I have...

I have often wondered the same thing!

Looking forward to hearing from one of our engineers on this one.

So, I tapped my friend Tom

So, I tapped my friend Tom Gasior on the shoulder about this tonight.

Tom is an engineer who runs big yellow locomotives (not just the ones on the SplitRock), mostly in Minnesota these days.

Here's our little convo...

Bill: what say you? http://model-railroad-hobbyist.com/node/24130

Tom: Ahhh. And I haven't even dug into the last request?

Interesting question. Notches don't equal specific speed. That has more to go with load and grade. I can be in idle going 60 and Mitch 8 going 2 mph

Bill: I know that   the real question I think is "do you finesse the speed" or "just leave it in the closest notch" when running.  My suspicion is that you are always working the throttle.

Tom: Depends on terrain. Most times you can leave it in one notch. The new PTC and other computer gizmos take control unless you have other than clear signals

Bill: so, hypothetically; if the speed limit is 35, and you're on level ground with 30 miles to go, and notch 4 would maintain 38, and notch 3 would maintain 30, what do you do?

Tom: Go between 3 and 4. No speeding... That's frowned upon. No going under speed, that's delaying a train.

Bill: so, constantly work the throttle wink emoticon

Tom: Most likely. Get up to speed. Then coast, and add a notch when it starts to fall off

Bill: yup, that's what I figured.

Tom: Most times trains settle in on a speed. The longer the train, the easier to run.

Bill: cool.

(reposted here with permission)

Sounds like

an automatic transmission on a vehicle. When starting a 7000lb vehicle you apply the throttle. As the vehicle gets up to speed you let off on the accelerator. If you are towing a 10,000lb trailer it will take more throttle to get that weight to speed. Once there you feather the throttle to keep it at double nickels. Sound about right? 

Bernd

And that's. . .

. . . the way it is. I do have some mainline throttle time myself -- CF7s on the old Washington Central, some Dash 8s and 9s elsewhere -- oh, 20-30 hours I suppose, nothing major.

But there is a fair bit of moving up and down (or left and right, depending on the control stand) between notches to keep a consistent speed. (But of course no greater than the maximum authorized speed, depending either on speed restrictions or other factors, such as hazardous materials).

Inertia, probably? I can't claim to understand all the dynamics inherent in a train -- long or short -- but they are prevalent. Have to be with all that tonnage moving at speed with the push-pull forces and centrifugal forces as well.

Thanks, Bill! 

Thanks, it is clearer now

Bill- thanks for the very clear and useful response from your friend Tom Gasor.  I knew a real engineer could set us straight. 

OK- so they watch the speedometer and keep manipulating the throttle between notches to maintain the exact speed they want.

When I made this post, I was thinking that it did not have any real modeling implications.  But, it does-- now we know how to really use manual notching with DCC sound decoders to emulate the real deal.  It is more than just notching down to a lower level once the train is rolling at a constant speed.  It is a "random" pattern of notching up and down between two notches.

I have never used any of the computer-based train simulators, but it would be great to have a program that modeled the real mechanical dynamics of a train and let you get the experience of manipulating the throttle to maintain a constant speed under a variety of conditions.  Clearly, this would take a lot of practice to get it right!

Another observation:  a given notch setting probably doesn't totally keep the diesel prime mover running at a particular rpm.  It probably stays close to a constant rpm, but I would guess it waivers a little bit as the mechanical load imposed on the diesel by the generator varies, which in turn is a result of the varying electrical load that the traction motors present to the generator.  I would think that if you carefully analyzed a sound recording of a loco pulling a train along at a "steady" speed, you would be able to pick out all the subtle changes in prime mover rpm.

Imitating this effect could be "the next big thing" in sound decoder realism.  "Real Notch." 

One question remains: why do they have the notching system?  Why not just have a continuous throttle on the diesel prime mover, like a car's gas pedal?  Maybe the answer is safety?  It might be too easy for the engineer to inadvertently "step on the gas?"  With notches, little risk of pushing the throttle farther than you intended.

Once again, we can learn so much from this forum.  A great place.

