Stopping train without E unit reversing?

@Gene H posted:

It was about the size of a accessory button and you only had to connect it to the center rail.

Yep, that would be consistent with adding a simple resistor in series with the "hot" rail (I suppose there's no reason it wouldn't work on the outer rail connection on a three-rail track, but in operation that would create a voltage differential on what is typically treated as a ground . . . and that's potentially a Very Bad Thing!😳).

@Steve Tyler posted:

Side thought: I don't believe I've ever heard why they call it an "E unit" -- anyone know where the term came from?

Engines with mechanical direction control had no suffix on the engine.  An E designation was added to engines with an electronic reversing unit -  - "e-unit.  The designation was later dropped when all products had the e-unit and the E was no longer necessary.  That device description got shorten to e-unit.

It's difficult to put yourself in a 1930s-1950s frame of mind.   But electricity was somewhat magical then as was remote control reversing of an engine.  I think the e-unit was on the top 10 list of innovations of lionel trains..  Cell phones, GPS, battery cars and internet were in the future.   Just like possibly space travel, teleporting and world peace seem distant today

@aussteve posted:

Engines with mechanical direction control had no suffix on the engine.  An E designation was added to engines with an electronic reversing unit -  - "e-unit.  The designation was later dropped when all products had the e-unit and the E was no longer necessary.  That device description got shorten to e-unit.

It's difficult to put yourself in a 1930s-1950s frame of mind.   But electricity was somewhat magical then as was remote control reversing of an engine.  I think the e-unit was on the top 10 list of innovations of lionel trains..  Cell phones, GPS, battery cars and internet were in the future.   Just like possibly space travel, teleporting and world peace seem distant today

Interesting, and certainly plausible, though I'm relatively sure the "E" would have been "electric" rather than "electronic", since E-units predated the use of the term "electronic" by quite a few years, and there's really nothing even vaguely electronic about a classic electro-mechanical E-unit -- after all, it's really just a special purpose stepping relay/solenoid.

BTW, I have a few spare E-units I'd trade for world peace -- any takers?

Dunks train control from about the 50's used automotive type light bulbs as resistors. The manual listed other bulb numbers but GE 1141's were installed in the units I had.

@Steve Tyler posted:

Interesting, and certainly plausible, though I'm relatively sure the "E" would have been "electric" rather than "electronic", since E-units predated the use of the term "electronic" by quite a few years, and there's really nothing even vaguely electronic about a classic electro-mechanical E-unit -- after all, it's really just a special purpose stepping relay/solenoid.

BTW, I have a few spare E-units I'd trade for world peace -- any takers?

Yes we have no shortage of rivet counters.

"E" for "escapement" as in, the mechanical basis for the unit's operation. (Or maybe designs A through D didn't work well.)

Interesting Dunk's box. Lamps soldered in -- must have been a treat to work on when one blew out and needed to be replaced.

@David Nissen posted:

Dunks train control from about the 50's used automotive type light bulbs as resistors. The manual listed other bulb numbers but GE 1141's were installed in the units I had.

Fascinating! Basically, just a box with two automotive bulbs that you can throw in series with the track, to defeat the reversing E-unit, very similar in function to the various resistors used in other configurations for the same purpose.

Back when I worked in a gas station, I sold many of those types of bulbs to replace customers' burned-out turn signal and rear lights. My first impulse in trying to defeat the E-unit was to throw a pair of E-10 base track voltage lights in series, but they turned out to be too high resistance/low wattage. The higher power automotive lights would have been about right (and obviously were!). Dunno how much the Dunk boxes sold for, but it certainly would have been possible to "roll your own" with a bit of creativity! I have to give a shout-out to the manual writer's creativity -- while more or less staying factual, s/he makes the Dunk box sound like a sure-fire cure for cancer!

@Steve Tyler posted:

And speaking of labors, my project is continuing to move forward. I dug out some more old track, and extended the basic figure-8 test bed to closer resemble the elongated figure-8 used in the Marx "Twin Train" setup. I also visited the local hobby shop to pick up a pack of insulating pins (I know I have most of the remainder of a pack somewhere, but haven't been able to find them, and hey, they're cheap, right? ), and while I was there I found and negotiated a cash deal on two legacy signal lights: a Lionel 151 with a broken semaphore arm, and a rusty Marx 454 with a defunct switch.  . . . .  I plan to drive both signal lights from the relays I got the other day. Oh, and the power resistors arrived, so I can start experimenting, designing and bread-boarding my prototype legacy block control system. It's even odds as to whether I can get it to work before the book I ordered at your recommendation arrives!

