Linn Westcott and realistic control.

Model Railroader Electronics and DCC
1

Linn Westcott did much work to perfect smooth realistic control of analog DC locomotives. Early on, Westcott was working to perfect the constant speed performance that is now achieved with BEMF in DCC decoders. While working on this problem, "It was Paul Larson who suggested he would rather have the train slow when going uphill so the operator is forced to give it more throttle as on an actual railroad. Likewise, when coming down one should be forced to use the brake to prevent excess speed." (MR, Feb.1962, page 62) Linn did provide load adjustments to allow for this sort of action with the True Action Throttle (TAT). Later, MR did a diffinative presentation of realistic control and simulation of train “Airbrake Valve Action” (MR, Aug. 1968). This included Dennis Blunt’s plans for a half size brake stand, with circuits to simulate the action of the train brake valve and the independent engine brake valve. This was a great presentation which raised the bar for realistic model railroad operation. Oddly, seven months later Linn published plans for the TAT IV which did not include the Dennis Blunt brake action. This presentation was much more subdued than “Airbrake Valve Action” and we might wonder what became of the enthusiasm for this “brake through” in realistic control.

The technical improvement of the Blunt circuits was that the brake functions did not cut out the throttle control like the TAT circuits do. This allowed the throttle to be partly “open” when the train brakes were applied. While this was the best practice to control slack action in the days of steam and the caboose, there does not seem to be much interest in “stretch braking” simulation for model railroading. DCC decoders do not provide for working the throttle against brake simulation. Again, we might wonder why Linn Westcott ran so far with the idea of perfecting mode

2

Rick, at the end of the day our models are powered by electric motors and our wheels don’t have any brake shoes being applied to them.

So, from where I sit, slowing is reducing the power to the motor, speeding up is increasing the power to the motor - no mater how it is accomplished.

As you might remember from previous conversations, I am a DC operator. I use Aristo Train Engineer wireless radio throttles. I feel that the push button operation of a throttle is more realistic than a “speed knob”. And, with the Aristo throttle, even at the lowest momentum setting which Aristo considers “none”, there is a “ramp up” and “ramp down” time to all throttle setting changes.

In my view, this provides for very realistic locomotive movement and behavior, without the added complexity of extra knobs, levers or settings on the part of the operator. I did not always feel that way, in fact I played with the Aristo throttle quite a bit before committing to them for the layout.

While a limited few modelers may want to sit in a chair with what simulates a control stand, most do not.

Most people in the hobby today seem very tuned in to the walk around idea, and most seem to prefer simple over complex. In fact much of the resistance to current DCC systems is the overly complex nature of many of the user interfaces currently offered - like the 32 button, two knob, no off switch?, Digitrax throttle.

For me operation is about a lot more than just operating the loco - it is about operating a small section of a whole railroad - so even with a crew of people, simplified operation of both the trains and things like CTC, makes operation easy and fun, but still very realistic to the viewer and to the operator - at least in my view.

So please explain exactly what is is you want someone to develop? It would require new user interface equipment (throttles) and that would be expensive, especially for what is likely a limited market.

3

Actually, Steve Hatch (Railway Engineering) has worked with the Tsunami decoders to tweak the CVs to get a push button brake (I believe with Digitrax throttles) simulation. Details were recorded a while back in the HOn3 Yahoo group.

As far as realistic simulations of braking action on grades - at the most basic, a controllable brake system is needed in the caboose. This would be possible in DCC with a decoder controlling the addition of friction to the caboose’s rolling. The practical limitations would be weighting the caboose sufficiently so the wheels just didn’t slide when the brakes were applied, yet keeping the caboose light enough to avoid stringlining issues.

Free-wheeling engine drives would be the next step, and very dangerous in more normal model railroad operations. I had a Lionel spur gear drive steamer that would free-wheel down a 5% grade if there were enough free-rolling cars behind it. The lack of control - shut the power off and the train keeps going downgrade - is most disconcerting. And it leads to f

4

[quote user=“ATLANTIC CENTRAL”]

Rick, at the end of the day our models are powered by electric motors and our wheels don’t have any brake shoes being applied to them.

So, from where I sit, slowing is reducing the power to the motor, speeding up is increasing the power to the motor - no mater how it is accomplished.

As you might remember from previous conversations, I am a DC operator. I use Aristo Train Engineer wireless radio throttles. I feel that the push button operation of a throttle is more realistic than a “speed knob”. And, with the Aristo throttle, even at the lowest momentum setting which Aristo considers “none”, there is a “ramp up” and “ramp down” time to all throttle setting changes.

