• EMD Field Loop Dynamic Brake Control

  • Discussion of Electro-Motive locomotive products and technology, past and present. Official web site can be found here: http://www.emdiesels.com/.
Discussion of Electro-Motive locomotive products and technology, past and present. Official web site can be found here: http://www.emdiesels.com/.

Moderator: GOLDEN-ARM

  by Pneudyne
Engineer Spike wrote:How difficult was the rewiring job to run the db control through the 27 pin jumper?
Another item from the previously-mentioned Staff book on the WP concerns this road’s GP-20 fleet. Staff noted that they were converted to potential-wire DB in a “very cheap and dirty way.” It was done by “trainlining the signal into the units, placing a few jumper wires on the right contactors, and hooking them up so that they operated in potential.”

Nevertheless, as the GP-20 would have had the same internal DB control system as the SD-24, it would have been amendable to conversion. In principle the DB control potentiometer could have provided an outgoing signal to the DB trainwire, and the load regulator positioner relay could have accepted an incoming signal from the DB trainwire.

So conversion of EMD units from field-loop to potential-wire was possible for the models from the SD-24 and GP-20 onwards, and the WP case shows that it was in fact done in practice. Conversion of earlier units, such as the GP-9 and its predecessors, would seem to have been a much more complex exercise, so may have been seen as not worth doing by some operators. By the time conversion was desirable, the units in question would have been approaching the end of the initial life in frontline service. Still, conversion should have been possible, and probably using standard parts. Rewiring to the export DB configuration of the period could have been one pathway.

  by Pneudyne
I need to correct my earlier posting on the Westinghouse implementation of field-loop FB.
Pneudyne wrote: From the above-mentioned schematic, WH evidently shunted the exciter battery field with a 1.2 ohm resistor during DB, and most of the field loop current would have gone via the resistor. Let’s say, at a guess, that with 15 amps in the field loop, about 13 amps went via the shunt resistor, and about 2 amps via the exciter battery field. That equates to just under 16 volts across the shunt resistor (and the battery field), which seems not unrealistic considering that up to four units could be in the field-loop.
The exciter battery field shunt resistor during DB was 1.5 ohms, not 1.2 ohms.

1.2 ohms was the value of each of the three switchable resistors in the DB unit selector switch.

From that one may see that the expected resistance of each “battery field” in the field loop was 1.2 ohms. With four units in the loop, all resistors in the unit selector switch were bypassed, and the total loop resistance was 4.8 ohms. At 15 amps loop current, the total voltage drop was thus 72 volts. 1.2 ohms of resistance was switched in at the unit selector for each “absent” unit in a consist to maintain the total loop resistance at 4.8 ohms. For a single unit, all three 1.2 ohms sections were switched in.

In the WH case, we can back-calculate that the exciter battery field nominal resistance was 6 ohms. In parallel with the 1.5 ohms of the shunt resistor that provided the expected per-unit loop number of 1.2 ohms. At 15 amps maximum loop current, the exciter battery field current would have been 3 amps.

  by Pneudyne
I have looked more closely at how GE implemented field-loop DB on the U25B, at least as described in GEJ-3815, the GE educational manual for this model.

It is evident that field-loop DB was provided for trailing units only. A U25B leading a mixed consist would itself still be under potential-wire control. Similarly, one is left to infer that a trailing U25B in such a consist would also be under potential-wire control from the leading U25B, although it would provide pass through for the field-loop system.

There did not appear to be any provision for DB control on a U25B operating as a trailing unit in a mixed consist headed by a unit with only field loop DB.

The field-loop was fed from the XB (DBX) trainwire via two resistors in series, one of 2.92 ohms and the other of 1.16 ohms, for a total series resistance of 4.08 ohms. There was no apparent provision for varying this resistance according to the number of field loop units in trail. Thus even for a single trailing field-loop unit, the series resistance was higher than the 3.6 ohms normally provided by the unit selector switch. Assuming that 74 volts was available on the DBX trainwire at full DB, then the maximum loop currents would be 14 amps for a single trailing unit and 9.6 amps for three trailing units. On the face of it, full DB would not be available from the trailing units.

