I ask this question because I just installed a new Tomar Industries double search light using the identical resistors on each of the four red and green wires that I used on the first Tomar Industries double search light some 9 years ago.
This newest signal is considerably brighter than the ones that I installed back in 2004.
So, do LEDs dim with age?
Rich
Alton Junction
Not that know of, Rich. Most likely it's better LED technology and manufacturing.
Did you double-check the actual resistance of the resistors before installing them? Always a good habit to get into whenever installing resistors.
Tom
https://tstage9.wixsite.com/nyc-modeling
Time...It marches on...without ever turning around to see if anyone is even keeping in step.
tstage Did you double-check the actual resistance of the resistors before installing them?
Did you double-check the actual resistance of the resistors before installing them?
Yep, I have been using the same resistors, with the same values, for years. I simply follow the instructions and the recommendations made by Tomar, the manufacturer.
They kind of do, but by an 'old' LED, we're talking a lot more than just a couple of years old. Such as, compare two early LEDs, from when they first became commonly available over 40 years ago - one that has been in continous use and one new in the pack (hmm, wonder what a NIB never used 1971 LED would go for...) you may be able to detect a difference. And you'd need a large sample size, not just 2, to be sure it wasn;t manufacturing differences, unless the two in question were from the exact same batch.
Which is what you are probably seeing - simply normal manufacturing variation from batch to batch. At the proce of today's LEDs, I doubt they are doing any super tight controls on process variation, at least not on the ones they sell for pennies a piece.
And then there's the dropping resistor being used - if you are using a common resistor, they are 10% tolerance. It is entirely possible that, with a pair of "1K" resistors, to have them 200 ohms different between each other - one the full 10% high, and one the full 10% low. It's VERY likely that there is a 100 ohm difference - one 5% high and one 5% low - because if the resistor was closer then 5% off of the nominal 1K value it would have been binned and sold as a more expensive 5% tolerance part. Though a 100K difference may not result in a visible difference int he brightness of the LED. But multiply that up if the resistor is bigger than a 1K - say the recommend and you uses a 2.2K, that could be a 440 ohm difference in values and very likely to be enough current difference to change the visible brightness.
Going back to the early LEDs, I remember wanting to experiment and not being able to afford many LEDs, they were well over $1 each. There was a circuit for an LED array oscilloscope in one of my electronic magazines - the LED array alone would have cost over $200 to build. At today's LED prices - maybe $15. I did eventually build one as my senior project in college since we didn;t have to pay for the components - by that time LEDs were well off the initial high proces, but there was probably about $50 or so in just the display. Hmm, I think I still have it around somewhere, too - I guess if I need some red LEDs I can start taking that apart.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
OK, got another question.
I grabbed another resistor and placed it in line behind the first resistor and the LED grew a little dimmer.
I don't want to simply replace the existing resistor with one of higher value because everything is already soldered and heat shrunk in place.
But, I could add another resistor in line behind the existing resistor. What is the resulting resistance?
Say, for example, that I have a 1.5K ohm, 1/2 watt resistor already in place. Then, I add another 1.5K ohm, 1/2 watt resistor next in line. Is that a 3.0K resistor value, or something less?
Or, I could just do some trial and error to match the brightness to that of other close by signals.
While there is probably some reduction on light output over time, LEDs are solid state devices and will eventually just fail. Unlike a lightbulb that has decreased output because tungsten vapour condenses on the glass envelope during operation, reducing light output.
What you are probably seeing is a combination of advances in LED fabrication and maybe a different supplier. Different suppliers will have subtle differences, even between batches. You may also be seeing the effect of different resistors. Again, different suppliers have variations between them. Or the tolerance of the resistors is different, maybe 5% versus 10%. They may sell them as 10% when in fact, they failed qualification as a 5% device by being 6% out. Resistors can shift in value over time as well, usually increasing in resistance.
Get a 5k pot from Radio Shack. Set it to max resistance. Put the pot in place of the original resistor. Slowly turn the pot shaft until you have the same brightness. Remove the pot and use your ohm meter to measure the resistance of the pot. Put in the nearest value. I have done that a few times. Not a big deal.
If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.
Yes, resistors in series just add, so a 1K plus a 470 ouhm in series gives you 1470 ohms total. In this case - it IS that simple.
Rich,
I agree it's an issue most likely caused by variance in resistor values. 10% ones are pretty notorious for catching folks up building circuits that need to be a little more precise to work well. 5% and 1% ones can get expensive, but do save this being an issue.