In my view, this provides for very realistic locomotive movement and behavior, without the added complexity of extra knobs, levers or settings on the part of the operator. I did not always feel that way, in fact I played with the Aristo throttle quite a bit before committing to them for the layout.

While a limited few modelers may want to sit in a chair with what simulates a control stand, most do not.

Most people in the hobby today seem very tuned in to the walk around idea, and most seem to prefer simple over complex. In fact much of the resistance to current DCC systems is the overly complex nature of many of the user interfaces currently offered - like the 32 button, two knob, no off switch?, Digitrax throttle.

For me operation is about a lot more than just operating the loco - it is about operating a small section of a whole railroad - so even with a crew of people, simplified operation of both the trains and things like CTC, makes operation easy and fun, but still very realistic to the viewer and to the operator - at least in my view.

So please explain exactly what is is you want someone to develop? It would require new user interface equipment (throttles) and that would be expensive, especi

5

Sheldon,

First, I should explain that I did work the steam excusion business and now work as a BNSF engineer, so my view is colored by 1:1 steam and diesel lovomotive operation. As a model railroader, I think about some of the challenges and joys of 1:1 railroading that have not yet been scaled down for model railroading. In addition, Bill Darnaby made a wonderful description of a train order meet with steam locomotives (Model Railroad Planning 1995, pg. 58). “You can actually hear the engineer bringing his train under control as it approaches the west end of Avoca siding.” This engineer must be prepared to stop since his train has no authority past the west end of Avoca until the opposing train is in the clear. If the engineer simply comes to a stop, then time and momentum will be lost. If the engineer comes in too hot, he will end up like Casey Jones going through the back of the opposing train if it is not in the clear. The trick is to set some air and keep working steam to keep the train stretched. If the opposing train is not in the clear, the throttle can be closed and more air set to bring the train to a quick stop. If the way is clear, the brakes are released and the throttle opened to get back up to speed. There is a lot of satisfaction in being able to accomplish this, and more drama than simply turning the knob or pushing the button to make the train stop on a dime and give you nine cents change.

And I do not blame anyone for not liking DCC sound. Sound systems are not yet able to “lip-sync” the dramatic action of an Avoca style meet.

Meanwhile, I do enjoy the two-button throttle on the radio remote that controls my Athearn HO Challenger. I have the hot lead to the main circuit board routed through a jack so that a battery car can be plugged in. On battery power, performance is flawless and “good enough” until something better comes along.

6

Fred,

The problem with the TAT and DCC decoders is that the throttle is disconnected when the brake is applied. Also, most DCC F7 brake functions toggle on and off, so there is no way to increase the brake effect, or to reduce it by adding more throttle. In 1:1 railroading, it is possible to work the throttle against the brakes to make precise stops. The Dennis Blunt circuits modified the TAT so that it was possible to work the throttle against the brakes (MR, Aug. 1968, pg. 48). This was all electronic simulation without a powered caboose or tender. The “brakes” were all in the worm gear of the open frame motors of the time.

Likewise, the implication was that the TAT could be adjusted so that it would respond by making the locomotive go faster when going downhill, so that the “brakes” would actually have to be “applied”. I have not seen any reports of this being done. It would be interesting to know if this can be done with a DCC decoder.

BTW, I did have too many cars behind a locomotive on too steep a grade. The locomotive slid down the hill and derailed with most of the cars. Most realistic train wreck (Layout Design Journal 40, pg. 4). Be assured that a permanent magnet motor will stop if there is any kind of resistance across the brushes. The motor tries to act as a generator against the permanent magnet field. Lots of resistance.

7

a basic throttle controls the voltage to the engine, making the engine go the speed set by the throttle.

When you consider the TAT throttles or others that mimic momentum, the throttle now no longer directly controls the engine speed/voltage, but the desired speed/voltage and the momentum circuit adds some delay to providing it to the engine. In other words, the circuit starts becoming a train simulator. On the TAT, there were controls to select the length/weight of the train that controlled how slowly it reacted.

it’s interesting that because each car has brakes, the time it takes a train to slow/stop is not necessarily dependent on the length of the train, while the length (weight) does affect its acceleration.

one approach to building a better simulator is to have separate control for the throttle and brake. The throttle no longer indicates the desired speed of the engine but the amount of power/force applied to the wheels. So shutting off the throttle doesn’t necessarily cause the engine slow, but simply stops maintaining speed. Application of the brake causes the train to decelerate. BEMF could be used to help determine if the train is going up or down hill based on the difference between the applied voltage and BEMF.

the processor within a dcc controller can mathematically simulate the speed of the train given the weight, based on the BEMF and application of throttle and brake. In other words, it calculates what the train speed should be and applies the necessary voltage to make it so.

there’s always emergency stop!

greg

8

If the goal is to make a simulator, this can be done in the same manner as Blunt did, the actual control system, DC or DCC, is totally irrelevent. For about the same level of difficulty, you could make a DCC throttle where there is a brake handle that reduces but does not cut off the throttle as it’s applied. In an analog fashion.

