Thanks Randy.
Built another copy of the lighting circuit for my son's E7. Worked great on the test bench, but didn't light on the layout. Time to put the tools away for a bit - too tired to go looking for bad joints, or to clean the track again ;-)
Lindsay
That should be enough to block it. In fact if it's across the motor brush terminals already you shouldn't be having this problem. Normally it's recommended to remove such capacitors, but that's for DCC, as they interfere with the decoder's PWM drive and BEMF detection. For straight DC power, it should absorb the momentary spikes that happen at the loss of contact.
This is I suppose a downside to the more highly efficient motors - Stewart uses high efficiency Canon motors in their Baldwin switch locos, and when I tested one of mine on DC before installing a decoder, cutting off the power at full throttle allowed it to coast for a couple of feet - with the LED headlight fully lit!
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
Thanks Mark, I'm across the root cause theory. Clean track and less joints definitely help (i.e. the problem is worse on the steel sectional track sections than the nickel-silver flex track lengths).
There is a 0.1uF ceramic cap soldered to the brush holders. Does the type of cap make a significant difference? The only green cap I have in the box of bits of the same capacitance is physically huge by comparison, and won't fit inside the car body (Lima NSW 44 class aka Alco World Series) on the large "ring field" mech. Moving it to a clear space will reduce its effectiveness.
Cheers,
The reverse light flickering is caused by poor track contact. It is really caused by the motor when the power is instantaneously disconnected. This is due to back EMF (electromotive force). To understand this you must first understand that current through a wire generates a magnetic field, and a changing magnetic field generates a current through a wire. The coils in the motor are really a series of very long pieces of wire, and the current generates the magnetic field the motor uses to turn. If a coil is receiving a current, and the current is disconnected, then the magnetic field collapses. This collapsing magnetic field constitutes a changing magnetic field, and generates a reverse voltage in the coil (this is called back EMF). In a motor this would normally be offset by the changing magnetic field that occurs as the coil moves past the fixed magnet(s) in the motor, but not always (depending on the position of the coil when the power fails). Therefore we occasionally, when the power fails and the coil is in a certain position, get a reverse voltage applied to the circuit. This reverse voltage is enough to ever so briefly light the reverse light.
This occurrence is common to any directional lighting system (although it will be more apparent in ours due to its high efficiency and speed), and cures are usually easy. Firstly cleaning the track and wheels of the locomotive should fix the problem 99%. If it does not, then a small capacitor across the motor in the locomotive will provide a permanent fix. The capacitor does not have to be big, say a 0.047uF to 0.1 uF polyester capacitor (often called a "Greencap"). These are small and should be able to be tucked against the motor somewhere.
Mark.
¡ uʍop ǝpısdn sı ǝɹnʇɐuƃıs ʎɯ 'dlǝɥ
I found a nice simple circuit using the LM334Z programmable current source to drive a LED. I set this up with a pair of LEDs in series (as the chip could only supply about 10mA). Works great in forward direction, but gets the usual flicker problem in reverse due to track joints, dirty track / wheels etc.
I found a variation of the circuit with a reverse-biased rectifier diode across the LED and tried it, but it's no different, and I fail to see the point of it.
Then I started thinking about putting a cap in parallel with the LED. The theory (untested!) is that in forward direction the cap will take a short time to charge (to the forward voltage of the LED), but once charged will have no further effect (although it might reduce flicker in the forward direction, but I haven't observed this to be a problem). In the reverse direction, when there is a spike caused by loss of contact creating a momentary forward voltage, the cap should soak this up and prevent the LED from coming on ... assuming the duration of the forward voltage is < about five times the time constant to charge the cap (to the forward voltage of the LED).
Any thoughts? Am I way off the mark?
Lindsay,
Sydney, Australia