Dennis

I suspect that is the way

I suspect that is the way Hermann Lemp designed it. I hope I have the name right.

https://en.wikipedia.org/wiki/Hermann_Lemp

I read an article in one of the magazines about him and discovered that in addition to the diesel control system he also held the patent that was used for automatic drip coffee makers, now there is a guy that invented practical things!

Here's a couple of video's

Listen as they throttle up.

Bernd 

Prime mover RPM

The engineer's throttle does not control the prime mover directly. The diesel motor has a governor that monitors rpm- either mechanically by the old "balls to the wall" mechanism, or electronically in modern units ( I presume). The target RPM of a mechanical governor is changed by a set of solenoids that tighten a spring- each throttle notch energizes a solenoid. That allows multiple unit operation-- the solenoids in each loco are connected in parallel and work in unison.

If the governor is working properly, there will be very little variation in rpm when the load is constant. If the engineer is playing the throttle to keep a speed between notches, there will be some pattern to the variations in RPM, and it would be possible to program that into a decoder. A simple mathematical proportion should do it: on the 128 step speed range, step 64 is notch 4 and 80 is notch 5. If a speed of 72 is selected, the sound should alternate evenly between the two notches. If a speed of 76 is selected, you should hear notch 5 for 3 seconds, then a second of notch 4, with some random variation of course. I don't know if this would be audible over the variation caused by changing back-emf though.

pqe

Alco's switching

One of my favorite engines to listen to while switching.

What Pelsea said

This is from Wikipedia:

Locomotive power output, and thus speed, is typically controlled by the engine driver using a stepped or "notched"  throttle that produces  binary-like electrical signals corresponding to throttle position. This basic design lends itself well to  multiple unit (MU) operation by producing discrete conditions that assure that all units in a  consist respond in the same way to throttle position. Binary encoding also helps to minimize the number of  trainlines (electrical connections) that are required to pass signals from unit to unit. For example, only four trainlines are required to encode all possible throttle positions.

David Stewart

What I hear

The audio in the videos Bernd posted while I was writing my previous is instructive:

In the first, there are two principal sounds- the whine of the traction motors and the chug of the diesels. The whine varies in pitch and loudness according to the acceleration (loudness) and speed (pitch) of the train. The prime mover pitch (more of a rumble) changes in steps, but the loudness follows the loudness of the traction motors. This makes sense-- as the load on the engine grows, the governor applies more fuel to the injectors.

In the second, the main variation is beating between the three engines-- they run at slightly different rpm, so the sound gets louder and softer as the phase changes. It is a testament to the skill of the mechanics that they are so well in tune. There's also a bit of doppler change as the locos go past the camera, but not much.

pqe

Sounds like the notching helps with MU ops

Binary notching

I found a pinout for the MU cable, which lists solenoids A, B, C, and D, so rather than a solenoid for each step,  the four are used in combination, as David points out. Here are some references:

An army manual,

Railway-technical.com

I'd like to see a diagram of the mechanism, as the code given in the second reference is kind of strange.

pqe

Awesome!

We've learned a lot here already.

This is a great thread folks!

Digging deeper

Pelsea- Thanks for the information about the real connection between the throttle and the prime mover, the speed governors, and the benefits it brings for MU operation.  I agree with you that the difference of moving up a notch, down a notch, would probably not be very audible in our models.  I can tell the difference on my layout between notch 1 and notch 8, but if I notch up from 3 to 4, say, I often am not sure that a change really occurred.

Bernd- Great video/sound recordings.  That adds a lot.

David- Notching as a way to simplify control of MUs makes sense.  Thanks for pointing this out.  I wonder: the earliest diesel box cabs, in the 1920s, had notched throttles, I believe.  But I would also guess that the designers of those locos were not anticipating that they might want to MU them?

Joe- If you are still listening, it would be great to see an MRH article, or a video, about "how a diesel electric loco really works."  Two parts: first, the details of how the loco operates, second: what the engineer has to do in order to properly move the train.  Add some details about having to watch the current draw from the generator (I think I've read it can go up to 1,000 to 2,000 amps short term when starting from a stop) and paying attention to the short-term rating maximums for the traction motors.  The few "cab ride" videos I have seen tend to focus on the view out the window and watching for signals, which is all good, but not the same thing.