Even better, JMRI supports those Rail Driver control stands, to make the brake work against the throttle there would be more along the lines of some scripting code. Then you’d have realistic controls to work with as well (they do, or did, have a traditional-style control stand as well as the more commonly seen console model - hopefully they still make that, since the console is totally inappropriate for my era).

–Randy

9

Have you tried Aristo’s version of direct radio control? also which version of the Train Engineer do you have, I have and use the two channel but just aquired the 10.

No, I never have used the HO onboard Train Engineer.

I have eight of the 10 channel Train Engineer throttles and use them in conjunction with a system of Advanced Cab Control that integrates signaling, CTC, turnout control, and cab assignment into one system.

With a CTC dispatcher on duty, engineers only need to run their train, just like DCC.

Without a dispatcher, many cab assignments are semi automatic and linked to turnout position. turnout positions and cab assignments are then handled at local tower panels that allow complete walk around operation.

The limitation of the Aristo onboard train engineer was the size of the receiver.

I have one of the new Revolution throttles , but have yet to do much with it.

Sheldon

10

If Linn Westcott were alive today, would he be working with transistors, ICs, FPGAs, micro/picoprocessors, …?

11

[quote user=“Rick Mugele”]

Sheldon,

First, I should explain that I did work the steam excusion business and now work as a BNSF engineer, so my view is colored by 1:1 steam and diesel lovomotive operation. As a model railroader, I think about some of the challenges and joys of 1:1 railroading that have not yet been scaled down for model railroading. In addition, Bill Darnaby made a wonderful description of a train order meet with steam locomotives (Model Railroad Planning 1995, pg. 58). “You can actually hear the engineer bringing his train under control as it approaches the west end of Avoca siding.” This engineer must be prepared to stop since his train has no authority past the west end of Avoca until the opposing train is in the clear. If the engineer simply comes to a stop, then time and momentum will be lost. If the engineer comes in too hot, he will end up like Casey Jones going through the back of the opposing train if it is not in the clear. The trick is to set some air and keep working steam to keep the train stretched. If the opposing train is not in the clear, the throttle can be closed and more air set to bring the train to a quick stop. If the way is clear, the brakes are released and the throttle opened to get back up to speed. There is a lot of satisfaction in being able to accomplish this, and more drama than simply turning the knob or pushing the button to make the train stop on a dime and give you nine cents change.

And I do not blame anyone for not liking DCC sound. Sound systems are not yet able to “lip-sync” the dramatic action of an Avoca style meet.

Meanwhile, I do enjoy the two-button throttle on the radio remote that controls my Athearn HO Challenger. I have the hot lead to the main circuit board routed through a jack so that a battery car can be plugged in. On battery power, performance is flawless and “good enough” until s

12

The “braking” of the motor in the engine is exactly the same input to the motor as reducing the throttle by a pre-determined amount at a pre-determined rate. Whether the throttle is disconnected when the brake in the model world is applied is a simple electronic function. When the brake is applied, the controller simply needs to know the state of the throttle, and resume throttle setting - preferably with a reasonable amount of momentum (or delay) - when the brake is released. The throttle setting at brake application can also be used to determine braking rate and deceleration due to brake.

The point I was making is that in our models, there is only a single input point - voltage to the motor. The motors are generally linear with voltage, except at the bottom and top ends. At the bottom, drop the effective voltage below a threshold (threshold varies with real load and pulse shape) and the motor stops turning, locking the drive train. At the top end, our motors generally have little RPM variation with increased voltage. The top end is not far off prototype behavior, but the bottom end is very unrealistic. Therefore, we try to avoid operating near the low end threshold of the motor by high ratio gearing and pulse application.

<

13

Sheldon, how big a layout do you have ??? It must be huge to need that many channels.

14

Greg,

Actually, the rehostat of the most basic throttle does not control voltage but power expressed as watts (volts times amps). Input would be the basic 12 volts modified by rehostat resistance. Some folks used Variac transformers to control voltage output to maintain constant speed. Amperage draw would vary according to load changes but voltage would remain constant. On a 1:1 locomotive it is actually possible to watch amperage drop and voltage climb as the train accelerates given a constant power setting. Probably see this with a model locomotive and a basic power pack and meters.

