I was finishing installing a NCE D13SRJ in a PSC HOn3 DL-535 loco I built from a kit. It was a breeze, everything worked fine until I tried to use any of the lighting. Nada.
Double, even tripled check my work and connections and all was good. The D13SRJ has 4 light outputs. It supposedly has jumper traces you can cut on the back of the board to enable the SMD resistors. Without separate resistors, I thought I'd try this for a neater install.
The D13SRJ does come with a special instruction that the rear traces are mislabeled. No problem. I'm using all 4, so will cut all four. Hooked up the two cheap LEDs and nothing. They had all been tested before use and still tested good with power applied.
Checked Blue wire voltage and it was good at 5+.
I thought maybe the outputs are fried somehow and everything else still works? Didn't make sense to me. To test that theory, I plugged in a fresh D13SRJ, called up the headlight and ZAP, fried to a crispy crust! OK.....I tried every CV I could think of and a hard reset to default settings and nuttin.
I gave up on the on-board resistors and went back to adding a small 1/4 watt one on each blue wire. Worked perfectly with the same wiring as before.
So what gives with the on-board resistors on this decoder? Are they some way incrdible high value that won't let a speck of light out of my poor LEDs? I know, because I turned the room lights off just to check. Not a photon.
Anyway, my problem got fixed, but I'm still wondering why this didn't work. This could sure neaten up some tight installs, but now I'm leery. I've had excellent luck with NCE decoders and their customer service is outstanding, but this has me scratching my head.
Mike Lehman
Urbana, IL
Mike, you give a pretty good explanation of your testing procedures in your post. But, it does raise a few questions in my mind.
On the first decoder, after you cut the four traces, did you measure the output at the four leads? Since the second decoder fried a low voltage bulb without the leads cut, I wonder what voltage is measured on that first decoder with the traces cut.
On the second decoder that fried the bulb without the traces cut, did you then cut the four traces and test another bulb? I just wonder if the first decoder failed once the traces were cut or if the same result would occur on both decoders.
Which decoder did you settle on for installation with the small 1/4 watt resistor added to the wire? The first decoder or the second decoder? I ask that because if you cut the traces on the first decoder, I would not expect the lights to work with that external resistor added to the wire. So, I am assuming that the decoder you settled on was the second decoder.
Rich
Alton Junction
I guess I am confused. I used to use D13SRJ decoders as my standard for a long time. I just checked the product page in case an update caught me off guard, but nowhere ont he NCE page do I see that this decoder even has built-in resistors.
WHat does stand out is the voltage measurement. Between what and the blue wire was this measured? SInce the resistors are on each individual function, with no resistor you should get more than 5 volts between the blue and a ground point. With the resistor, 5V should be fine, a little high but the meter has different characteristics than the intended LED and affects the circuit under test.
ANd what LEDs are you using? Do they maybe already have some resistor in them? The clue is the voltage rating, if it's a white LED and rated for anithign other than about 3.2-3.4 volts, it has some resistor already. There is no such thing as a 5 volt or 12 volt LED, they are just the same LED with a resitor incorporated into the case.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
Rich and Randy,
It was late and my explanation was probably not as clear as it could be.
Rich,
No lamps, all LEDs with this.
There were two identical D13SRJ decoders involved. The first had the traces that bypass the resistors removed, which should have enabled the resistors. I was able to verify everything was wired right when checking over the original install EXCEPT to verify that the individual light grounds worked (white, yellow, violet and green).
I finally swapped out the D13SRJ with the cut resistor bypasses for an unmodified D13SRJ. After verifying the substitute decoder worked as far as motor control, I decided to verify if the white wire worked. Knowing it would fry the LED if there was no resistor in place and was otherwise working, I turned on the front light -- Fizzzt! Dead LED and wiring verified as correct, as was the decoder EXCEPT for those pesky onboard resistors.
So I verified everything was wired correctly. Adding the separate resistors worked like it always does. But something about the onboard resistors is clearly wrong.
Randy,
Well, NCE's documentation of the on-board resistors leaves much to be desired. I've used these things on a DA-SR decoders previously IIRC, so it is something that should work.
The original printed user info supplied with the D13SRJ says NOTHING about these. Where I found out about them was through a small insert added to the user info that clearly refers to them. It notes that the said traces on the bottom of the board are mislabeled (F, R, 1, 2), but I presume they are usable.
After cutting the traces in my failed attempt to utilize the onboard resistors, I soldered a link back across the pads, installed the separate resistor and it worked fine.
I measured 5+ volts between the Blue wire and the track. It was actually around 6 volts, but without a load on it that's not out of line. My understanding is that the Blue wire typically carries 5+ volts and the resistor helps drop that to the appropriate lower ~3+ LED voltage.
I'm using Miniatronics Golden White LEDs and am sure there is no resistor installed in them, otherwise it wouldn't have blown when I activated the output without any resistor in place.
I spoke to Matt at NCE about this.
The D13SRJ does not have built in resistors. There are solder points on the bottom of the decoder, so you can add resistors. If you cut the traces and do nothing else, you have cut the electrical path to the outputs and that is why the LED didn't work. That is what was prompting me in my earlier reply to ask whether you tested the outputs for voltage.
Welllll, that does explain things.