This resistor calculator may come in handy dealing with adding resistance in series or parallel to correct:
http://www.quickar.com/bestledcalc.php
Mike Lehman
Urbana, IL
I have used carbon film resistors for many years which are 5 percent.
http://en.wikipedia.org/wiki/Resistor
richg1998 I have used carbon film resistors for many years which are 5 percent. http://en.wikipedia.org/wiki/Resistor Rich
That's what I use as well.
I have a fairly large selection of 5-packs from Radio Shack on hand. I have had these resistors for a while now, but the Tomar Industries Double Search Light that I installed today is brand new, just purchased. So, I suspect that it is the LED, not the resistor.
I have a whole bunch of Tomar single and double search lights on my layout. What I have observed is that the older the search light, the dimmer (or maybe I should say less bright) the LED. I have always used resistors with the same values from the same source (carbon film resistors from Radio Shack).
If it's a case of using the LEDs as supplied in a kit, there can be a big difference in light output solely because of variation in batches. I'm sure Tomar buys in large quantities, but equally sure there's variations in batches.
It could also be the case that improved chemistry, substrates, etc means that what you're see may be brighter LEDs in production, even if the spec is technically the same. It may not be so much any dimming of old LEDs, but that the new ones are general brighter.
These Tomar Industries search lights are pre-built, assembled and wired, and ready to install right out of the box.
Tomar does not supply the resistors, but it does recommend 1.5K for the red and 680K for the green. I have faithfully followed this recommended value with each resistor that I have installed for these bi-polar LEDs.
My power supply is the DC side of an MRC 1370 Railpower transformer. It has an 18 volt maximum, and I have the rotary dial turned all the way up. Yesterday, I experimented by dialing it down a little, and that dimmed the new search light LEDs, but, of course, every other LED connected to the power supply dimmed as well.
I do like Richg1998's suggestion of using a pot to determine the ideal resistor value, and I will do that today. I will report back with the results.
richhotrain tstage Did you double-check the actual resistance of the resistors before installing them? Yep, I have been using the same resistors, with the same values, for years. I simply follow the instructions and the recommendations made by Tomar, the manufacturer. Rich
Actually, my answer to Tom's question was not entirely accurate.
When I said, Yes, I check the actual resistance before installing a resistor, what I meant was that I selected the resistor with the resistance value stated on the package.
After I re-read this thread, especially Randy's reply, I put my ohm meter on several different resistors.
I make sure that the ohms aadjustment properly set to zero and I installed a new battery before testing.
I am finding variances of 25 to 30 percent, unless I am doing something terribly wrong.
I measure resistance on a 330 ohm resistor and get a value of 250 ohms.
I measure resistance on a 560 ohm resistor and get a value of 440 ohms.
To assure myself that my multi meter is working properly, I tested the AC volts and the DC volts on my power pack and I get exactly accurate output voltages.
If I am doing this correctly, it is no wonder that the LEDs on the new signal are so bright.
Can resistors deteriorate sitting in a plastic box for years?
Not likely, but one thing to keep in mind is calibration of the meter. Just because the volt function may be dead on does not mean the resistence calibration is correct. Also, the resistence function relies on the internal battery, if the voltage is slightly low due to it being an older battery, this will throw off the results, even though there is plenty of power to drive the display. Same applies to an analog meter, plus you have to be wary of parallax when viewing the pointer against the scale.
Most meters will list int he specs somewhere the accuracy of each range. So you worst case for a given resistor reading would be (assuming a 10% resistor): nominal value - 10% of nominal - meter tolerance
So if it's listed as being accurate within 10 ohms on the 1K scale, a 330 ohm resistor could read as low as 290 ohms, assuing everythign else is operating within parameters, such as the battery. And then there's the contact between the probes and the resistor leads, and the probes and the sockets on the meter. When you start getting into lower resistor values, a slight difference in the pressure applied between the probes and resistor leads starts to become a factor. Stick something bigge ron, like a 10K resistor and see what you get.
I use 4.7K SMT resistors to make detection wheelsets. They are something better than 10%, they might even be 1%, I'd have to go look on the box. With the silver conductive paint and a clean wheelset, my fancy but older desktop Fluke (which has not been calibrated in over 15 years) reads about 5.16K. I think it's pretty accurate. i don;t have a small enough set of probes to measure one fo those tiny resistors directly. I'll try some of my 1K resistors I use for headlight LEDs as soon as I can get back to the bench. I know this meter is quite sensitive - it can tell the difference between a piece of #14 wire and connecting the ends of half the spool.
Checking the battery is something I always do if I suspect an incorrect reading. I often forget to mention that when I suggest using a meter. Another good reason to have more than one meter.
I wound up adding a 5.6K ohm resistor behind the 1.5K ohm resistor already in place to tone down the LED's intensity.