It would be great to get hands on a "how to operate a diesel loco" training book from the 1950's, say.  It would also be neat to get access to the training material that today's railroads use to train loco engineers.  Anyone know where to find such stuff?

This whole discussion reminds me of one of the reasons I love trains: huge machines moving heavy stuff.

Dennis

You ain't kiddin'!

"Why don't they use a

   Keep in mind the 8 notch solenoid system was designed a long time ago. With today's industrial know how,computers,and solid state devices they probably come up with a more efficient system if starting from scratch. ......DaveB

Lets add a second element to making an engine move.

I believe it's called transitioning. It's where the traction motors are all in series at start up and as the speed increases they transition from series to series-parallel. That's where the two motors of one truck are in series and the two trucks are in parallel. Then they transition to all motors in parallel. Am I correct on this? I think that's how they worked.

Bernd

"how to operate a diesel loco" training book from the 1950's

Our club library has manuals for over 20 diesel locomotive and a few steamers, but of course the info is on the web. So for Operators Manuals check out  http://www.rr-fallenflags.org/manual/manual.html. GE had posted some training information on some os their loc's along with some videos (eg.

Thanks for the links to the operator's manuals

Ken- Thanks so much for the links to the fallen flags organization's collection of operator's manuals.  A treasure trove.  I have gone through a first pass of the Alco FA2 operation manual and it is filling in a lot of gaps.  It does discuss the ability to change the motor connections in several variations of series and parallel that Bernd mentioned above.  You could either have the control system make those changes automatically, based on speed, or you could control those transitions manually.

I couldn't find anything though which said which button you push to make the Alco send out its smoke cloud.

Many hours of good reading ahead.

Dennis

Baldwins

As an aside, most Baldwins used an air throttle, that was more or less continuous. Which is why Baldwins couldn't MU with EMD/ALCO/GE units with an electric throttle.  Later Baldwins used an 8 notch electric throttle, so they could MU with EMD/ALCO/GE but not other Baldwins.

Oh, hey...

oh, hey, that's EASY...  it's the Start Button!

Why Notching?

As Auto-files we spend a great deal of time trying to compare systems to the operation of automobile which just complicates things. What has been missed in this discussion is that notching was the result of a design (by an Electrical Engineer) to control the amount of power delivered from the motor generator set to the traction motors. A design which required contactors (big relays) to switch the configuration of the wiring from the generator(s) to the motor field and armature windings ( Operation and Maintenance of Diesel Locomotives, p. 75. fig 45). An intrinsically nonlinear system. I am not going to dig into the electrical theory (explanations are available in references and most modelers could care less), but at low speed (notch 1 & 2) motor windings were often wired in series (p. 78. fig 46(1)) while at high speeds the windings were switched to parallel (p. 78. fig 46(2)). Between notch 1 and notch 8 (or 10 or other if you are not in North America) the motors may be wired in series, parallel, series / parallel (p. 78. fig 46(3)), resistance added or several hundred other configurations all controlled by big, noisy relays. So as the engineer (driver) notched up or down he was essential rewiring the locomotive (if you ever get a chance to ride in one of these older locomotives stand by a High Voltage cabinet and listen to the clatter [and arching - like Lionel engines real locomotives also generated a bit of ozone]). The control of the governor was linked to the expected power demand of each contactor (notch) configuration.

So Why Notching? In the early 20^th Century no other reliable, power efficient system, which would fit in a locomotive, existed for adjusting the power between the generator and the traction motors. There were several competing systems like the Baldwin and ALCO pneumatic systems (rode in an ALCO DL 109/110 set and a PA but no abnormal smoke), but none of these systems proved as reliable as the GE - Limp system especially in MU ops. No simple, reliable, linear (gas peddle like) system existed until the development of high voltage rectifiers and semiconductor control systems. Although rarer some modern locomotives still rely on relays to switch the configuration of generator / motor wiring (it is simple, less expensive, reliable and easy to repair).