--- In digitrax@yahoogroups.com, Don Crano wrote: Let's see if this helps a little. First we have to assume we understand how the DCC signal/power is made. Often called a DC bi-polar square wave. A fancy way to describe how the DCC power at the rails is made. It is actually a DC source being pulsed to two poles around a center voltage of zero volts, thus bi-polar. This generates an AC square waveform across the rails, and as typical this results in an average voltage of '0' volts. The average voltage is the sum of the instantaneous voltages in a half-cycle waveshape, divided by the number of instantaneous voltages. This is a very common method of generating an AC current, such as small inverters used for AC power from a DC source. Small ones used for mobile use in automobiles, etc, to larger ones used for back up or solar to invert DC to AC. As an example. You can easily duplicate this with a common DC cab. Apply full throttle, then switch the reverse switch back and forth as fast as you can. If you could switch it fast enough, the loco would never move, it would only vibrate at the frequency you could alternate the reverse switch. With an electronic switch or oscillator it is no problem to do this at a rate of 9kHz or so, such as DCC is switched. The end result non the less, by hand or by electronics, you create a square wave AC, by generating a bi-polar DC square wave. That is as long as the poles are switched at an equal rate, thus an average voltage of zero volts. If the poles are not switched at equal times, it is no longer an AC wave form. This is because if one pole or the other is a longer time period, the average voltage will move towards that pole. Now once we stretch the time period of one pole or the other [ plus or minus] a DC component is added. The pole that is stretched determines the direction of the loco, and the amount of time or stretch sets the speed via the added average voltage. And again if this increased average voltage is of a pulse voltage, the loco would move forward, and if a minus value the loco would move backwards. How does a DCC command station do all this: As per NMRA DCC St&Rp's: All timing measurements are done between zero volt crossings. In a "1" bit, the first and last part of a bit shall have the same duration, and that duration shall nominally be 58 uS, giving the bit a total duration of 116 uS. Digital Command Station components shall transmit "1" bits with the first and last parts each having a duration of between 55 and 61 uS. A Digital Decoder must accept bits whose first and last parts have a duration of between 52 and 64 uS, as a valid bit with the value of "1". In a "0" bit, the duration of the first and last parts of each transition shall nominally be greater than or equal to 100 uS. To keep the DC component of the total signal at zero average voltage, as with the "1" bits, the first and last part of the "0" bit are normally equal to one another. Digital Command Station components shall transmit "0" bits with each part of the bit having a duration of between 95 and 9900 uS with the total bit duration of the "0" bit not exceeding 12000 uS. A Digital Decoder must accept bits whose first or last parts have a duration of between 90 and 10000 uS as a valid bit with the value of "0". From this we can see several things we should note. First the '1' bits are pretty well locked in timing wise, that is no real variation in timing, and that a decoder should allow as much as +/- 6 uS variation from the nominal 58 uS of the command station timing. But more important is the '0' bit timing as related here to DC operations on DCC track. As noted above if we stretch the duration or timing of one side of the pulses or the other we add average voltage to the AC component. As noted here nominally >= 100 uS and equal to keep the average voltage at zero volts, but allowable from 95 to 9900 uS per half bit. This allowable timing variation is what allows what is called pulse stretching. As an example if we have '0' half bits, of say the plus [+] half at 9900 uS and the negative [-] half bits at 95 uS, this would result in a high positive [+] average voltage. The result is a DC motor would power a loco forward at some high rate of speed. Or inversely if we have the positive [+] half bits at 100 uS, and the negative [-] half bits at 200 uS. Now we have a negative average voltage that would in a DC motor that would power the loco in the reverse direction at a some what slower speed. Now what happens when we place a non-decoder equipped loco on the DCC rails: Placing a DC loco on DCC rails, will cause a singing [armature vibration] from the DCC bi-polar frequency as it sits idle. This will also cause some added heat. This is not typically damaging to the motor as long as the motor has a stall rating that exceeds the amplitude of