be in the house to be used as a night lite!
Thus replacing the evil penguin with glowing red eyes stained glass lamp that held the job for 2 years.:(
 

·
Premium Member
Joined
·
1,220 Posts
Well for me it will always be paralell mainly because I am no electronic wizard. My 1/350 'E' has about 46 leds not counting the strobes. and it runs off a 9v wall wart. I have actually left it on for two days once over a w/end as we went away. Got home on the sunday night and had to think what the gow was from the study...:rolleyes: The only part that had any noticable warmth was the area of the saucer where I had Randys nava board mounted. Except for space issues I will usually have the resistor on the + leg of the led then shrink tube it for insulation and ease of glueing. I did put a set of series in my Chariot but the second time I turned it on to show a mate only one side of the dash was working :cry: So parralell and good wire managment are the only way I'd suggest.

Cheers,

Alec
 

·
Registered
Joined
·
432 Posts
Well for me it will always be paralell mainly because I am no electronic wizard. My 1/350 'E' has about 46 leds not counting the strobes. and it runs off a 9v wall wart. I have actually left it on for two days once over a w/end as we went away. Got home on the sunday night and had to think what the gow was from the study...:rolleyes: The only part that had any noticable warmth was the area of the saucer where I had Randys nava board mounted. Except for space issues I will usually have the resistor on the + leg of the led then shrink tube it for insulation and ease of glueing. I did put a set of series in my Chariot but the second time I turned it on to show a mate only one side of the dash was working :cry: So parralell and good wire managment are the only way I'd suggest.

Cheers,

Alec
Glad the Nav board is still going strong, nice test job Alec, I mean it in a good way mate! Randy
 
1 - 17 of 17 Posts
Top