How do you fix Fastrack controller LEDs?

The bulbs are burning out due to some circuit design issue.  I just got finished replacing the LED's in three of mine, got tired of seeing dark out there.

I send a query to Lionel to see if there's a mod to the switches to stop them from cooking the LED's, obviously they are driving them with excessive current.  When three out of ten switches fail, it's not a component issue, it's a design issue!  Those LED's should outlast most of the people that buy the switches if the circuit wasn't frying them.

So this is on the controller and not the led on the switch lantern? Seems like the controller schematic is pretty clear and the signal voltage is +/- 5vdc

Sorry, I mis-read that, doubtless due to the fact that I have those LED's on switches biting the dust regularly!

I don't know why the controller LED's would be going, they don't seem overloaded, at least according to Chuck's diagram.

Once I had to replace the 220 ohm resistor in the upper right corner of the switch pcb. I think the problem was the lantern not lighting...

That could be an area of question if lantern leds are burning out.

I would check the yellow wire connections, especially if you extended the wires.

It's apparent that the lanterns are getting too much current, just don't know why.

Hey gunrunnerjohn. I just sent you an email at the verizon email address. Or do I need to send it to the gmail one.

Either works, got it.

oky doky

Gentlemen,

   I have one Command Control switch with an out LED also, it came, NIB that way and I did not want to send it back to Lionel for warranty repair at Christmas time.  I have been placing the manual controller track side with numbers visible to identify how the turn out sets.

If one of you guys can fix the LED, I will pay you to fix mine, after the Christmas layout is taken down.

PCRR/Dave

I got mine from Henning's Trains here. They have one in stock. John, can you fix my other three for $4.95 apiece? I'm sure your expertise is worth a whole lot more than that.  I have a set of instructions on the procedure from an old post if anyone else is interested. I need to work on my soldering skills some more before I tackle one. I still have a couple replacements left before I have to try one.

That's not necessarily the problem in every case.

Like Chuck says, there could be other causes.  OTOH, I've seen a bunch of these, not only with my own switches, but others as well.  This was the first time I decided to actually see why they were out, I was actually surprised to find it was all bad LEDs!  Chuck is the first guy I've heard that had the LED out that it wasn't the LED.

You can test the LED on the lantern bar with an AC transformer and a 1k resistor.  Just dial up a low voltage on your transformer and use the resistor in series with the power to the two contacts on the lantern bar.  If the LED doesn't light, it's bad if you're using AC.  Assuming that's what you find, you can take the lantern bar apart and see what you have to do to fix the LED.  Truthfully, I'd probably have to charge more than they're worth to fix it, so I'd probably just wait for Lionel to have them in stock again.  I just used plain white 3mm LEDs to repair them, that's what is in them.

Guns,

   Looks like I messed up when I did not send it back for warranty repair when it was faulty NIB.  If the labor is that intensive I will keep using it the way I have been.  It's already programmed and works properly.  Bummer!  IMO any time you as the operator can not simply replace the light in a switch, it's poor engineering design.  LED or not.

PCRR/Dave

It's not "that bad", it's just taking the lantern bar apart, soldering a couple fine wires onto the LED and reassembling it.  Truthfully, I don't know why the LEDs are failing in any case, that's what I'd like to know.

I bench tested a switch and the current to the LED is only around 9ma, that should be fine.  I replaced it with a 100 ohm resistor to see if that changed anything in the circuit, that gave me a little over 10ma, still well below the LED rating.  I have no idea what is taking the LED's out.  If I were to guess, I'd say voltage spikes in the reverse direction are cooking them, but that's just a guess.

For five bucks, it's not all that difficult to change. You probably already spent over a hundred; might as well have it right. That's why I figured John couldn't do it for that, so I wasn't going to ask.

Guns,

   IMO it should be able to be popped out top side, like the old type switches and replaced when bad, by the owner operator.

PCRR/Dave

Maybe so Dave, but LEDs should never have to be replaced.  Look at any Lionel locomotive with LED's nowadays, you don't just "pop them out" and replace them either.  You end up replacing a small light board as a unit typically.  On some of the GP light boards, I can't even find a source for the odd shaped LED's they used if I did want to replace the LED.

No argument there, Dave. Taking the switch off the layout is harder than changing out the switch stand. That job is the same as switching sides with it. That is why I chose to get the whole stand from Henning's and go that route.

I dont have the switch. If input is 5 VDC the 470 ohm resistors are more than adequate. Perhaps voltage spikes from the coils are making it to the LEDs.

Dale H

No coils, only a very small DC motor to accomplish the switching function.  I'm at a loss what takes the LEDs out.

Why not set your meter to measure peak voltage on the LED and ,then operate the switch?

Dale H

I have a left hand O36 switch that is DOA from a bad board and of course the replacements are not available. My LED was intermittent and dim on that switch for a long time until one day nothing. It ran on track power and then later using the accessory power.

The LEDs that are going out are not the ones in the controller, but is the one in the lantern on the switch. We don't know how that led is wired. I think it has a 220 resister associated with it but don't know what the voltage is. It's so bright that one could guess maybe 220 ohms from 5 volts. If I get inspired further I'll take one apart and measure just to see. 