That might not have been quite as bad as it looked. The EMD field-loop system of the era essentially controlled TM field current in DB. TM armature current in DB was proportional to both field current and locomotive speed. As speed increased, lower field currents were needed for a given armature current. Thus over some of the speed range at least, full DB (maximum grid current) was available at DB handle settings, and so field-loop currents were well below maximum. On the other hand, the GE control system, by virtue of its heavy decompounding, effectively controlled TM armature current, so that full DB required more-or-less a full handle setting over the whole speed range. Whilst exciter battery field excitation was proportional to handle position, the net excitation was not, as the amount subtracted by the differential action was proportional to TM armature current. So the net excitation decreased with increasing speed.

Staff, in his WP book, provided a good perspective on this difference between EMD and GE:

“Finally, the problem of dynamic braking in consists of mixed power was always present. General Electric and Electro-Motive type of control differed in that the GEs possessed a grid current control, whereas the EMD was one of tapered or field current control. To achieve full dynamic braking on a GE, it was necessary to be in full braking position. This was not true of the EMD. If the consist was mixed, the units were not always producing the same braking effort.

“With a GE on the point, the rating was calculated on the GE unit which, had there not been automatic regulation on each trailing EMD unit, would have overexcited such units. If an EMD was on the point, the engineer went by the ammeter but he was not getting full braking out of the trailing GEs. In such situations, it might be necessary to “widen out” on the dynamic braking lever to produce the equivalent of No. 8 throttle position on the GEs.”

GE’s apparent initial choice of relatively simple field-loop configuration for its U25B may not have been acceptable to all operators. In one of his earlier excellent posts, Typewriters (Will Davis) noted the following comment from GEJ3816A:
Typewriters wrote:Some locomotives may be equipped with a multi-position dynamic braking MU Selector Switch. This switch is used to furnish full braking effort at all locomotive speeds when operating U25B locomotives and loop type locomotives in a consist with a U25B leading. Position the switch according to the number of Loop Braking units in the consist and as specified by railroad rules.
The selector switch mentioned would have included in its function the switching into the field loop the appropriate amount of resistance for the number of field-loop units in the consist. The rationale for it, namely availability of full braking effort at all speeds, was tacit acknowledgement that the simple circuit with fixed series resistance in the loop did not allow that.

  by Pneudyne
Typewriters wrote:ALCO Century series locomotives...

ALCO Products publication TP-447C - Operating Manual, Century Series Road Locomotives / D.C. Transmission, Rev. August 1967.

Page 46 - Dynamic Brake Unit Selector Switch (if used)

1. When operating all ALCO units in multiple:
a. Place unit selector switch on all units in No. 1 position.
b. Do not install field loop dynamic braking jumpers between units.
2. When operating ALCO units in multiple with units of other manufacture:
a. Place unit selector switch on all trailing units in No. 1 position.
b. Place unit selector switch on lead unit to correspond with number of units in consist.
c. Install field loop dynamic braking jumpers between all units.

We can see thus that ALCO also offered field loop compatibility on the Century series locomotives through at least the third quarter of 1967, just looking at one manual.
Interesting here is what the above operating instruction implies about unit conditioning in respect of dynamic braking controls.

For trailing Alco units, the unit selector switch was in the #1 position regardless of whether the lead unit was providing potential-wire or field-loop dynamic brake control. Thus such units were evidently able to accept either potential-wire or field-loop incoming DB signals, simply on the basis of whichever happened to arrive. Of course, field-loop signals would arrive only if the field-loop jumpers had been connected.

For a leading Alco unit, one would assume that the #1 position of the selector switch deactivated the field loop. On the other hand, the #2, #3 and #4 positions would have activated the field loop with the appropriate amount of series resistance, and I imagine would have deactivated the potential control – to avoid sending duplicate control signals - although local (to that unit) DB control was still by potential.

