Manual Transition

I don't quite get the idea of manual transition. Could someone please explain how this works, versus the modern alternative of auto transition? It seems only to be present on first gen hood units such as GP7 and GP9, etc... I'm a juice-jack fan. And don't even get me started on how all the cocks and levers work in steamer. I'm 26, so manual transition is foreign enough to me. So when I looked at the controls of the ATSF 2-6-2 in the city park yesterday, I got a headache and went to the bar... However, I'm a pretty mechanically-inclined guy, so I pick up pretty quick if you have a cursory explanation.

Mods note: edited two posts, combined into one.

A rather simple question, with a rather complex answer. I will post some info, and as always, you can use the "search" feature, as this topic has been explained in several parts of the RR.Net forum, already. Enjoy:

Propulsion system operation
As previously explained, the locomotive's control system is designed so that the MG output for any given prime mover speed will be constant and ideally will be exactly matched to the maximum horsepower produced by the prime mover at that RPM. Due to the innate characteristics of traction motors, as well as the way in which the motors are connected to the MG, the generator will produce high current and low voltage at low locomotive speeds, gradually changing to low current and high voltage as the locomotive accelerates. Therefore the net power produced by the locomotive will remain substantially constant for any given throttle setting.


Typical main generator constant power curve at "notch 8".In older designs, the prime mover's governor and a companion device, the load regulator (LR), play a central role in the control system. The governor has two external inputs: requested engine speed, determined by the engineer's throttle setting, and actual engine speed (feedback). The governor has two external control outputs: fuel injector setting, which determines the engine fuel rate, and LR position, which affects MG excitation. The governor also incorporates a separate overspeed protective mechanism that will immediately cut off the fuel supply to the injectors and sound an alarm in the cab in the event the prime mover exceeds a defined RPM. It should be noted that not all of these inputs and outputs are necessarily electrical.

The LR is essentially a large potentiometer that controls the MG power output by varying its field excitation and hence the degree of loading applied to the engine. The LR's job is relatively complex, owing to the fact that although the prime mover's power output is somewhat proportional to RPM, the MG's output is not (which characteristic was not correctly handled by the Ward Leonard elevator drive system that was initially tried in early locomotives).

As the load on the engine changes, its rotational speed will also tend to change. This is detected by the governor via a change in the engine speed feedback signal. The net effect is to adjust both the fuel rate and the LR position. Therefore, the prime mover RPM and torque will remain relatively constant for any given throttle setting, regardless of actual road speed.

In newer designs controlled by a “traction computer,” each engine speed step is allotted an appropriate power output, or “kW reference”, in software. The computer compares this value with actual MG power output, or “kW feedback”, calculated from traction motor current and MG voltage feedback values. The computer adjusts the feedback value to match the reference value by controlling the excitation of the MG, as described above. The governor still has control of engine speed, but the LR no longer plays a central role in this type of control system. However, the LR is retained as a “back-up” in case of engine overload. Modern locomotives fitted with electronic fuel injection (EFI) may have no governor, however a “virtual” LR is retained.

Traction motor performance is controlled either by varying the DC voltage output of the MG, for DC motors, or by varying the frequency and voltage output of the VVVF for AC motors. With DC motors, various connection combinations are utilized to adapt the drive to varying operating conditions.

At standstill, DC traction motors are connected across the MG in a series-wound configuration (that is, the motor field windings are connected in series with the motor armature windings), generally with two motors in series with each other. In this configuration, MG output is initially low voltage/high current, often in excess of 1000 amperes per motor at full power. When the locomotive is at or near standstill current flow will be limited only by the DC resistance of the motor windings and interconnecting circuitry, as well as the capacity of the MG itself. Torque in a series-wound motor is approximately proportional to the square of the current. Hence, the traction motors will produce their highest torque, causing the locomotive to develop maximum tractive effort, enabling it to overcome the inertia of the train. This effect is analogous to what happens in an automobile automatic transmission at start-up, where it is in first gear and thus producing maximum torque multiplication.

As the locomotive accelerates, the now-rotating motor armatures will start to generate a counter-electromotive force (back EMF, meaning the motors are also trying to act as generators), which will oppose the output of the MG and cause traction motor current to decrease. MG voltage will correspondingly increase in an attempt to maintain motor power, but will eventually reach a plateau. At this point, the locomotive will essentially cease to accelerate, unless on a downgrade. Since this plateau will usually be reached at a speed substantially less than the maximum that may be desired, something must be done to change the drive characteristics to allow continued acceleration. This change is referred to as "transition," a process that is analogous to shifting gears in an automobile.