Well, I've got a few things to tuck in before I declare *full* success, but after my recent labors, I'd estimate I'm about 90% there (and the book is still more than two weeks from delivery!). I'll write everything up eventually, but a brief progress report:

First, the question that I asked to start this thread has been definitively answered: yes, it is possible to stop an engine, yet allow a small trickle current to hold the E-unit in place and prevent it from reversing direction. Among the many options (including Dunk's!) for creating this condition, I ended up using four 10 ohm, 10 watt ceramic resistors (two parallel pairs wired in series, to handle the power that might have to be dissipated if there ever was a prolonged dead short across the track) that trickles power into the control section when a train is stopped there.

Second, the 12 VAC relays I selected seem to be working fine, activating immediately when any rolling stock enters the "sensing" section. At times there's some chatter caused by intermittent contact and/or dirty track, but I don't think that will impair operation even if the condition persists after cleaning, etc.

Third, when only one train was run on the track, it operated entirely normally. Each time it traversed the sensing section, the signal (which faces toward any train approaching the crossover from the other direction) changed from green to red as it should, but since there *was* no other train on the track, normal one-train operation was not impaired, and the E-unit operated normally.

Fourth, I was able to briefly run two trains on the track (slowly!), but discovered the power drawn by those engines (two Marx 666 smokers, in one test) exceeded the power capacity of the small transformer I was using. Another small transformer similarly failed, repeatedly popping the circuit breaker as the first had. So, I was unable to conduct a full test of the setup, proving the second train will stop in the control block until the other train clears the sensing section, then resume motion in the same direction. To simulate this, though, I was able to use a single caboose to intermittently trigger the sensing section, which changed the signal from green to red and reduce the power to the control section to a trickle, stopping the running train. When the caboose was lifted off the track or rolled forward out of the sensing section (simulating a second train leaving the sensing section), the stopped train resumed operation without reversing.

So, other than the absence of a full two-train test run, I believe all the critical elements and functional capabilities of the Marx "Twin Train" set have been successfully duplicated, as I hoped to do. It should be noted that I made one executive decision: rather than try to precisely duplicate the Twin Train set's "bump and run" configuration, where the first train to reach the control section will be stopped until the other train approaches the control section from behind, which then allows the first train to proceed, rather than sensing a potential conflict at the crossing and stopping the approaching train until the possible conflict is avoided, as a true block control system would. In other words, I implemented something closer to a true (albeit simple) block control system, one that could potentially be extended to multiple blocks or extended in more complicated block control systems (BTW, that's why I purchased 4PDT relays, which might be better suited to more complex decision trees to produce a more robust block control system).

Stay tuned . . .

O.K., I declare victory, even with a few niggles remaining to iron out:

A few footnotes:

- To solve the lack of power, I ended up carefully matching the output phase of *two* small transformers (SAFETY NOTE: hook them up wrong and you'll short BOTH out!), which together in parallel provided enough power to run two engines on the same circuit (neither was powerful enough alone, and their circuit breakers had kept popping when I tried).

- Even with the extra power, I was not able to get both vintage Marx engines running properly (one or the other always seemed to slow to a crawl), so to replace one of them, I substituted a (somewhat newer) Lionel engine I'd had on the shelf. As it happens, it had a forward-neutral-reverse-neutral operating E-unit, which drove me crazy trying to sync it with the other (forward-reverse-forward) Marx 666's E-unit, so I cheated a bit by looking for and throwing the Lionel's E-unit lockout switch (yes, it had one I'd never noticed in over 30 years!). So, that engine did not actually have to rely on the E-unit-defeating current trickle through the resistors when that engine was halted, because it would not reverse in any event.

- The three track section "Control Block" proved sufficient: the lightly-loaded (weighted tender only) Lionel engine tended to roll forward a bit more than the Marx 666, which had a tender, car and caboose in its consist, but both stopped well short of the crossover section.

- The eight track section "Sensing Section" I set up (the entire loop on that side of the crossover) is probably a bit too long, and keeps the relay and red light activated much longer than necessary, though in doing so it absolutely ensures a close-following train is locked out of the crossover in time (with longer consists, I can envision the approaching train clearing the Control Section before the first train reaches the Sensing Section, which might leave the back end cars on the second train vulnerable in the crossover as the first train approaches it). I may experiment with moving the insulating pin a section at a time, and see how that affects operations.

- As I noted before, the Sensing Section activation is a bit ragged and intermittent, but so far the control function has worked well. In other words, the momentary repeated dropouts have not proven sufficient to enable the stopped train to advance out of the Control Block and into danger, though I will see if cleaning and repairing the Sensing Section track will reduce the problem.

- The Marx signal block I bought had some intermittent ground problems, so I substituted a crossing light instead, with red and green bulbs. When I'm satisfied the signal block is operating properly, I'll restore it to trackside, and add the other signal I bought to the other track at the crossover (colors inverted, of course!).