Another reasonable inference is that Alco units equipped with this kind of DB control could receive and act upon field-loop signals as well as provide field-loop commands. From item 2b above, one would be counting all units in a consist, not just the non-Alco ones.

Another “mixed-DB case” might have been the members of the UP GTEL fleet that were retrofitted to MU with trailing diesel units. As far as I know UP started the retrofit work in 1958, and in the early days at least, EMD GP-9 diesel units were used in conjunction with the GTELs. If this MU capability included DB control, then it surely would have been of the field-loop type because that is what the GP-9 was fitted with. Again as far as I know, the first batch of 4500 hp GTELs had Amplidyne control; certainly the equipment layout diagram showed that they had Amplidyne exciters. The second batch (“Veranda” type) were shown as having auxiliary alternators, so would seem to have had an early version of GE’s static control. Anyway, most likely both batches had GE’s potential-wire DB control. In the 8500 hp GTELs, DB control was evidently separate to motoring excitation control and was quite simple, thus: “In the dynamic braking control, excitation for the 12 traction motor fields connected in series is furnished by the diesel-generator set. Field no. 1 on this generator is controlled from the engineman's throttle by means of a rheostat and battery power. When the voltage across either the braking resistor or the motor armatures exceeds the motor field voltage, the generator differential field no. 2 is energised through the rectifier. This serves to maintain a relatively constant armature current in the traction motors regardless of motor speed.” The DB control rheostat would likely have been a higher current device than the potentiometer normally associated with the potential-wire system. And the control might have been characterized as being more like a single-unit field-loop system. I’d guess though that the field-loop for trailing diesel units – if such was fitted – might have been separate from the internal “loop”.

Also interesting from a DB perspective was the SP diesel-hydraulic fleet. The photographic evidence shows that the K-M and Alco DH units were equipped with field-loop jumper receptacles, and that they operated in MU with the SP F7 fleet amongst others, in both leading and trailing positions. That they were equipped with field-loop DB control is a reasonable deduction. In the case of the K-M hood units and the Alcos, the field-loop equipment was likely installed as built. The K-M cab units originally had all-pneumatic motoring and braking controls, and were retrofitted with compatible MU controls by SP. As well as providing field-loop control signals for trailing diesel-electric units, the SP DH fleet must also have been capable of having their hydrodynamic braking controlled from the field-loop system, which would have been an interesting exercise. Unknown is whether the SP DH fleet also had potential-wire DB control. If not, then the Alco DH643 would have been very unusual as an Alco locomotive fitted only with field-loop DB control. Also, by 1964, US domestic locomotives from any builder were either potential-wire only of “universal” with potential-wire and field-loop, so field-loop alone would have been unusual anyway.
Typewriters wrote:It's an interesting topic.

Indeed. And it would appear that much of the detail has not surfaced in the railfan world….

  by Pneudyne
Another question that comes to mind is whether locomotives fitted with field-loop DB control were ever involved in early distributed power operations, in the late 1960s. I think it is reasonable to assume that locomotives fitted with head-end DPU equipment were most likely relatively new models, and thus would have had potential-wire or perhaps universal DB controls. The Locotrol system transmitted DB commands in 8-step form, and was probably built around potential-wire DB control. At the remote end, these commands would have been converted into 8 voltage steps for the potential wire, possibly by a relay-switched resistor ladder forming a stepped potentiometer. Equally though the 8-step commands could have controlled a stepped rheostat that was part of a field-loop. So even though the head-end consist might have necessarily had potential-wire DB control, the remote consist could have been either field-loop or potential-wire.

I suspect that a look at Southern operations in the late 1960s might provide some clues. At that time the Southern then used Locotrol remote control cars rather than fitting out individual locomotives for remote control. This would have allowed remote consists to be drawn from a large pool without having to fit every last member with Locotrol equipment. And the Southern probably had a large fleet of older locomotives with field-loop only DB that it might have wanted to use as remote power. So there is some chance I think that its remote control cars were fitted to work with both potential-wire and field-loop DB consists.