Transition methods include:

Changing the traction motor connections from series or series/parallel to parallel. In parallel mode, the back EMF developed by the motors will not increase as rapidly as in series operation, as the now-parallel field will develop a magnetic flux strength that is independent of armature current. Therefore, armature current can continue to increase without causing an increase in field current, preventing the latter from changing the rate at which back EMF can increase. In some cases, resistance may be introduced in series with the field winding to accentuate this effect. This type of transition is known as "motor transition."
Reducing motor field current while operating in series mode by placing resistance in parallel with the field. This has the effect of increasing the armature current, producing a corresponding increase in motor torque and speed. This method is variously termed "field shunting," "field diverting" or "weak fielding."
Reconnecting the two separate internal MG stator windings from parallel to series to increase the output voltage. This is called "generator transition."
In older locomotives, it was necessary for the engineer to manually execute transition by use of a separate control. As an aid to performing transition at the right time, the load meter (an indicator that informs the engineer on how much current is being drawn by the traction motors) was calibrated to indicate at which points forward or backward transition should take place. Automatic transition was subsequently developed to produce better operating efficiency, and to protect the MG and traction motors from overloading due to improper transition.

In a very simple way, manual transition is exactly that. A selector lever on the control stand is ratched up, or down (actually left or right), to manually move from series, to series parallel to full parallel or return. This can also be controlled, on some older locos, by shutting off the throttle, once you drop below a pre-determined speed, then opening the throttle back up (this is known as manual reverse transitioning). EMD, GE and Alco for the most part, used an electric type transition/throttle system, allowing for MUing the locos with other builders units. Baldwins were somewhat unique, in having air operated equipment, and were able to MU with other air operated equipment only. Alcos were built this way as well, until the Century Series came along. The transitioning speed in an auto loco was pre-determined in the loco speed settings, in the controlling wiring of the loco. Set at very slightly differing speeds, for a very "generalized" definition, locos are set to transition between 23 and 26 mph automatically, under a pre-determined load regulator setting. This prevents all the locos in the consist from making transition simultaneously, and tearing the train apart, from the sudden surge in tractive loss, then effort, as the actual transition is made. Let those with credible info add to the discussion here. Thanks, and hope this helps a bit. Regards, Dave

An engineer I work with said he had 5 SD40-2's go from series to parallel at the same time resulting in a broken knuckle about 4 or 5 cars back. I know this isnt supposed to happen but is it possible? I know when I have a few GE's on the point if you have a SD40 back there you can feel it making transition if you are in a high throttle position.

Tadman wrote:...versus the modern alternative of auto transition?

The modern alternative is no transition-- on four-motor units anyway. On C-Cs the alternator/generator might be built in two halves that can be "transitioned", but everything since the SD50 has had traction motors in permanent parallel.

That doesn't explain the very sudden, very violent sounding/feeling transition that occurs, on SD-50's and SD-60's, where the loco doesn't drop it's load, then slowly builds it back up, but suddenly the loco "pops" into transition, with a bang, and the revs of the engine don't alter significantly, nor is there a drop in TE or it's accompanying drop in speed, as the loco unloads........... The "kick in the pants" can be felt clearly in the seat, as the loco quite suddenly transitions, and maintains a full load while doing it.
GN asked "can it happen"? Yes, it can, does and has. The reasoning behind the staggering of the transition speed by a few tenths of a MPH between locos, over a few MPH total range, is supposed to eliminate that. Imagine starting a train, of grain, and four SD-40-2's are pulling in the 8th, approaching 24 MPH. Suddenly, they all transition at the same time. All four locos simultaneously drop their load, as transition occurs. The dropping of load causes the cars to roll against the power, bringing in a good piece of slack. Those ponies suddenly come back to life, and 4 big locos then get a solid yank on the train, as they load their way back up to 900 amps. that slack that reached 10-15 cars into the train, is suddenly ripped out. Anything that is even microscopically compromised is now subjected to a violent DB pull in excess of the continuous DB limits for that piece of equipment. Ker-POW goes the weakest link. The locos are pre-programed with a specific transition speed, withing a very tight range to eliminate that from happening.
The other problem, with locos having the same, or very close transition speed, has happened to me. Pulling a train on the Riverline, for ConRail, I had the pleasure(?) of a solid GE consist, of 4 B units. As my train reached transition speed, the locos all seemed to transition at the same time. The sudden loss of tractive effort immediately reduced the speed, and the locos made reverse transition after failing to make a successful forward transition. As I got to that speed again, the same thing happened over and over. It took cresting a grade, before I could build the speed to overcome the locos, that "couldn't" transition, due to the weight of the train, and the amount of time the transition took. It made for a very long, and very frustrating trip.

Interesting. Thanks, Dave and all.

I have notices that on SD40-2s, they go into parallel at about 23 mph. They go into series if the speed drops below 18. If you are above 23, and go to notch 1, then they do backward transition. They will go back to parallel if the throttle is in notch 5.
I have noticed that some units restore power slowly, after transition. The ones that jump right back up are the ones that might cause the train to come apart. I advise throttling back, if you have one that does rough transition, just before you expect it to happen.

