Notched power control on locomotive

[oltmannd]

The signal for the MU wires comes from the control stand.  The throttle lever is connected to cams with switches.  That is standard. What the locomotive does with that information can vary...

[Overmod]

Erik_Mag

PWM for control makes sense as the duty cycle (i.e.pulse width) can be recovered with all sorts of voltage drop from the lead locomotive to the final trailing locomotive.

There was a recent thread over on the MR forum which is highly amusing in this context, regarding whether certain modern full-scale locomotives use DCC...

[Erik_Mag]

PWM for control makes sense as the duty cycle (i.e.pulse width) can be recovered with all sorts of voltage drop from the lead locomotive to the final trailing locomotive.

[Pneudyne]

Overmod

In either case, the diesel engine rpm would be governed within very close limits, so "locomotive road speed" control would have to be done with variable field... which would inherently be stepless PROVIDED there were some way to trainline a proportional signal to trailing units.  I am not familiar with how that would be (or was) done.

 

 

EMD had developed a notchless power control option for its GM6C and GM10B prototype electric locomotives, so presumably had worked out its own method for relaying electrically a continuously variable signal to trailing units.  Presumably the same could have been used for diesels, although that would probably make them incompatible with standard eight notch diesels for MU purposes.  On the other hand, the electrics had the option of eight notch control for diesel MU compatibility.

 

In worldwide electric locomotive practice since the mid-1960s, one method used for conveying a continuously variable control signal to trailing units in MU was to convert the variable voltage into a PWM signal that was sent down the jumper cable (possibly on a screened wire) and decoded locally in each trailing unit.

 

 

Cheers,

[Pneudyne]

Pneudyne

 

 

 

 

I have seen it said (right now I can’t find where) that the E-44 fleet was equipped to MU with diesels, but I have not been able to confirm this.  

 

That was is Cunningham (*), on p.67.  In describing the PRR E-44. It was said:

 

‘Dual controls for bidirectional operation were included, as were dynamic braking and a capability of multiple operation with diesel locomotives equipped with standard control systems.’

 

 

(*)  Joe Cunnningham; New Haven EP-5 Jets; NJ International, 1991; ISBN 0-934088-26-8

 

 

Cheers,

[BaltACD]

blue streak 1

Know that some AEMs had notchced and others called "P?"  Riding clocker Engineer complained train sluggist with 2 AEMs.  RFE went back at stop and cut second unit out.  Performance better with just the lead unit.  Some one who understands it please explain.

Sure the RFE cut the 2nd unit out, or did he correct a improper connection.  I know you can only go by what he said - but the truth has been know to be a rare commodity for some individuals.

[blue streak 1]

Know that some AEMs had notchced and others called "P?"  Riding clocker Engineer complained train sluggist with 2 AEMs.  RFE went back at stop and cut second unit out.  Performance better with just the lead unit.  Some one who understands it please explain.

None of those problems with Diesels if using an electric  motor(s). The ICTs certainly will work well under wire. How are the QSKs handling these known lags?

[oltmannd]

I'm going to try to add and or simplify.

8 notches?  Three solenoids in the Woodward Flyball Governor.  2^3 = 8 distinct engine speeds.  (Purists will note that there are actually 4 solenoids in there - the 4th for shutdown from the control stand.  Later used to get low idle and low-low idle speeds).  Three (four!) wires plus ground for MUing.

U25s and their funky 16 notches?  My recollection was it gave the engineer some control over turbo lag and engine smoke.  He could do the half notch and get the next engine speed without the accompanying load.  Let the engine and turbo spool up a bit before putting the full juice to it.  Most U25s I ran across in the late 1970s had the half notches wired out.  The controller would notch, but it did nothing.

A basic difference between EMD (up through Dash 2) and GEs was how HP was regulated.

EMD let the governor have control using the load regulator.  Put it in notch 8 and the engine would come up to speed and increase the fuel rack to a set point, using the load regulator (controlled by the governor) to get the fuel rack setting to the set point.