the signal. And as long as the track voltage is =< +/-18V, the lower the better. Now once we get the loco moving, as noted above, this is done by what is called pulse stretching, here one half bit or the other of the '0' DCC bits are being stretched, this applies a DC component to the DCC power. Now the loco will move, speed dependant on how much the pulses are being stretched, and direction based on if the + or the - half bits are being stretched. At this point the motor is seeing both DC and DCC to some degree. But as more DC is being applied [pulses being stretched more] then the motor will start to act more typical of DC, the singing reduces and the motor cools back down. As we can see if a DC motor is going to be placed across DCC rails, make sure the track voltage is =< 18V typical is N = 12.5V, HO = 15V, Large Scale = 20V. And do not let a DC loco sit Idle on DCC rails, this is where the most heat will be built up. Not much of a problem if the loco is moving at much of any speed, the DC component is taking over. Notes of importance: The reason one should never let a DC motor sit idle on DCC rails is multiple. And include such as added heat, armature vibration and demagnetization of PM's. And none are good for prolonged motor life. Heat and vibration are detrimental to bearing life, and demagnetization to motor torque. None mean your motor will smoke on the spot, or even over time. It simple means your motors life is going to be shortend by some period of time. Most of us whom have had or have DC layouts have learned over time it is not a good idea to power reverse our locos at speed. The effects here are pretty much the same. A motor sitting idle is being reversed constantly, granted many times a second, in short bursts of full track voltage, but the result is still the same over a period of time. Or simply ask the Mfg'er of your 12Vdc motor how they feel about applying AC, square wave or not to their DC PM motor. Again as noted above, as the motor begins to move, stretched pulses are applying a DC component, and taking over the AC component. Of more concern is the core-less motors, not typically found in most loco's. But due to their design, they do not have the ability to dissipate heat as a common iron core motor does. Simply do not use a core-less motor across DCC rails. Note: core-less motors should be operated only with pure DC, or with a High-Frequency PWM output decoder of around 16k or better yet around 20+k. Pulse stretching also has a negative effect on system bandwidth. If we look at a non-stretched packet stream on the rails, '1' half bits at 58uS and '0' half bits at 100uS, then look at stretched '0' half bits at say 9000uS. It is not to hard to see that in any given period of time, the stretched pulse will result in less packets being able to be sent. This can, and usually does, result in notable system delays of decoder responses to commands. The more activity AKA more decoder command being made, such as clubs or large layouts with more operators at throttles, the more this will become noticable, even to the point of possible interference of proper operations. And last but certainly not least, we all know a typical decoder today has the ablity to operate on DC track, this is via automatic power conversion. CV29 bit 2. When bit 2 is OFF, we tell the decoder, if no DCC signal stop. When bit 2 is ON we tell the decoder to convert to any new source availible, via optional CV12, or typically to DC operations. We also know that from above a A Digital Decoder must accept bits whose first or last parts have a duration of between 90 and 10000 uS as a valid bit with the value of "0". So pulse stretching should not be a problem for a typical decoder. But what happens if power conversion is ON, CV29 bit 2 ON, but the decoder for what ever reason looses the DCC signal, this means most likely it has no idea what the packet timings are, so it does what it has been told to do, convert to DC, and will see the DC component across the rails from the stretched pulses. Typical result is what we call a runaway. At least till it can sync again with the DCC packets to it's address. All this is nothing more then attempt to explain the hows and whys, as well as some of the negative effects of using a DC loco on DCC rails. And not meant to say in any way it should not be used. I use it all the time. The first thing I do with any new loco, is the same as I used to with DC. Run it on the layout for a period of time without a decoder, both directions, but typically never idle for any more period of time then required. I make sure it is going to run as smooth as possible, and check for strange noise, once one gets used to the added sing from the AC component. Remember Always Have Fun and Enjoy!, Don Crano Akron, Oh NMRA #096211 Moderator Digitrax User Group