If any detailed pictures could be found of the ends of the 1960s Southern remote control cars, then the presence or absence of field-loop jumper sockets would I think be a good indicator.

  by Pneudyne
Pneudyne wrote:

Westinghouse XM-781 compatible control stand and system, optional on Baldwin locomotives and standard on Lima-Hamilton locomotives. The XM-781 options included DB with either field-loop control (EMD-compatible) or potential-wire control (Alco-GE compatible), although the actual availability (and use) of the latter needs to be confirmed. Did Lima-Hamilton ever offer a DB option?
That Lima-Hamilton offered a DB option was kindly confirmed by Allen Hazen, and Will Davis advised that the known system used by Lima-Hamilton was field-loop, as recorded in the thread Lima transfer units on Pennsy.

Whether Westinghouse ever offered a potential-wire DB option with its XM-781 controller, and if so whether it was ever used, remains unknown. Perhaps that is a question that Will Davis might be able to answer.

I mentioned upthread that EMD used potential-wire DB control on its export models from the start, and that the associated US patent referred to the need to obtain adequate control in locomotives of 40 000 lb axle loading.

That axle loading number connects with an item in the Kalmbach book “Our GM Scrapbook”. On page 70, at the start of the chapter headed “These 567’s Spoke Any Language”, it was written: “Soon after World War II a call came up from Down Under to General Motors’ Electro-Motive Division for an unorthodox F3. Australia wanted to begin dieselizing with America’s most popular unit, but optional C-C and A1A-A1A wheel arrangements were required to hold the axle loading to 40,000 pounds, and a multigauge truck that would fit either standard-gauge (Commonwealth Railways) or 5-foot 3-inch track (Victoria Railways) was needed.”

So that reasonably confirms that EMD’s potential-wire DB control system was developed initially for the Australian business being pursued by Clyde-GM. Commonwealth Railways’ first EMDs did not have DB, but those of Victorian Railways did.

  by Pneudyne
Pneudyne wrote:Whether Westinghouse ever offered a potential-wire DB option with its XM-781 controller, and if so whether it was ever used, remains unknown. Perhaps that is a question that Will Davis might be able to answer.
In fact Will Davis might already have answered it. This is from the “Baldwin Diesel Shortcomings” thread in the Baldwin-Lima-Hamilton forum:
Typewriters wrote:Actually, Baldwin diesel locomotives WERE available with the Westinghouse "common standard" eight notch electric throttle as an option. Many were built this way - some or all of the New York Central's RS-12 units, and the SAL's (I believe), most or all of the Erie's roadswitchers, some of the Reading's roadswitchers .. these come to mind first.

This was the same throttle that was standard equipment on the Fairbanks-Morse "C-Line" locomotives, and was also standard on all Lima-Hamilton locomotives. According to Kirkland, this throttle could be purchased set up to MU with either EMD or ALCO-GE locomotives.

-Will Davis

I guess it depends upon how one interprets the closing sentence that I have put in bold type.

EMD and Alco-GE shared the same four-wire, eight-notch throttle control. Their respective control and MU systems differed in terms of jumper receptacles and their pinouts (relatively minor), and in the form of the dynamic brake (major). I’m inclined to take that sentence as encompassing the major and not just the minor differences. It would seem hardly worth making that statement if it referred only to the jumper receptacles.

So the Wemco XM-781 throttle, when setup to MU with EMD units would have included the option of field-loop dynamic brake control. Whereas when setup to MU with Alco-GE units, it would have included the option of potential-wire dynamic brake control.

  by Engineer Spike
From your detailed answers to my questions, the 18/20/24 series, although built with field loop, were easy to convert to potential wire. The UP discussion about consisting the turbines with GP9s leads to another question. Didn't UP use the Alco 2 jumper my system, as well as having the 27 point? Even before this, did they mix the FA and F units? Perhaps UP came up with a potential wire conversion earlier. I know that this may be odd, as UP's non EMD purchases went from FA Alco, to turbines and U25B. It seems like a long drought for anyone but EMD.