GEs run with load regulator at max.  Excitation is regulated by the locomotive's control system.  

EMD will balance at a set fuel rack point.  GE will float the rack to get the HP.

Consequently, a GE would always try to do it's rated HP, while an EMD might do more or less depending on fuel temp (and other things).

This difference allowed GE to do all sorts of whacky things over the years.

Some U boats had 1-5-8 speed schedules. (but 8 notches of power).  When you notch out from 1 to 2, the engine goes to notch 5 speed, then when you notch from 5 to 6, you get notch 8 speed.

Later, there was "skip 3 double 6" on some Dash 7s.  Notching 1 thru 8 got you  1, 2, 4, 5, 6, 6, 7, 8 engine speed.

These were all about smoke and trying to minimize turbo lag.

If you've ever wiped a throttle on a CHEC Dash 7, you know it takes 80 seconds to get to full load with more than half coming on in the last 20 seconds.

[Overmod]

mvlandsw

Amtrak's F40PH's had a notchless throttle when the diesel engine was running at its maximum rpm to supply headend power to the train. They reverted to a standard 8 notch throttle when not supplying headend power.

Clarify something for me.  The F40PH had two distinct methods of HEP.  One involved running the diesel engine at 843rpm (a precise number, related to the number of poles in the special HEP alternator and the HEP line frequency).  The other involved running the engine at 720rpm to get 60Hz off the traction alternator, probably the same as the U34CH.

In either case, the diesel engine rpm would be governed within very close limits, so "locomotive road speed" control would have to be done with variable field... which would inherently be stepless PROVIDED there were some way to trainline a proportional signal to trailing units.  I am not familiar with how that would be (or was) done.

[Pneudyne]

Re the GE 16-notch control, I have not found a single source that describes this in detail.  Thus what follows is constructed from several sources, along with some deductions, so might not be correct in all respects.

 

It was used, with three-field excitation, on the U25B/C, U28B/C and U50 models, and with electronic/PWM excitation on the U30B/C and early U33B/C models, and also on the Alco C855. 

 

With a 16-notch locomotive leading a consist, then any other 16-notch locomotives trailing in that consist also operated in the 16-notch mode, even if they were separated from the lead locomotive by one or more eight-notch locomotives.

 

When a 16-notch locomotive was trailing in a consist led by a standard eight-notch locomotive, then it defaulted to the eight notch mode.

 

As already said, the engine had the customary eight operating speeds, which notchwise were shared as follows:

 

Notches ½, 1, 1½          1st speed

Notches 2, 2½               2^nd speed

Notches 3, 3½               3^rd speed

Notches 4, 4½               4^th speed

Notches 5, 5½               5^th speed

Notches 6, 6½               6^th speed

Notches 7, 7½               7^th speed

Notch 8                         8^th speed

 

Basic main generator excitation was controlled in 16 steps or eight steps according to mode, and by separate means for each mode.  Thus the whole-numbered notches in 16-notch mode did not have to align with their counterparts in the eight-notch mode.

 

That gave 16 or eight basic main generator curves according to mode.  And the basic curves determined the tractive effort increments available along the standstill (IR) line.

 

In either mode, there were eight constant power hyperbolic curves, one for each engine speed, imposed by the governor operated load control system, and superimposed upon the basic main generator curves.  Thus in the 16-notch mode, there was some sharing of these curve portions, at least with the three-field excitation system.

 

But the electronic/PWM excitation system was also able to construct close approximations to these constant power curves, leaving the load regulator to do the final trim (for maximum power notches at any engine speed) and to provide an override in the event of an engine malfunction.  This same capability was also used in later eight-notch locomotives that used fewer than eight engine speeds.  Thus for the U30B/C and U33B/C, each of the sixteen notches had its own constant power curve sections.

