Replacing bulbs in Kato 206 light kits

[Bernard]

I have lots of the Kato 206 interior lighting kits and since I've changed over to DCC I chose to remove them from my trains after have one car started to melt. Because of costs, I don't want to get all new lighting kits so.....using the existing kits, is it possible to replace the incandescence bulbs with LEDs? If so where can you get LEDs that fit into the housing unit for the lamp?

[CaptOblivious]

You can, but it is an incredibly tedious process. I've done it with one set that had the bulb-type lights permanently installed at the factory:

http://akihabara.artificial-science.org/2010/01/18/kato-651%E7%B3%BB-dcc-pt-1-upgrading-interior-lights/

[inobu]

Good install Capt O.

Inobu

[Bernard]

Don - Great article and nice install......and it has convinced me not to try and replace the bulbs. Now what to do with all those 206 lighting kits I have. 

[CaptOblivious]

sell 'em on eBay! Surely there are folks out there who prefer bulbs and aren't using DCC?

Did it really begin to melt one of your cars? I hadn't actually heard any first-hand stories of this happening yet. Pictures?

[Bernard]

sell 'em on eBay! Surely there are folks out there who prefer bulbs and aren't using DCC?

Did it really begin to melt one of your cars? I hadn't actually heard any first-hand stories of this happening yet. Pictures?

It was my Sonic head car and I got it just in time! You can see where the "soft" spot was starting. My LHS has a sigh that they won't be held responsible if you use the 206 sets and run DCC.
Here is a photo but looking at it again, it doesn't show the soft spot where the bulb would be.
 

[inobu]

The problem is based on which DCC system you are using and its output track voltage. The decoders are rectifying track voltage back to a straight DC voltage. The potential heating problem is linked to this conversion.

Incandescence bulbs are heat-driven light emissions with no control or regulation. The more you give it (bulb) the more it will take until it blows. Our problem is, when we convert to DCC our base or track voltage is higher than the designed specs for the bulb. We normally run our DC track at 6 - 8 volts even though its maximum voltage level is 12 volts. The only way track power reaches 12 volts is when we run full throttle which we rarely do.

Here is the twist. Some decoders regulate the output voltage and some don't. The "drop in" decoders regulate the output voltage to the LED and do not require resistors. I tested Digitrax drop in decoders and found 2.5 volts  on the LED (Fwd Rev lights) leads of the board. On the wired decoders the output voltage on the Forward and reverse light lead (white and yellow) have 13.8 volts (Zephyr DSC50). This means that the lamps will be running on almost 14 volts constantly. Cooking your roof like an Easy Bake Oven for this reason we should not run bulbs in our cars. 

The remedy is controlling the voltage and we accomplish that by introducing a resistive network or resistor. Even that creates a problem. The reduction in voltage with a resistor creates a by product "heat". Heat is linked to watts. Because the resistor is going to generate heat we need to match the resistor's wattage rating with the amount of current or voltage flow we are going to reduce. This is a must for LED's

I know this seems to be going off on a tangent but this information helps one to understand what is behind the lamp issue and the importance in stepping down the voltage for the LED to operate properly.

I think Capt O solution is the best way to go, the only caveat it match up the right resistor to step down the voltage for the LED. Another thing is not all LED have the same forward voltage. You need to match all components with their perspective ratings in order to have a solid lighting setup that wont prematurely burn out.

Inobu

[CaptOblivious]

I had thought it was simply because DCC was an AC signal, and that bulbs get hotter when powered by AC than by DC?

Reason, as I understood it: Bulbs draw more current on AC, because the in-rush current (about 10x steady-state current) is drawn when the bulb is powered up. On DC, the bulb gets powered on and stays on. On AC, and esp. square wave DCC, the bulb is powered on and off rapidly. Although the repetitive in-rush current isn't as much with high-frequency DCC (because the bulb doesn't have enough time to cool all the way down between cycles), it's still quite high, and the resulting increase in power (power = voltage * current) has to be dissipated as heat. The increased voltage of DCC, you rightly point out, is certainly not helping!

(Bulbs connected to the outputs of decoders should not be a problem, however, since they provide rectified power.)

That said, now that I think on it, Bernard shouldn't have to switch out the bulbs for LEDs. He need only provide a rectifier, like the one I used in my post, wired in just like a decoder, or an actual decoder capable of handling the in-rush current of a bulb (Kato FR11 CANNOT handle this in-rush current, so that is sadly not a solution).

[inobu]

Capt O

Yes the current rush is the major factor that contributes to that inherent problems between DCC and bulbs but even if it were converted to straight DC the increase in the operating voltage from the booster is actually the operating DC voltage of the decoder. The bulbs were not a problem on a DC track because the average voltage on the track is 6 to 8 volts. Place the Bulb on a DCC track and it at a constant 13 to 14 volts. This increase in operating voltage over DC tracks contributes to heat soak of the plastic just the same, not as fast as alternating waveform of DCC but it still has its effect. Our cars are not designed with ventilation and cannot dissipate the heat from the bulb. In my opinion these combined factors make we weary.   

The suggestion of the LED has merit in that it resolves multiple issues. Power consumption, heat generation and it supports the Green initiative (lol).   

As you probably have read I am weary of heat. The plastic used in the trains are getting thinner and thinner and will warp easily. Bulbs are heat driven components which can contribute to a warping problems. The further you stay away from heat the better off you will be.

I'm having issues with light my self. I would be really ticked off if a $.05 bulb ruined my $200.00 train.   

Inobu

[CaptOblivious]

I'd be ticked too!

In the interest of full disclosure, I did not replace the bulbs in the headlight unit of my Super Hitachi. I haven't tested the voltage they're receiving from the decoder, but at least its a straight DC signal. I could always stuff a current-limiting resistor in there if I had to, I think, or a diode to give a 1V drop or so.

[inobu]

I'd be ticked too!

In the interest of full disclosure, I did not replace the bulbs in the headlight unit of my Super Hitachi. I haven't tested the voltage they're receiving from the decoder, but at least its a straight DC signal. I could always stuff a current-limiting resistor in there if I had to, I think, or a diode to give a 1V drop or so.

It depends on the decoder, I tested the Digitrax DZ125's and they have 13.8 volts on the light leads. The funny thing is the "drop in" decoders have 2.3 volts on their lighting leads. I'm guessing that they supply full voltage on the wired decoders which allows for lamp and function diversity.

I removed all of my light until I can configure the boards for the supplied track voltage. .

Inobu 

[Bernard]

inobu - I think you made a great observation about no ventilation inside the cars plus the heat of the bulbs. I did exactly what you did, removed all my interior light kits.... but I don't have to worry about melting a car shell.

[Martijn Meerts]

Very few wired decoders are designed to handle LEDs directly. I'm guessing it's mainly because wired decoders tend to be installed (in Europe at least) in older trains which use bulbs.

Often it's easy to replace the resistor on the lightboard, sometimes it's possible to use the resistor already there.

[CaptOblivious]

Bernard: Another option:
http://www.snjpn.com/ngdcc/lu1k/lu1kj.htm
Some members have ordered from this guy before; he's apparently willing to send stuff to the US, but without warranty support. I can vouch for his English, as well. ¥2000 for four units means about $6–7/car (not including shipping!), so the cost may not be worth it.

[Bernard]

Thanks, Don for the link......maybe I can say everyone on the train went to sleep and decided to turn off the lights and that's why all the cars are dark