I got the impression there were, because one of the features of this decoder is "Full support for LED lighting." I can see how the insert can be read the way it actually is -- without resistors, but capable of taking them -- but the wording is also ambiguous enough once combined with "Full support for LED lighting" it can cause exactly my confusion.
I don't think I'm going there, at least with SMDs. If it doesn't have leads attached, anyway.
But really doesn't EVERY decoder provide "Full support for LED lighting" since I have to add resistors to all the others, too, whether SMD or full size?
In fact, if I was NCE, I would put the darn resistors on the D13SRJ. Really, far more people are going to utilize a feature if it's only a scratch-off with a knife tip than there will ever be willing to start soldering SMDs to that decoder. That sort of thing is a competitive advanatage. Frankly, I'm now thinking about switching my future installs to boards that do provide this as a true feature, rather than what amounts to less than full "Full support for LED lighting."
Without digging through all the literature, do other NCE boards offer on-board resistors? I've had really good experience with them. However, like most hobby-level installers, having to install add-on resisters is one of the biggest PITAs about DCC, so anything to ease this is a plus.
The only one I know with a resistor at all is the SW-SR or whatever for the P2K SW switchers. It's an extreme pain to access and repalce the rear light with an LED so they put a resitor for the stock bulb on the decoder, plus there are sodler pads to replace the front light with an LED and it also has the resistor, just sodler the LED to the decoder and done.
Since I almost always hardwire decoders and bypass factory boards, I'm always adding a resitor to the LEDs. You only need 1/8 watt, so they are pretty small. I usually cut one LED lead and one resistor lead very short, wrap them togehter linearly, and sodler there. Then slip some shrink tube on - a piece small enough diameter to slip over the resistor and all, since it doesn;t get hot.
A lot of the TCS decoders have resistors in place, especially the Atla/Athearn board repalcement ones. I've switched to using exclusively TCS for non-sound, although I mostly hardwire in T1's unles they don't fit, it's much cheaper than the specific board repalcement types, for the expense of a few minutes witht he soldering iron and a couple of cents for a resistor - don;t buy them at Radio Shack. Since a 1K resistor is 99.9% the one you want, buy a box from Mouser, they are a coupel of cents each. I do use the Miniatronics LEDs, with the clear cases witht he little yellow dot on the junction they look like a lense with a bulb inside on locos with good light pipes that let you see the LED, better than an LED with an all-yellow case. Plus you know what you are getting, compared to the far cheaper giant lots from some Chinese seller on eBay. My locos are all first-gen diesels so I want golden-white color, not sunny white or the bright white that looks bluish.
For complex installs with lots of lights, Ulrich has a nifty board with surface mount resistors on it to make a centralized wiring board, connect allt he decoder function wires to it, and then just run each LED on a basic pair of wires, the intermediate circuit has distribution points for the blue wire so no multiple splicing, and resistors for the current control.
Matt at NCE did indicate that the DA-SR decoder has an on board set of resistors, but only for LED lighting.
Here is what the DA-SR manual says:
Built in resistors for LED lighting: We provide built-in resistors if you wish to use LEDs for lighting. These resistors are disabled at the factory with a “U” shaped shorting link built in to the circuit board. To use these resistors you must cut the link associated with each LED.
Thanks for your patience and all the useful info. So it was the DA-SR that I've done this with. At least I haven't totally lost it
Yep, I actually have one of those Ulrich boards. Didn't think to use it, but would have also worked. For some installs, it's just too big and that's where the on-board resistors would be so handy.
At least I know there's alternatives out there. I've pretty much finished my fleet conversions and mostly I buy sound-equipped when I do.It's the stuff I build that I have to figure out DCC on.
This NCE install is a placeholder for what I hope eventually is Tsunami sound. They don't have quite the right sounds yet for the 6 cylinder in the DL-535. The Micro-Tsu, which I'd like to use, only provides two light outputs and I need at least 3. So I'm being patient, which is helped by the fact my modeling budget is already depleted by recent Blackstone arrivals. Those folks in Durango end up with most of my $$ one way or another, but I'm REALLY NOT complaining...
Bring on the next run of drop-bottom gons!
mlehman ... But really doesn't EVERY decoder provide "Full support for LED lighting" since I have to add resistors to all the others, too, whether SMD or full size?..
...
But really doesn't EVERY decoder provide "Full support for LED lighting" since I have to add resistors to all the others, too, whether SMD or full size?..
No, not really. To clarify what NCE means by "Full support for LED lighting", they are not talking about the hardware, but the decoder software. Since LED's respond differently than incandescent bulbs, ligting output algorithms designed for bulbs do not look quite right with LED's. If all you are doing is on/off, then it doesn't matter, but if you are using special effects, like flashing ditchlights or Mars lights, then it does. Decoders with full support for LED lighting have a setting that modifies the outputs to look correct with LED's.
Yep, I understand that now -- it's marketing terminology that's clear as mud to the average end user.
All I'm saying is that there is a marketing opportunity to produce boards that actually offer full software and HARDWARE support for LEDs. I'm really not sure why the relative scarcity of such decoders exists at this point in the relatively mature DCC market.
Maybe for the good of the fraternity, some secret, arcane initiation rites must continue to exist. In the case of DCC, that's resistors.
OUCH!
Darned paddle.