 

Now to the excitation means.  In 16-notch mode, the leading locomotive controlled main generator excitation for all other 16-notch locomotives in a consist.  One trainwire (I don’t know which #), known as SN, was assigned to signal (when at +74 volts) to any trailing 16-notch locomotives that they were being led by one of their own kind, and so to take their excitation current or excitation control voltage from the XB (dynamic brake excitation) trainwire, rather than derive it locally.  Absent +74 V on the SN trainwire, the trails reverted to eight-notch operation, with excitation controlled locally from a resistor matrix switched by the throttle control trainwires (AV, BV, CV and DV).

 

The 16-notch locomotives had a commutated resistor matrix in their master controllers (KC99 or KC102) that provided 16 voltage graduations for motoring excitation control, and with reasonable current capacity.  This was fed to the XB trainwire.  In locomotives with three-field excitation, the XB trainwire was used as the source of exciter battery field current.  In locomotives with electronic/PWM excitation control, the XB trainwire was used as the source of a control voltage input to the electronic system.

 

A consequence of this arrangement was that the 16-notch sequence did not have to be exactly interleaved with the eight-notch sequence.  But I do not know whether or not that was the case.  Nonetheless, one may envisage a situation where say notch ½ in the 16-notch sequence was not far removed from notch 1 in the eight notch sequence, giving some “misalignment” at the low end.  If so, I’d also guess that alignment was achieved by around the 3^rd or 4^th notch, and maintained thereafter.

 

Dynamic braking control was similarly arranged.  In 16-notch mode, the leading master controller commutated resistor supplied voltage and current to the XB trainwire, which was then used by any trailing 16-notch locomotives for their exciter battery fields directly (three-field) or for exciter battery field control (electronic/PWM).  In eight-notch mode, the three-field locomotives used the XB trainwire voltage to control a follower relay which, via the governor overriding solenoid, in turn controlled the load regulator rheostat.  This way, they did not load the XB trainwire, recognizing that some potential wire (PW) dynamic brake control systems used on eight-notch locomotives had quite limited current delivery capability.

 

I am not sure, but I think the baseline XB trainwire current capability of the early GE 16-notch master controllers was something like 10 amps, enough also for a reasonable level of field loop dynamic braking if required, but that 15 amps capacity, for full field loop capability, was optional.  The 16 master controller commutated rheostat voltage steps were different for the motoring and dynamic brake cases.

 

My understanding is that GE’s main reason for adopting 16-notch control for the U25B was to provide finer graduation of starting and low speed tractive effort, in recognition of the fact that it had a higher power output per axle (and so per ton of adhesive weight) than previous US domestic diesel-electric models,  This was based upon its experience with GTELs (which had 20-notch control) and various electric locomotives (with multinotch control) that also had relatively high power-per-axle numbers.  In this context, overlapping of the curves in their constant power sections at higher speeds (as would happen with three-field excitation) would not have been seen as problematical.

 

A full set of main generator curves (both 16- and eight-notch modes) for say the U25B would well illustrate the foregoing, but such appear to be unobtainium.

 

 

As Overmod has outpointed, 15 rather than 16 notches would seem to be more logical for full interleaving with the eight notch system, at least if the eight notches were the same for both sequences, which was not necessarily so for the GE 16-notch case.  In fact such a 15-notch system was used for the Alco DH643 diesel-hydraulic.  With diesel-hydraulics, apart from part-filling the 1^st converter at low speeds, engine speed is the only readily available power control parameter, since the load imposed by torque converters is simply speed dependent, following the propellor law (cubic curve).  (Although let’s acknowledge that in the 1950s, Krupp offered a converter with variable pitch impellor blades, which could be adjusted by a load control governor not unlike the kind used for controllable pitch propellors in the marine world.)