I'll write my whole quest up in a bit more detail and post it soon, for those who might be interested in the process and sources, and again when(if?) I get around to implementing all this on my actual layout in due course!

As promised, I've attached a write-up of my project experiences so far. Just for giggles, I decided to do it up like a term paper, footnotes and appendices and all. The careful observer might notice that portions of the paper strongly resemble portions of previous on-line discussions here and elsewhere, though I was far from compulsive about attributing other's numerous specific contributions to the project (sorry -- but thanks again to all who weighed in!). I've also included references to and copies of some relevant source materials, if you'd like to do a deeper dive. If/when I get around to attempting to incorporate the "Twin Train" set functionality I've been exploring in my bread-boarded experimentation onto my layout, I'll add an addendum (and probably a video or two!). Enjoy!

Over the years, I have experimented with some pretty advanced layout based train control systems. What exactly are you trying to accomplish?

@BOB WALKER posted:

Over the years, I have experimented with some pretty advanced layout based train control systems. What exactly are you trying to accomplish?

Bob,

Steve's trying to replicate what Marx called "Twin Train".  Two trains purchased as a set in one box, and configured to run simultaneously around a single track route, nicely spaced, with no collisions occurring between them.

This is not anywhere near "pretty advanced".  It's simply a back-in-the-day kind of approach to a little automation.

Thank you Steve for the effort.  It's always fun to learn something new, even if it's 65 years old.

Mike

@Mellow Hudson Mike posted:

Bob,

Steve's trying to replicate what Marx called "Twin Train".  Two trains purchased as a set in one box, and configured to run simultaneously around a single track route, nicely spaced, with no collisions occurring between them.

This is not anywhere near "pretty advanced".  It's simply a back-in-the-day kind of approach to a little automation.

Thank you Steve for the effort.  It's always fun to learn something new, even if it's 65 years old.

Mike

Absolutely right, Mike.

When I first came across references to the Marx "Twin Train" set and checked out the videos of it in action, I was fascinated by the clever way Marx had simulated a block control system in a simple, self-contained set, albeit in a very simple and limited situation. The Gilmer paper (which I attached in full as Appendix 1) went into considerable detail about how Marx did it, and how limited their system actually was, and how difficult it would be if the attempt was made to expand it into a true block control system.

My quest was really in two parts: first, I wanted to experiment with the "Twin Train" system to see if it (or some variation) could be incorporated in my layout, and to do that, I needed to create a home brewed version of that system, since the originals are almost unavailable, and as the Gilmer paper elaborates, the original hardware, while clever in design, had built-in flaws that would limit its use outside a simple configuration. And second, I'd like to see if a "Twin Train"-type system could be modified to allow its development into a more robust block control system, in the manner explored by Gilmer.

In the process, I discovered "back in the day" resources that set forth in precise detail how the system I'd been pursuing had in fact been developed and largely perfected around or even before the time I got my first train set, some 70 years ago! One of those sources I did not cite or excerpt in my paper is a publication I'm now reading, "Operating O and O-27 Trains, A Comprehensive Guide to the Design, Construction and Operation of a Layout for Lionel Trains" (which was published in 1964 but is largely a reprint of material from 1940!), which was commended to my attention by someone who had implemented an electro-mechanical block control system on his layout some 44 years ago!

Frankly, reading this book is like stepping into a time machine, popping the reader back into a world where "boys" (always boys!) are encouraged (always in the passive voice) to take over any available space in their homes and build miniature train worlds that would have bankrupted most families if fully implemented! I remember that same vibe from reading "Popular Mechanics" and "Boy's Life" issues (for the latter, both my own real-time subscription in the late '50's and early '60's, and my Eagle Scout uncle's bound copies from the late '20's and early '30's) as a callow youth. So, what am I trying to accomplish? Is "revisiting your youth" an available option?

And here's another footnote in the spirit of OGR being a discussion forum.

While the thread topic says "without the E unit reversing," what if you let the E-unit*** in the engine change direction but do it quickly so that the E-unit returns to the same moving direction.  So the idea is to press the controller's "Direction" button 3 times.  This can be automated using a low-cost timer relay module...about $3 shipped to your door from Asia or a bit more for faster delivery from a US seller on eBay or Amazon.

I did a proof-of=concept for automatic reversing of a conventional control out-and-back engine in this thread.  That required "automatically" pressing the controller Direction button twice to stop the engine, then send it back in the opposite direction.  If the same relay timer is programmed to "press" the Direction button three times, this should restart a stopped engine in the original direction.

No question this is a half-baked idea.  And yes the engine lights blinking on/off 3 times to advance the E-unit back to the original direction can be off-putting.  OTOH, there may be applications where this inexpensive configuration might be a good fit.  That is, you'd think there might be some applications where you want more than the motor starvation voltage on the track when the engine is stopped.