Was it beyond economical sense to convert GP9s? Like I originally posted, CPR did, but the rebuilds had completely redone electrical systems, with many modern components.
  by Pneudyne
The UP was certainly noted for using the Alco-GE 2-jumper MU system (21- and 12-pin sockets) long after the single 27-pin jumper became the de facto standard. For example that oddity was recorded in the “Trains” magazine 1968 December article on MU, as was the fact that “Y” cables were used to mate UP locomotives to Burlington locomotives with regular 27-pin sockets.

That mention appears to have been on the cusp of a change by UP, though. For example, the published photographic evidence shows that the U50C fleet was fitted with single (presumably 27-pin) sockets from new, whereas the U50 fleet had been fitted with dual sockets. One or two late service pictures of U50s shows then with one MU socket blanked off, presumably after the active socket was converted to 27-pin.

The U50 fleet shows no evidence of having been fitted with field loop jumper sockets when new. So by 1963-64, the UP had evidently standardized on potential-wire DB control, at least for new locomotives.

Interestingly the (small) SP U50 fleet seems not to have been fitted with field loop receptacles either, whereas at least some of the SP U25Bs had these, at least in their early lives. Bu then I think that the SP had tended to operate separate pools for EMD and Alco-GE locomotives, so may have been able to easily keep separate field-loop and potential-wire locomotives for the most part.

Returning to the UP, that it converted earlier EMD units to potential-wire DB is certainly a possibility. For a GP9, one possible way of doing it would have been to convert its DB circuitry to the equivalent export model configuration, using stock components. Or possibly a pseudo-GP20 configuration would have worked. In that case a micropositioner – if not already in place for humping control – would have been added, and the load regulator rheostat would have formed part of a “single-unit” local loop, with 3.6 ohms (3 x 1.2 ohms) permanently in series rather than switchable via the unit selector. Either the control stand rheostat could have been replaced by an EMD-standard DB potentiometer, or used as was, in series with a 4.8 ohms (4 x 1.2 ohms) resistor, with the control potential tapped off between the rheostat and the resistor. That would have been a bit wasteful, though – 15 amps through the resistor and doing nothing else useful just to obtain 74-volts potential for full DB. (Still, the WEMCO approach to field-loop DB was also somewhat wasteful, with most of the loop current going through divert resistors and only a small amount going through the exciter battery fields.) I hasten to add that the foregoing conversion suggestions are purely layperson’s speculations, and may well have fatal flaws that I cannot see.

Whether the UP ever mixed EMD units with Alco FAs in MU I don’t know. Stagner (1) recorded UP’s severe dissatisfaction with its Alco road locomotive purchases, including the fact that the Alco fleet was moved to the Eastern District to get the best out of them, but said nothing about mixed operations.


(1) Lloyd E. Stagner; Union Pacific Motive Power in Transition 1936-1950; South Platte Press, 1993; ISBN 0-942035-24-0. This was one of several similar and very useful books by Stagner that recorded how locomotive fleets were used and moved around during the steam-to-diesel transitional years.
  by Engineer Spike
After rereading the thread, there was mention of EMD's change from using a high powered rheostat in the control stand, and its change corresponding to the 18/20/24 series. Might it have been in the 9 series? Midway through the 9 series, they went from the can to the cash register style control stand. I think it was around 1956-57. New Haven GP9s of '56 had can, while B&M' of '57 had cash registers. I have noted the same split comparing the stands and blue card build dates on BN's ex CB&Q, C&S, and GN SD9s.

Would this make the late 9 series easier to convert?
  by Pneudyne
Yes, you are right about that. I checked some of the operating manuals available at this site: http://www.rr-fallenflags.org/manual/manual.html" onclick="window.open(this.href);return false;.

