Tuesday, June 14, 2016

Electrical wars, Part 10: DCC refinements

In a recent post in this series about my efforts to conquer electrical problems, I depicted my installation of a circuit breaker device which permitted separate layout districts, each with its own circuit breaker. That post can be found here: http://modelingthesp.blogspot.com/2016/04/electrical-wars-part-8.html .
     My initial solo operation with this arrangement worked great, and so did my first operating session with the new arrangement. That is, it worked fine until a persistent short occurred (a car wheel resting on a gap that we didn’t find right away), and I discovered that the circuit breaker for that district wouldn’t reset. If layout power were turned off, then turned on again, it would reset fine, but if we shorted the track with power on, it would not reset. This was mentioned in my post about recent operating sessions (at: http://modelingthesp.blogspot.com/2016/06/spring-operating-sessions.html ).
     As always with the trouble-shooting of such problems, I am very much offering the short version of the story. Disconnecting wires to various layout areas was tried, moving locomotives and cars was tried, etc. at some length, and various hoped-for remedies were tried, but there’s no need to repeat all of that. My friend Ray DeBlieck directed me to an explanation of the problem, and also a solution.
     The clearest explanation is an essay by Mark Gurries, posted in 2004 to the NCE-DCC Yahoo group. It can now be found on-line in a variety of places. I think probably the most informative discussion is on the site at http://www.members.optusnet.com.au/nswmn/cb_sound_reset.htm , which has further description of the issue and solutions to it, as well as the full text of the original Gurries article.
     In brief, the problem with older circuit breakers such as the NCE EB3 I have (installation of it was described in the post cited at the top of the present post) is that with multiple sound-decoder-equipped locomotives in a single district. the “inrush current” to capacitance in the decoders as soon as re-set begins, fools the circuit breaker into thinking there is still too much current flowing, meaning the short is still present, and it opens the breaker again. Net result, it fails to reset. I’m not an electrical engineer, just quoting from the Gurries piece.
     The solution is to insert a resistance into the circuit, usually suggested as an automobile brake lamp, so that the lamp lights when a short occurs, and thus is present when the breaker re-sets. The light goes out when reset is completed. Here is the suggested arrangement for an EB3, with the track power leads at right.

The automotive bulbs I chose to use are the widely available no. 1156. These are recommended in a number of web discussions. They are available as either filament bulbs or LED lights, and I chose the filament bulb. I also bought the automotive sockets for them.
     The bayonet sockets for no. 1156 lamps have a 3/4-inch  diameter. I just drilled holes that size in a small rectangle of plywood and painted it flat black, to serve as a mounting board for the bulbs. The mount went underneath a fascia support so it is out of the way, but since the bulbs are very bright when lit, there would no problem spotting the light.
     When all wiring was done according to the diagram above, I tried the classic DCC short method, namely a quarter placed on the track. This of course trips the circuit breakers, and the bulb for that district comes on. As soon as the quarter is removed from contact, the circuit re-sets and the bulb goes out. That is just what was supposed to happen. Then I put four sound-equipped locomotives in one district, so I could test the “inrush current” correction feature. The circuit breakers still worked the same as just described, no problem. Here is a view of the operation:


Yes, the bulb really does have a blue cast, as the photo shows.
     This is really “short insurance,” for a persistent short rather than a momentary one, but that was the problem in the last operating session. With this additional circuit, it shouldn’t happen again. (He said cautiously . . .)
Tony Thompson

3 comments:

  1. Three of us operated the West Virginia Northern in an industrial park in Creswell, Oregon. Based on the WM & B&O, it featured six mines and a collery spread around three walls. Wiring was a nightmare!! A true rat's nest of long strands of spidery webs set to ensnare. Kenny was the fellow who knew it like the back-of-his-hand. All was well till Kenny split for Colorado to climb mountains... Well, long story short - it wasn't long before failure proceded closure.
    But, it did teach me to document the schematic well.
    Regards,
    Harry

    ReplyDelete
  2. Tony, you mention that you opted for the filament version of the bulb, instead of the LED version. The casual reader should know that the filament version is REQUIRED for this solution to work.

    ReplyDelete
  3. Thanks, Jeff. I assumed that would be the case, as the filament warm-up provides a changing resistance, and that seemed like it was important. But not being an EE, I only surmised that, so I appreciate your clarification.
    Tony Thompson

    ReplyDelete