 

Thus, 15 notches required 15 engine speeds.  The Woodward PG governor with its standard electro-hydraulic speed control could in fact be configured to provide 15 speeds.  (Four binary solenoids, A, B, C and D means 16 positions, of which one was shutdown, so 15 running speeds.)  But this would have given a sequence that was not directly compatible with the eight-notch sequence.  So in the DH643 case, the A, B, C and D solenoids were left in their usual state, to provide eight speeds.  Then the overriding solenoid (ORS), not required for a diesel-hydraulic, was purloined to provide an additional speed control facility, evidently used to provide the seven intermediate speeds.  Logically, a spare MU trainwire would have been assigned to this function.  The motoring notches were 2 through 16 on the GE master controller (KC102), notch 1 being a “dead” position.  Notches 2, 4, 6…16 corresponded to the regular eight notches, with 3, 5,....15 being the seven interleaving positions.  All 16 notch positions on the master controller were used for hydrodynamic/dynamic braking control.

 

Thus the DH643 could interwork with eight-notch diesel-electrics.  If coupled to say a GE U25B, each would see the other as an eight-notch locomotive.  (Although given their master controllers, 15-notch interworking would have been possible with a little extra complication.)

 

In the diesel-hydraulic case the reason for using multinotch controls was empirically derived.  Both DB, Germany and BR, UK found that the early six or seven notch controls used in the 1950s were just too coarse in practice.  Thus there was a shift to 15 or 16 notch and continuously variable pneumatic systems.  JNR I think moved from seven to 14 notches for the production DD51 class.

 

 

Cheers,

[Pneudyne]

Overmod

Pneudyne in particular will weigh in here, but ISTR the E-44s had 26 tap-changing notches, and may have been set up to have 'two intermediate notches' between 8 positions that would have made them nominally diesel-MU compatible.  (There are pictures of E44s coupled to GG1s, but every one I've seen has been with the trailing type 'dead-in-train' with the pans down.)

 

The E-44 had 29 tap-changing motoring notches, and 17 dynamic braking notches.

 

I have seen it said (right now I can’t find where) that the E-44 fleet was equipped to MU with diesels, but I have not been able to confirm this.  If so, one might expect an additional MU jumper socket at each end, but none of the pictures I have seen show more than the paired electric jumper sockets at each side.

 

Nonetheless, had that mixed MU capability actually existed, one may derive the required notch mapping by indirect means, from the work that GE did for the DoT in modifying one E44 to incorporate chopper control.  The modified E44, referred to as an E44/E60, had a standard eight notch controller (as I think did the Amtrak E60CP) that conferred MU compatibility with diesel locomotives.  But the E44/E60 was intended to be used in MU Trail behind an unmodified E-44.  The notch-mapping matrix for this was given as:

 

Standard E-44 notch range mapped to E44/E60 notch in each case

 

Motoring:

 

1,2                1

3,4                2

5,6                3

7,8                4

9–12             5

13-16            6

17-20            7

21-29            8

 

Braking:

 

1,2                1

3,4                2

5,6                3

7-9                4

10,11            5

12,13            6

14,15            7

16,17            8

 

Logically this map could also have been used for the E-44 plus diesel case. 

 

I don’t think that the E-44 was set up for backward MU compatibility with earlier PRR electric types.  The E2b prototypes though were setup to MU with the P5a fleet.  This is probably why they had 32 volt rather than the expected 74 volt auxiliary electrical systems.  Differing control voltages is not a barrier to mixed MU working – and it has been done - but typically it does require voltage conversion relay sets both ways.

 

 

Cheers,

[mvlandsw]

Amtrak's F40PH's had a notchless throttle when the diesel engine was running at its maximum rpm to supply headend power to the train. They reverted to a standard 8 notch throttle when not supplying headend power.

[jeffhergert]

Erik_Mag

The GE 16 notch throttle only had 8 engine speeds, but with each speed having two generator load settings.

 

The newer GE's (Wabtech's now I guess.) still have that half notch setting, but only when the Trip Optimizer (Energy Management System with auto throttle/dynamic brake control) is engaged.  A human engineer only has the 8 notches. 