I called Tomar Industries to ask if they were using some new super bright LED, but they seemed unaware of any difference in their LEDs.
Dunno, I cannot explain it.
richhotrain I wound up adding a 5.6K ohm resistor behind the 1.5K ohm resistor already in place to tone down the LED's intensity. I called Tomar Industries to ask if they were using some new super bright LED, but they seemed unaware of any difference in their LEDs. Dunno, I cannot explain it. Rich
Nice that you came back with a reply. That is a lot more than I would have thought.
Hmm
We have a looooong dark hall with no outlet for a night light.
A nifty solution was to string up Christmas lights along the ceiling and run an extention cord to the nearst outlet.
We replaced the old minilights with 2 strings of LEDs about 3-4 years ago {the bluish white ones}.
They run about 6 hours a day at night, and about 4 hours every morning {10 hours total}, every day of the year on a timer.
There doesn't seem to be any difference in light/dimness since they went up.
-G .
Just my thoughts, ideas, opinions and experiences. Others may vary.
HO and N Scale.
After long and careful thought, they have convinced me. I have come to the conclusion that they are right. The aliens did it.
Yeah, everyone including Tomar seems to agree that LEDs don't gradually degrade.
As solid state units, they either work or they don't. They eventually just fail, but they don't start to dim.
My guess is that these newest signals have brighter LEDs that require more resistance to match the brightness level of the older signals.
richhotrain After I re-read this thread, especially Randy's reply, I put my ohm meter on several different resistors. I make sure that the ohms aadjustment properly set to zero and I installed a new battery before testing. I am finding variances of 25 to 30 percent, unless I am doing something terribly wrong. To assure myself that my multi meter is working properly, I tested the AC volts and the DC volts on my power pack and I get exactly accurate output voltages.
Just to update something that I reported earlier, I did some minor maintenance to my ohm meter, and then re-tested the resistance value of several resistors. The readings that I get are pretty much dead on with the labeled resistance values. So, it was the meter, not the resistors, that was at fault.
That said, I still had to add a 5.6K resistor in line with the existing 1.4K resistor to dim the LED sufficiently to dim the brighness to an acceptable level.
richhotrain I am finding variances of 25 to 30 percent, unless I am doing something terribly wrong. I measure resistance on a 330 ohm resistor and get a value of 250 ohms. I measure resistance on a 560 ohm resistor and get a value of 440 ohms.
Radio Shack makes some notoriously bad electronic components in terms of quality. But 25 to 30 percent would be bad, even for Rat Shack.
Remember you can't measure the resistence across the resistor if it's hooked in a circuit. (ie: One or both ends of the resistor still hooked to your power pack) That will alter your readings.
First thing I do when I get a pack of resistors is sit down with my volt meter and measure the resistence across each tape package. I then write the number on the tape that holds the resistors in place. That way I know precisely what I'm dealing with.
As to LED's dimming. Yes they dim. Some more then others. There are two types of LED failure.
1) Pitting of the anode caused by electron bridging the gap. Excess heat causes increased rate of break down at the junction (gap). Running excess current will cause excess heat.
2) Phosphur failure. Creating white LED's that cover a good spectrum is difficult. Most have a blue-ish tinge. To correct for this and to lower cost manufacturers the coat the inside of the LED or bulb with Phosphur. The LED itself emits UV light, which is then used by the Phosphur (Stokes-shift) to energize it and make the Phosphur glow. (Similar to CFL lights)
With #1 I've seen rated half lifes as high as 100,000 hours. That' means you have to burn it for 100,000 hours before it becomes half as bright. That's 50 working years @ 40 Hrs/week!
With #2 I've seen typical half lifes anywhere from 20,000 hours (typical) to 60,000 hours. It all depends on the amount and quality of phosphur used.
Don - Specializing in layout DC->DCC conversions
Modeling C&O transition era and steel industries There's Nothing Like Big Steam!
A final update to this thread.
Let there be no doubt that Tomar Industries is using brighter LEDs nowadays.
I sent in a double search light signal to repair a burned out green LED on the lower light.
The faulty LED was a 3-leg bi-color one.
I sent it in for repair with the wiring and resistors attached.
When it was returned, the new green LED was considerably brighter than the older green LED in the upper light.
I'm curious!! What is, or what were the tolerance level of the resistors,you used?? 5,10 or 20?
Cheers,
Frank
zstripe Rich, I'm curious!! What is, or what were the tolerance level of the resistors,you used?? 5,10 or 20? Cheers, Frank
It is a 5% Tolerance Carbon Film Resistor.
And, note, that it was the same resistor, not a new or different one.