Bought a Proto 2000 USRA 0-8-0 while on vacation, along with a basic decoder (NCE P2KSR) to get it going. Got it home, got the decoder in, got it running no problem. Great. But the lights came on for only the briefest moment and wouldn’t come back on again. After a little research, I discovered that with this particular engine, the lights need to have resistors added when using this particular decoder.
Got new leds. And resistors. Put them in. Nothing. Reversed the decoder. Nothing (save that the engine ran backwards). The leds test out fine on pure track power when put in series with a 470 ohm resistor, but quietly blow without the resistor (I bought extras so I could test). Figuring that the decoder’s light output might be below track voltage (and being unable to find any literature to confirm or deny this), I hooked an led straight to the decoder’s rear light output and powered it up. The led lit up and blew (which I kind of expected), but this time it took the decoder with it.
Bottom line? Not impressed with NCE at this point. A) the decoder has no indication as to which is the proper orientiation for 8-pin installation. B) there appears to be no specs available to define the voltage of the lighting (or any) output. C) The decoder’s entire instruction manual (online and paper) is basic and contains no specs at all
I’ll be returning to Digitrax decoders, thank you very much.
Page 2 of the instructions under where it says recommended lighting pretty clearly states that “if you use LEDS we recommend a 1K ohm 1/4 watt resistor in each function lead”. Decoder pin 1 is also identified on the diagram on the same page.
Also on the same page it says “always test your decoder installation on a current limited programming track before trying it on full power”.
Sounds like the resistors you used weren’t high enough in resistance. I have not had any problems with NCE. When I have fried one or two of their decoders, it has been negligence on my part. Theyre customer service has been awesome,and theyve even fixed my goofs under their warranty policy for free (normally $10 plus shipping.)
In my non-sound units, I run almost exclusively NCE. They run flawlessly. Ive only had two Digitrax decoders one of which was a low quality sound decoder. It wouldnt program when I installed it. Digitrax’s customer service was so poor I still have a bad taste in my mouth from it.
The only decoder I blew in a fleet of 63 locos is a NCE N14IP. It was not the decoder fault but the rather piece of junk it was installed into. Namely the first run of DCC ready MTH SD70ACe…
So what you are saying is you blame NCE because you screwed up? Funny how personal responsibility has gone away. Like blaming the cop because you got caught speeding.
How is NCE supposed to know what your track voltage is? Orientation of the 8 pin plug can be found with very little research at all. Pin one is always the orange wire or some sort of mark on an integrated board decoder. And it is common knowledge that LEDs commonly use 1K or more resistors for reliable operation. Buy a meter and learn how to use it before blaming someone else.
I have well over 100 decoder instals across my workbench. Only popped 2 decoders and both were Digitrax plug and pray. After Digitrax weaseled out of warrantying them and wanting to charge me for fixing them I swore off Digitrax. I doubt they miss me much.
In the heat of the moment I suppose it was. I was working with an unfamiliar product doing something I’ve never done before. My own inexperience combined with what looked at the time like less-than-stellar documentation cost me a decoder. NCE’s fault? I don’t suppose I can blame them for not speaking the same language as me. Perhaps I need to learn how to translate a bit better.
Pete, if I may be so bold as to ask your advice, what resistor/led combo has worked for you in the past?
If I can fit it in I use a 3mm white or sunny white with a 1/4 watt 1K resistor. Some applications I have used surface mount LEDs with up to 3K resistors. I have never put less than 1K on any LED and have had no problems. Once I did a loco and burnt out an LED and it turned out to be a bad resistor that was miss packaged. After that I take a few seconds and ohm the resistor with a multimeter.
Maybe,it’s just me, But I’m wondering why?? You started your post,with Hooray!! and then,changed to becoming,really up-set about it… I read and viewed,the click on post that was provided,to you and it looked,pretty straight forward,to my eyes, Oh WELL!!
Sorry to be confusing, Frank. I was thoroughly frosted at my engine, and the hooray was sarcastic. All for good, though; Pete’s offered some very helpful advice that I’m looking forward to trying out.
I have comfort with TCS as when I mess up installs for whatever reason TCS replaces them. Not only are they Goof Proof - but Brian from Buffalo Proof! [:D]
Replace any and all factory boards and hard wire the decoder whenever possible. That way you will know when you need a resistor - aka, ‘ALWAYS’ since there is no factory board standing between the decoder and the LED. Now, where this LED is depends ont he decoder, some decoders alreayd have current limited outputs for LEDs. Others require an external LED.
The minimum correct resistor for white LEDs is 1K. You can squeak by with less, but the current will be near the LED’s limit. 1K keeps it nice and low. Sometimes the LED is too bright with 1K, and you need to use a higher value. But I’ve never seen a case needing a smaller than 1K resistor. So stock up.
The function outputs on the decoder are limited in what kind of current they can deliver, regardless of the type of load being connected. You actually got very lucky the first time when it did not blow the function output. Usually when the LED or light blows, it takes the function output with it, no retries, no do-overs.
You can either replace the decoder, or use a different function and map it to F0 for your headlight - this may require bypassing the factory board, see #1. Of course, bypassing the factory board means the P2K-SR is not the correct decoder, in fact you can get away with the much less expensive D13SRJ. See another reason I always do #1.
I stopped using Digitrax motor decoders a long time ago, because the ones with BEMF, well, the BEMF kinda stinks, and the ones without, you cna usually get others without BEMF cheaper. For a while I was using the NCE D13SRJ - you cna get 10 packs for under $120, that’s less than $12 per decoder. For a non-BEMF decoder, the running qualities are excellent. And it’s a silent running decoder so no buzzing like the cheapy Bachmann decoder that they have the gall to ask $20 or more for. I have since switched to all TCS for motor only decoders, they have excellent BEMF motor control that is nicely self-adjusting, no fooling around with
Furthering the boneheadedness of this whole effort, reading up on the Tony’s Trains website clues me in that I almost certainly installed the LEDs backwards. For a guy who worked in electronics manufacturing for a decade, this is pretty durned embarrassing…
Thanks to all for endeavouring to drag me out of the congealed mud of my own stupidity…
If all else fails, it did, follow the instructions. Don’t know what the instructions mean, do research. The Internet is loaded with all kinds of useful DCC info if you do not need to be led by the hand.
Backwards LED won’t hurt anything, definitely not if the resistor is there. It just won’t light. No current will flow, nothing will fry. But yes, the blue common wire on the decoder is the plus side of the circuit.
Well, not exactly - the problem was using no resistor the first time. Depending on track voltage, 470 ohms results in current right at or maybe a milliamp over the limit, so it should work ok without instantly making a flashbulb. It’s definitely too small a resistor to keep there, but it shouldn;t cause a failure, at least not for a few hours. That LED might actually be ok, the decoder was stressed from blowing the one without a resistor and running it over 2/3rds of the limit per function output might have just been too much. A second one without resistor though - forget it, the decoder’s going to go with that one - pretty robust that it didn;t take out the function with the first blown one.