I believe only the lead engine is so enabled.  Any trailing GE's are run only in the 8 notches.

Jeff      

[Overmod]

It may help for you to read up on Frank Sprague's design of MU control, and modified Ward-Leonard control of gas-electrics, as a background for how diesel-electric locomotives with AAR 8-notch control are set up.  You should also review previous posts here about series/parallel and EMD's later use of permanent parallel with stages of field weakening.

The eight notches basically control the rotational speed of the diesel engine, with a system of solenoids controlling the engine speed governor to within a few rpm.  As load on the engine at a particular speed increases or decreases (by varying the field of the traction alternator/generator) the governor automatically adjusts fuel flow to the engine so it makes greater or less power at that rpm.

This is not intended to work like the accelerator on a road vehicle: when the additional fuel 'maxes out', the train will accelerate up to where the train resistance (e.g. approximated using the Davis formula) balances the power delivered at the railhead by the traction motors.  To 'go faster' you'll select a higher notch, where the engine will have more power strokes per minute to be 'maxed out' in fuel feed, and therefore produce more power.  (Note that the amount of exhaust volume determines the effectiveness of turbocharging, and this is usually nonlinear notch-to-notch, which is something to be aware of).

The electrical load varies with applied field, as noted above, and this is done automatically rather than with an additional power control.  The GE 'fifteen-notch' throttles modulated the field differently to give the effect of the 'half-notches'; the system fundamentally defaulted to AAR 8-notch when anything but a 15-notch GE was in the power consist.

This is very different from the 'throttle' control used on PRR electrics.  The GG1 famously had a 22-notch controller that worked by tap-changing -- the highest 5 positions were blankable, to be used only in passenger service (or as someone on TrainOrders noted, when a freight crew wanted to get home in a hurry).  Pneudyne in particular will weigh in here, but ISTR the E-44s had 26 tap-changing notches, and may have been set up to have 'two intermediate notches' between 8 positions that would have made them nominally diesel-MU compatible.  (There are pictures of E44s coupled to GG1s, but every one I've seen has been with the trailing type 'dead-in-train' with the pans down.)

 

[CSSHEGEWISCH]

Perry Babin

Are the 8 notches the only points on the controller where power changes, like an electric motor on a fan where there are only a finite number of speeds or is it a variable control where you could put it between notches to vary the power slightly (less than it would at the next fixed notch)?

 

Early Baldwin locomotives (and others) were equipped with air throttles that acted as a continuous variable speed throttle.

[Erik_Mag]

The GE 16 notch throttle only had 8 engine speeds, but with each speed having two generator load settings.

[Pneudyne]

There are a couple of earlier threads that provide some background information, namely:

 

 

‘diesel notch power settings’, at:  https://cs.trains.com/trn/f/741/t/283879.aspx?page=1;

 

and:

 

‘Locomotive MU systems’, at:  https://cs.trains.com/trn/f/741/t/291088.aspx?page=1.

 

 

Cheers,

[BaltACD]

The PRR GG-1 electrics had 32 notch throttles, I believe.

[tree68]

Perry Babin

Are the 8 notches the only points on the controller where power changes, like an electric motor on a fan where there are only a finite number of speeds or is it a variable control where you could put it between notches to vary the power slightly (less than it would at the next fixed notch)?

Short answer, yes.  At one time, I think it was GE that came out with a 16 notch throttle.  If connected via MU to an 8 notch locomotive, the 8 notch loco simply ignored the 'in between' notches.

The key here is the 8 notches require only a few wires between locomotives (via the MU cable).  Think octal/base 8.

Changing the throttle varies both the electrical power to the traction motors and the Diesel prime mover.

It's always satisfying to have a train "settle in" without having to vary the throttle.

[Perry Babin]

Are the 8 notches the only points on the controller where power changes, like an electric motor on a fan where there are only a finite number of speeds or is it a variable control where you could put it between notches to vary the power slightly (less than it would at the next fixed notch)?