开云体育

Mike, You can't fool Mother Nature. 40v for 200nSec across a 100 ohm resistor 8,000 times/sec is only 3.2 micro-watts. You have something wrong with your set-up. Wrong value of C or shorted C or a polarity sensitive C or the resistor is somehow wired directly in parallel with the track. With your R-C networks installed you should NOT see any more voltage ringing at the track. The ringing IS gone...Right?

However, there IS a difference between temperature (as in Hot) and average heating power (as in watts). [A burning match may be hotter than the electric stove element, but does not have the heat energy to boil a pot of water.] The fine print on many resistor data sheets shows that a lowly 1/2 watt resistor may have a temmperature rise of 70C...thats 90C and too hot to hold on to with bare hands when the room temp is 20C. "Rated" operating temp may indeed be 100C. If you want it to run cooler, you need to increase the case size to add more surface radiating area. i.e. - a bigger watt rated resistor. But yeah, it should not get hot quickly, or begin to smoke. That's a good sign that there is something wrong.

re - The R-C network
The voltage spikes you see at the track without the R-C are created by the DCC booster attempting to rapidly change the polarity of voltage stored on rail capacitance when fed by wires with a significant amount of inductance. Using your example, the remote rail to rail capacitance and layout wiring inductance of the DCC bus form a resonant L-C with a 1/2 period of ~200 nano-seconds (0.2 uSec). Current must flow in to (or out of) the rail capacitance in order to flip rail voltage polarity. The faster you attempt to flip it (power FET switching speed inside the booster) the more peak current must flow to the rail capacitance. This causes higher and higher current and subsequent voltage ringing in the wiring-Rail L-C. You can calculate the energy content of the actual ringing which represents energy flowing back and forth between the L of the wiring and C of the track...only if you know the actual circuit values...But it is rather small. However, there is little or no energy absorbing damping to keep the L-C 'tank' circuit from voltage 'peaking' and ringing. Adding an external resistor to the L-C circuit will provide damping to the point where it limits the voltage peaking effects and doesn't ring any more. But we don't want to add series resistance as that would reduce power available to our locos on the track. A parallel resistor can be used, but it will also see the rms value of all the DCC track voltage as well and heat up accordingly (14.25 volts continuous, squared, divided by R ohms). A capacitor of the right value placed in series with a damping Resistor in parallel with the track will the reduce resistor dissipation if sized so that the cap will charge up to the full track voltage on each voltage transition but much more slowly than the ringing we are trying to suppress. [100 ohms X 0.1 uFD = 10 uSec, maximum DCC transition time is ~2uSec (I'm sure someone will correct me), NCE/Sys-1 boosters switch in ~0.6 uSec. Digitraxx boosters switch much more slowly and don't generate the ringing problem.]

In brief, current flows through the resistor only during the DCC voltage polarity transitions, about 4 X 8,000 times per second. And the 40 volt spikes aren't there when you connect the R-C, becauses the ringing energy of the track L-C (a much smaller value of C) is being damped and absorbed by the resistor. But whatever energy flows in and out of the external added capacitor must also flow through the series resistor, causing it to heat up. The extreme and easy to understand case is when you fully discharge a capacitor into a resistor. The entire energy content is 1/2 X C X V^2. If you discharge it only once, the energy content is so many joules or watt-seconds, all absorbed by the resistor (regardless of the ohms value of the resistor). If you do it repeatedly, multiply the watt-seconds of each event by the number of events per second to get the average watts being dissipated as heat in the resistor. But a resistor will heat up with current flow in either direction...so charging the capacitor up to a certain voltage will cause a series resistor to heat up with the same energy content as discharging it. When you apply the DCC square wave of voltage to the series R-C network and assume that the capacitor becomes fully charged up to the the appropriate DCC level (the peak voltage is gone on your scope traces, right? That's why we put in the R-C) the cap charges and discharges 2 times each per DCC cycle [0 to + to 0 to - to 0] So the resistor sees 1/2 X C X V^2 X 4 X the number of times per second (~8,000) watts. Notice that you don't want to make the external capacitor too large, as it does determine resistor watts.

An R-C network used in this manner is referred to as a "snubber".
You need to locate it out where the ringing to be suppressed is occurring...At the end of the DCC bus wires. [i.e.- Where the inductance of the wiring meets the capacitance of the track.] It does no good if placed at the booster terminals.

DonV