The Dumont Museum has two large train layouts and an extensive Oliver tractor display housed in a steel building. The smaller layout is about 15’ x 20’ with fairly simple loops of track on the flat and in a mountain. This layout operates properly.
The larger layout is located in an adjacent room that is 60’ x 100’. This layout is composed of two sections with interconnecting bridges that cover about 45’ x 90’. This layout has problems with Legacy operation due to weak track signals.
Lyle Dumont and I have spoken about the problems on a number of occasions, but have been unable to resolve the issues. I volunteered to visit so that we could tackle the problem first-hand.
Day One – Thursday, Oct. 11, 2012
The first step was to hook up and oscilloscope to the Track output of the Legacy Base. I used an isolation transformer on the oscilloscope to permit me to establish an arbitrary ground reference. (The Legacy/TMCC track signal is referenced to ground through the wallwart. I thought I might need to reference to other points.)
A quick examination showed that the 455 KHz track signal was being cleanly chopped off on the lower half cycle. The wire to the track was removed to unload the output, but the output was still cleanly chopped off.
I have seen this distorted output before, including my own Legacy Base that I acquired because it needed repair. The output amplifier components include several capacitors and a driver transistor.
Conclusion #1 – The Base was probably damaged by a nearby lightning strike that caused damage to several electronic devices in the adjacent Dumont home and the air conditioning thermostat in the museum.
Action #1 – We borrowed the Base from the smaller layout, and it had a proper unclipped signal of 5 volts peak-to-peak without any load.
Next, we moved on to examine the troublesome areas on the layout. The older east section was operating fairly well, but the newer west section was very poor. This newer area had been constructed with aluminum foil duct tape laid under all of the track roadbed, on top of the tables. (Foil tape had also been added to the east section, but well after construction. This foil is under the table, in some cases wrapping over support beams and other construction obstacles.)
Initially, the foil was tied together, but not connected to earth ground. We soon tried tying the foil to ground, but this caused a serious drop in track signal level and bad waveform distortion on the negative half cycle. We experimented with several configuration of foil grounding, but could not clean up the waveform.
Conclusion #2 – You can actually have too much foil! In this case we had somewhere between 300 and 500 feet of track with foil very close to the track. Most of the track in this area is Gargraves-like track with non-conductive ties, but there is also some tubular S-gauge track. The equivalent area of 500’ of 2-inch-wide tape is about 80 square feet of aluminum foil! This foil is located in close proximity to the other half of the command signal, which is on three rails of the track, forming a very significant equivalent capacitor. The amplitude drop and distortion on the lower half of the track signal are typical for the output stage when it is overloaded with capacitance. I didn’t have a capacitance meter, nor did I have an assortment of capacitors that I might have used to try to simulate the observed waveform with lumped values. This will need to wait until I return home.
With the massive amounts of foil acting as an enemy, we moved on to Plan B. Lyle had installed some discrete wires next to the track. Some of these wires were exposed, but he had covered some of the wires with glued gravel ballast.
Some of the wires worked well, but some produced the heavy loading we saw with the grounded foil. We finally noticed that some of the staples holding the wire next to the track were stapled into the foil, and some of these staples pinched the wires enough to cut through the insulation to connect the wire to the foil. The bare wires were fairly easy to troubleshoot, but the ballasted areas were a big problem. In some cases we chose to abandon the buried wires and temporarily add new wires on top of the ballast.
The wire grid of earth grounds was considerably less capacitance than the foil, giving us less signal drop, but still about a 30% drop and some negative-loop distortion.
Our test engines were a Legacy Santa Fe 3751 4-8-4 and a Milwaukee 261 4-8-4. Lyle had found that the steamers were more troublesome than the plastic-shell diesels. (Does anyone have information to share on the command signal performance of these locomotives?)
Even with the wires running between tracks, we still had some areas where the steamers would flash their headlights or even stop.
Conclusion #3 – I am going to build a combination track signal booster and signal strength meter that can be used to measure drops in signal level due to loading by capacitance. The booster will feature a symmetric amplifier with some voltage gain but primarily enough output current capability to avoid distorting one half of the track signal.
When we finished after two days, things were better, but not yet perfect. Lyle will do the beta testing of the MANCO Track Signal Booster/Meter.
Lyle and his wife Helen were wonderful hosts. I thoroughly enjoyed my visit and their great hospitality. Lyle and I worked long and hard, but we had a great time. (And then there was that terrific grilled pork sirloin sandwich….) I feel that I am very lucky to be able to visit and help wonderful people like the Dumonts!
I have built a booster amplifier based upon the vacuum tube design of Jim Lefevre, adding a few embellishments of my own. I hope to test the amp on some large layouts in the near future.