Re: Sub Bus "beyond" Main Bus
Problem can potentially appear when the bus wires go beyond 30 feet. It is not a cliff where at 31 feet everything starts to fall apart and decoder stop working. It become a shades of gray problem that only gets darker as the bus length gets longer.
If you have OLD existing layout wiring, you can go beyond 30 feet with a RC filter installed at the end of the track bus.
If you are installing new wiring, you can twist the cable and go a lot farther without installing a RC Filter. Somewhere close to 50ft or so, a RC filter can start to become advisable again. The actual distance depends on how well you twisted the cables.
Twisting the track bus and installing RC filters are both about preventing potential problems.
Where do the numbers come from?
When one reports a layout has no problems with excessive long track bus runs, the recommendation is often based on the false idea that there will be a clear indication of a failure when something goes wrong. Things can go wrong long before one actually sees something go wrong.
Decoder in moving train receives DCC packets regularly the contain speed and direction information. The assumption is the decoder may be on dirty track and momentarily lose power in which all speed and direction information is lost by the decoder. To keep train running reliably, the DCC standard by design require redundant speed and direction DCC packets be sent. Hence when the flywheels get the engine past the dead spot such that DCC power is restored, the decoder knows right away what speed and direction it need to be at to continue pulling the trains without any surprises. When a invalid DCC packet is received, it is thrown away by the decoder as if it was never sent. But because of the redundancy of the sending the DCC packets, eventually they get through and no problem APPEARS to exist because the train keeps moving. This is the key to DCC's robustness and high tolerance when dealing with less than ideal electrical transmission medium due to the naturally intermittent wheel to rail electrical contact.
Now lets say the moving train decoder goes out on the layout that has long track bus wires. The decoder will start to experience GROWING high DCC packet error rates due to growing noise and other electrical distortions of the DCC signal created by the wire properties. This is on top of the existing problem is of the trains own creation due to the less the perfect electrical wheel to rail contact. The number of successful DCC packets getting through falls as the engine get farther away. The point is problem are present, but you may not see them.
So how does severe problems look like? Sever problem are potentially observable problems. The problem shows up in the form of lost of control. However to know that one has lost control, the engineer has to want to make a big changes to the train speed or sound the horn in the area where the failure is occurring to see it. This results in a delayed response in the speed change -OR- the horn never plays. Some people will mistakenly think they did something wrong on the throttle and dismiss the problem as their fault. The point being is that one seldom or never makes any adjustments in the trouble are at the end of the long bus run, they may never notice the problem for the train may eventually get back to an area were there are no more problem and all appears to be fine.
There are scientific ways to detect problem long before they are visible and that is what the recommendations are based on.
DCC Website & NMRA DCC Clinics: www.markgurries.com