Engineer Spike wrote:I have notices that on SD40-2s, they go into parallel at about 23 mph. They go into series if the speed drops below 18. If you are above 23, and go to notch 1, then they do backward transition. They will go back to parallel if the throttle is in notch 5.
I have noticed that some units restore power slowly, after transition. The ones that jump right back up are the ones that might cause the train to come apart. I advise throttling back, if you have one that does rough transition, just before you expect it to happen.

The SD40-2 units come with two types of Rate Control Modules. Some units have the RC11 card and some units have the RC12 card. The RC11 card is a much faster card then the RC12. Lets say you would do a throttle sweep, The RC11 will take 11 to 16 seconds to build 50 VDC at the RC11 testpoints 13 to 14 to reach full power. On the other hand the RC12 will take 18 to 36 seconds and it is not uncommon for it to take as long as 44 to 46 seconds to reach full power. When you have got 50 VDC at the RC module pins 13 and 14 you have max out. So when the units restore power quicker then others, it is due to the RC card. Also when your unit makes transition, another thing happens, you bump the ORS Valve inside your governor and the load regulator will move toward minimum field and slowly build back toward maximimum field after the unit makes transition (to reduce excitation and drop power). The SD40-2 units also have a circuit called transition delay. Lets say you make transition at 23 mph and the load behind you is so great that it slows the train and you make backwards transition around 18 mph. The transition delay circuit should come into play and then allow the SD40-2 to go past 23 mph up to 28 mph before the unit will make transition again. I want to touch on another thing, the SD40-2 unit also has a resistor and a capacitor in the CDR circuit (Contactor Delay) which should hold the "S" Contactors in a split second so the generator circuit can decay some what, to cut down on flashing and arcing to make a smoother transition from Series to Parallel and save on contactor tips.
You will also notice a difference between the RC11 and RC12 in Dynamic Braking. With a RC11, you should reach 700 amps of Grid Current much faster then you would with the RC12.

but suddenly the loco "pops" into transition, with a bang, and the revs of the engine don't alter significantly, nor is there a drop in TE or it's accompanying drop in speed, as the loco unloads...........

To understand this, if you're into performance cars think of it as a board-shift!

Manual transition (EMD units) was phased out on the F-7. They came equipped with a toggle switch in the left electrical cabinet on engine room side of fireman side cab (over the shunt field contactor) door, where you could set that unit in manual or auto. Transition levers to operate trailing manual units were included on newer locomotives up to the SD-40-2's, if the railroad ordered them.

To avoid the "Snap slack action", one had to reduce the throttle to #6 position before the mechanical interlocks on the lever would allow you to move to it to or from #3 (parallel) position. Once one became accomplished at using manual transition you could reduce throttle to #6, operate transition lever over the hump from 2-3, and have throttle back in number 8 position without hearing any decrease in engine RPMs. This was a little hard on the power contactors, but they never felt anything in the caboose or diner.

I dug out my operators manual for F7s and it describes the four stages of transition. 1,series-parallel; both motors in each truck are in series and both trucks are in parallel. 2,series-parallel shunt; same connections but the fields are shunted with resistance. 3,parallel; all four motors are in parallel. 4,parallel-shunt.
As was stated before starting with the F7, automatic transition was built in to all new EMD engines. After placing the handle in the first position it was only used when older units were in the consist.
Its like changing gears to get into the best power range for your needs. On Septa engine 52, our SW1200, has four positions of transition; off, switch, series, and automatic. Switch allows the engine to load quickly for yard moves, series allows for maximum pulling power, and auto is for road use.

GN 599 wrote:An engineer I work with said he had 5 SD40-2's go from series to parallel at the same time resulting in a broken knuckle about 4 or 5 cars back. I know this isnt supposed to happen but is it possible? I know when I have a few GE's on the point if you have a SD40 back there you can feel it making transition if you are in a high throttle position.

I believe it, some GM's transition very rough, especially units with PTC(Positive Traction Control). They make transition(up or down) so hard that they will rip the train apart. You see that ammeter drop to zeo and just pray the train stays together. That is the one thing I like about the new SD70M-2's, they don't make transition nearly as violent.

That's because some of them just aren't making transition period

Getting back to the manual transition side of things, on manual transition equipped units such as E and F units which pins are used for transition? Also after looking through one of my SD40-2 manuals the SD40-2s could be equipped with a rotary snap switch on the control stand on special order by the railroad for manual transition control of units that require manual transition and has gotten me wondering becasue from what it seems like from reading the few E unit manuals I've got that E units have 3 levels of transition while, from what I've heard and read, F units have 4 levels of transition and also has me asking the question of why do the E units have 3 transition levels while Fs have 4?