For the F9, two manuals are available, one dated 1954 February and without an edition number, so presumably the first issue, and another dated 1957 July, and labelled as the 2nd edition.

The latter refers to the newer DB control system, as used on the SD24, etc., and also shows the newer control stand, in which the throttle handle doubles as the DB handle. The former does not include a description of the DB system, but from the schematic one may see that it was the of the original field loop type with the heavy current control rheostat. The control stand was of the older type, wherein the selector handle controlled the DB.

The only GP9 manual at that site is the 4th edition, dated 1958 July. It refers to the newer DB control and newer control stand.

And there is an E9 manual dated 1954 September, no edition number but presumably the first. This shows the older DB control and control stand.

So mid-production run, give or take, and probably during 1957, there was a change of both the DB control system and the control stand for the “9” series. As you say, the later “9” series would thus have been easier to convert to potential-wire DB control.

Interesting though is that some export models, such as early G12 production, had the older control stand, with the selector lever controlling the potential-wire DB system.

  by Pneudyne
It does seem logical that, having introduced the export model potential-wire DB control using a load regulator positioner relay, EMD would then extend this development to its domestic field-loop DB system, thus eliminating the need for a heavy current rheostat in the control stand and instead using the existent load regulator rheostat. One assumes though that the positioner relay was something that was either not available at all or not well enough developed for locomotive use when the field loop system was first designed.
Pneudyne wrote: One may imagine that EMD originally developed its field-loop system as being a relatively simple way of ensuring synchronized main generator excitation of several units during DB operation. The loop enforced the same current through all battery fields, and did not depend upon for example the correct calibration and operation of “slave” mechanical devices in trailing units. At the time, four units was likely the maximum locomotive consist envisaged, and the available battery voltage was enough for four main generator battery fields in series (at maximum DB), but perhaps not more than that.
That notion is more-or-less confirmed in US patent 2304895, which appears to be the original covering the EMD field-loop dynamic braking control system.

EMD was breaking new ground, and I imagine had found little in existing DC electric locomotive practice that was either easily transferable to its diesel-electric locomotives, or as simple as desired. DC electric locomotives with regenerative or separately-excited rheostatic brakes typically had a separate motor-driven exciter for electric brake excitation, often as part of a combination motor-exciter-blower unit. The braking exciter was of relatively low power, and although it necessarily had a low-voltage, high-current armature, it had a low-current battery field system that operated at auxiliary voltage. This field could be controlled stepwise by a switched resistor network, the current being low enough that relays rather than contactors could be used. Typically, but not necessarily, there was a control trainwire for each braking notch, just as there was usually a trainwire for each motoring notch. Two or three MU jumper sockets to accommodate the large number of trainwires was not unusual for DC electric locomotives. In the EMD case, where the main generator was used for DB excitation, higher currents were required and the set of suitable resistors and contactors that would be required on each unit was probably an unattractive proposition, as would have been a multiplicity of trainwires. In comparison, the field-loop system was relatively simple.

  by Engineer Spike
There seems to be even more questions that keep popping up. One post said that the 18/20/24 series was built with only field loop. Santa Fe had a large fleet of SD24s. At the time they also bought the Alligators from Alco, which were the competing model. Did they mix them? If so, then the Alcos would have had the optional field loop, as referenced earlier.

When I was on BNSF, I ran many SD39, 40, and 45. I never noticed the double style boxes for 27 pin and fl jumpers. Same on the GP30 or 35. Did they segregate the 24s, until the SD26 program, or were they converted to potential before the 26 program?

How about the UP? They ran their SD24 fleet until the late 1970s? Were they only in dedicated service, away from the general pool, or converted? Was fl still available on the -2 line? I read D Day on the WP years ago. Apparently the GP20 was out of main line service, and not mixed with the GP40-2, or at least wouldn't have had functioning db.