Now we just need to figure how to send the proper signals for a DCC decoder to interpret as power and brake applications.

15

My layout room is a dedicated 22’ x 40’ heated and cooled space above my detached garage.

When complete the layout will fill the room and consist of a double track mainline on two levels about 8 scale miles long, or about a 500’ actual run.

The mainline will support the operation of four to six trains. There is a separate passenger terminal, engine terminal, main yard and industrial belt line area that will support the operation of three operators, two working the yards, one on the industrial belt line.

Additionally there will be a single track “branch-line” representing a separate railroad that interchanges with the ATLANTIC CENTRAL and that line will require one throttle.

Hidden staging will be provided for about 30 trains, typically 35-40 cars in length.

Full operation will require a crew consisting of eight engineers, two dispatchers, yard master and several conductors for local switching.

The layout is also set up in such a way that four mainline trains can be on “display loops” if desired and this will still allow for separate operation of the other areas, yards, passenger terminal, etc.

While big, I would not call it “huge”, I know a number of modelers with layouts larger and/or more complex than mine.

I consider my layout large, but relatively simple, designed for long trains and realistic operation while still maintaining good display operation for non railroaders.

Sheldon

16

The thing is, none of that matters. It’s all a simulation anyway, You’re getting too caught up in the details of how Blunt’s design worked. The EFFECT of it was that as you apply the brake, the loco slows but not as fast as it would if the throttle was closed first, and if you let off the brake, the loco will speed back up. This might be easier to accomplish with a potentiometer throttle, but I’m confident it could be done digitally as well - the throttle and brake in Train Simulator work prototypically, even down to independent and dynamics. True actual train braking will never work in small scales - the physics doesn;t scale - so it’s always a simulation.

–Randy

17

Agreed!

And so, I suggest that it is infinitely more practical, in our selectively compressed little worlds, to simply learn how to slow or speed up the train in a manner that reasonably reflects how it would appear to the casual viewer.

I find the Aristo Train Engineer very good at this, better than most DCC setups. Without any programing or other “fiddling”, most locos respond in a very smooth and prototypical manner. The ramp up and ramp down times built into the TE provide the perfect amount of delay to simulate throttle and brake actions while still allowing precise control for switching.

We are not in the cab, we don’t have “seat of the pants” feel or onboard instruments to go by - we are observers - even as we a are miniature operators.

But that is just my view, if I wanted that real life experience, I would drive up to Strasburg and apply for a job.

Disclaimer - my modeling goals are more focused on running a “railroad”, not just one locomotive.

Rick, ideas like your battery powered Challenger are interesting, but at the end of the day I need and want to operate everything from a GE 44 Tonner to a C&O H8 on the same system. And I need

18

Thanks Sheldon!!! I was wondering as I have a large (15x30) and a single large dogbone with a yard at both ends, max 3 operators. Main is the Train Engineer and one yard (very small with a carfloat interchange) is a standard MRC 2500 with the other main yard being a handheld control linked to a 2500 as it is a much bigger yard.

19

That’s exactly why I prefer the encoder throttles over the potentiometer ones. Since it can take several truns to go from stop to full speed, I can very easily accelerate and decelerate in a prototypical fashion, no jackrabiits, no slamming on the brakes. And I don;t have to fiddle with the decoder settings. I do typically put a little momentum in my locos - basically the least common denominator, mainly, light loco moves. If I would just crank open the throttle with a train, it would be too fast to be realistic. But since th encoder makes tiny little steps, I can run up to speed very slowly, simulating the load of that train.

It also comes in handy with sound - though I know this doesn’t appeal to you, with many sound decoders, the momentum controls the ‘loading’ effect on the prime mover. There, it needs more momentum programmed in, simulating the heavy load, which you can bypass by slowly opening the throttle. For a heavy train - crank open the throttle, prime mover loads up, loco begins just creeping. You can throttle back to reduce the prime mover, or else if the throttle is at the speed you want anyway, as the loco reaches that speed, the prime mover will settle down. All pretty neat and effective as long as you don;t have the thing cranked up to toy train sound volumes.

Consider making that Strasburg trip around the July 4th week - I’ll be there with the RCT&HS big modular layout in the museum.

–Randy

20

The throttle circuits used to simulate a real train should be easy to implement on a Digitrax system as long as the circuit does not rely on reading the motor’s bemf and as long as the circuits output is smooth(actually even a pulsed output would be usable if fed through a resistor capacitor filter). Just feed the circuit’s output into a Zephyr’s jump port.