Guns,

    IMO the LED's are actually damaged in shipment or faulty when shipped to Lionel to begin with, I do not believe it's from normal usage.  As you indicated they should really last a life time.  I do not think Lionel accounted for a problem like this, when they engineered the LED into the FTCC switches.  I do not think it was planned obsolescence either, just a poor engineering design for owner/operator replacement.  I do not think they ever planned to have to actually fix any of their switch side LED's.  However because the LED's are foreign made with little quality control/testing, so many of them are actually shipped to Lionel faulty to begin with, this may really be the true problem.  Heck the 990 Legacy units are not fully Quality Control tested, and they are a fairly expensive electrical piece, there is no way every little LED is ever going to get fully tested.  Every 20th LED is probably shipped to Lionel faulty, it's the way they do business today in many of these foreign Companies.  I can almost bet that everyone who owns 20 or more FasTrack Switches has at least 1 or 2 bad LED lights in their switches on average.

PCRR/Dave

I can attest that i have 19 switches on the layout and I have had to replace one as it started out as a dim lantern on the switch and one day the entire controller board died. Lionel does not have these available either which is a bummer.

How did the board fail...did the switch not work either or just the led go out?

I don't believe the LED's are damaged in shipping Dave.  The LED is a very robust package, and all of my failures came well after the switch was in use.  Add to that the fact that it was the physical LED bulb that is bad, and shipping damage becomes even more remote of a possibility.

I buy tons of cheap LED's from overseas, I've used hundreds of them.  I don't get failure rates like the Fastrack switch lantern bulbs, I really don't think this is a component problem.

What kind of meter do you have that measures instantaneous peak voltage?

My Fluke 87 does. I just did a demo for me showing connected to a CW-80, it shows a max of 13.91 vac and a min of 13.76 vac and an average of 13.85 vac.

This is interesting as to the thread about the CW-80 varying, which is most annoying when used for lighting. Certain settings can show quite a difference in light intensity with this rather un-damped feedback loop as in the CW-80. Kind of causes one to wonder who designs these. Maybe I'm being too harsh, and maybe it's mostly the variation of parts values. But then design speaks to where to use tight tolerance and where you can use loose specs, eh?

I measured the average current and it was only half of the rating, so unless they're getting reverse voltage spikes, I can't imagine excessive current is the issue.

When I had a job, (before it was moved to Mexico)  I used to have a Fluke which measured peak voltage. It would remember the maximum peak cycle, such as a voltage spike during the measured interval. That meter bit the dust a long time ago. All I have now are cheap meters, $10 Sears ones. 

You could hook a zener diode 5 volts in the circuit and hook a bulb or something to it. A spike would light it up briefly.

I just suspect some kind of spike in this instance,but I don't know. This would slowly degrade the LEDs. 

You have a simple circuit, 5 volts input DC, in series to an LED with an adequate resistor. I only see 2 possibilities.

A faulty LED, made in bad batches.

A voltage spike or surge, generated off something. 

This is just my take on it.

I had some MTH budd cars burn out their red LEDs markers. MTH says in their schematic and description they are rated at 6 volts and 6 volts is circuited to them. I never heard of 6 volt LEDs.  When I replaced I reduced the voltage with a resistor and my own regulator. They still light adequately and I have not burned one out since. 

One thing I learned is that EEs do not always do things right,no disrespect to you EEs out there. 

Dale H

I think it's manufacturing variances in the LEDs. 

I just measured one of the new CC switches and the led has 3 vdc across it from a 100 ohm resistor which drops 1 vdc. So that's 10 ma.

I measured the NON CC switch LED and it has 2.9 vdc across it coming from a 68 ohm resistor which has 0.9 vdc across it. So that's about 13 ma. Somehow there's a diode and a 220 ohm resistor associated with the supply to the LED, but the bottom line is 13 ma.

None of this sounds like a bad design.

Which begs the question of why they're dropping like flies.

I'll have to agree with Dale, I've personally screwed up some circuit designs.

Guns,

   No matter how you look at this, it's still a problem to the operator/owner when the LED is faulty, with so many of these happening Lionel should stand behind the product and fix them for no cost.  Then maybe this LED problem would get straightened out, before the product was delivered/sold to the customer. 

PCRR/Dave

I can't argue that it's apparently a product defect of some kind, but I wouldn't hold my breath for Lionel to fix them after the warranty.

Guns,

   I hear that, however just like the 990 Legacy has a extended warranty, maybe they should do the same with these FTCC LED lights in their seriously expensive switches.  I do love these low voltage Command Control switches, great product, the faulty LED is a pain however.

PCRR/Dave

From your lips to Lionel's ear.

I'm not as worried, I'm getting good at replacing the LEDs.

If the LED is powered form a 7805 circuit,there should be a .05 np capacitor on the regulated end. If this is missing or defective or wrong value spikes could make it through the circuit. I would like to have a meter on the LED in operation when the turnout is thrown to and fro. I don't know how much backward EMF a small cheap DC motor can generate. 

A possible fix,if it is spiking is a 5 volt zener and a .05 tantalum cap across the LED leads or something across the motor leads,if that is the culprit. .  Or put a much larger resistor in series to the LED. I don't have the switch in question nor a schematic so I can't go much further. Kind of like a lot of train stuff,you have to guess,very frustrating.

Dale H