***Edit added for clarity.  By E-unit I am referring to 4-position types that cycle thru Fwd-Neutral-Rev-Neutral on each momentary power interruption.

@stan2004 posted:

And here's another footnote in the spirit of OGR being a discussion forum.

While the thread topic says "without the E unit reversing," what if you let the E-unit in the engine change direction but do it quickly so that the E-unit returns to the same moving direction.  So the idea is to press the controller's "Direction" button 3 times.  This can be automated using a low-cost timer relay module...about $3 shipped to your door from Asia or a bit more for faster delivery from a US seller on eBay or Amazon.

I did a proof-of=concept for automatic reversing of a conventional control out-and-back engine in this thread.  That required "automatically" pressing the controller Direction button twice to stop the engine, then send it back in the opposite direction.  If the same relay timer is programmed to "press" the Direction button three times, this should restart a stopped engine in the original direction.

No question this is a half-baked idea.  And yes the engine lights blinking on/off 3 times to advance the E-unit back to the original direction can be off-putting.  OTOH, there may be applications where this inexpensive configuration might be a good fit.  That is, you'd think there might be some applications where you want more than the motor starvation voltage on the track when the engine is stopped.

Honestly, I had to reread your post a couple of times to figure out what you were on about, but then the light bulb went on and I think I see what you're proposing.

I suppose your proposal might be useful in some situations where you are running exclusively three-position E-unit engines, and can program a single pulse to pause, followed by a three-pulse cycle to resume motion forward when desired, and are willing to employ modern circuit boards and electronics as control mechanisms, but in my situation 1) I'm running almost exclusively two-position E-unit legacy Marx engines, which have no "off" position available to use to hold the engine in place, and 2) I've already confirmed that a single sub-dollar ten ohm power resistor is capable (with the control relay and isolated center rail in the block section) of providing precisely the control I need, and is exactly the sort of simple, old-school electro-mechanical solution I was looking for (note: I ended up actually using four of those resistors to get the wattage rating necessary to survive a dead short, but one alone would suffice for routine duty).

In fact, I just used a variation of essentially the same technique to ameliorate another problem that cropped up as I modified the figure-8 on my test bed to more closely resemble the actual circuit on my main layout (an embedded figure-8): I found that the variable stop-and-go demands of the two engines on the weak power supplies I'm currently using sometimes resulted in over-speed derailments on the curve approaching the block section. So, I used four more of the same resistors (all in parallel this time, to yield 2.5 ohms) to reduce voltage on a single section of track just before the curve to slow the train and forestall derailments. Not that I can claim credit -- I got the idea directly from a boy's publication from mid-last century!

Steve

To add to my post on this topic one 2 trains on 1 track of 7-11-2022

"It also lead be to block every section of track on the layout except the 33 Marx switches.  Most blocks are 5  to 10 track sections long , allowing parking trains anywhere on the layout.  It was then simple to add a track section, with isolated outside rail on one rail to act as a switch ahead of the blocked section, to operate the two relays that control one of the blocks, with the wire going to that section to have the 5 ohm, 100 watt resistor in it, to not let the eUnits cycle.  In fact, each loop has one isolated track section and two blocked sections, one on each ends of the isolated track section after 5 or 6 sections of track normally blocked, so the 2 trains on 1 track can operate in CW or CCW directions."

To make adding a single DPDT slide switch to allow 2 Trains on 1 Track to operate the in each direction easier I have included the wiring circuit below.

Below is a circuit I used to wire the double throw, double pole slide switch that provides operation of trains in CW or CCW directions.  I had to look at the bottom of the control board to figure out how wired the DPDT switch as there was not an original diagram.  I was real smart in those days!

Charlie

@Choo Choo Charlie posted:

Steve

To add to my post on this topic one 2 trains on 1 track of 7-11-2022

"It also lead be to block every section of track on the layout except the 33 Marx switches.  Most blocks are 5  to 10 track sections long , allowing parking trains anywhere on the layout.  It was then simple to add a track section, with isolated outside rail on one rail to act as a switch ahead of the blocked section, to operate the two relays that control one of the blocks, with the wire going to that section to have the 5 ohm, 100 watt resistor in it, to not let the eUnits cycle.  In fact, each loop has one isolated track section and two blocked sections, one on each ends of the isolated track section after 5 or 6 sections of track normally blocked, so the 2 trains on 1 track can operate in CW or CCW directions."

To make adding a single DPDT slide switch to allow 2 Trains on 1 Track to operate the in each direction easier I have included the wiring circuit below.

Below is a circuit I used to wire the double throw, double pole slide switch that provides operation of trains in CW or CCW directions.  I had to look at the bottom of the control board to figure out how wired the DPDT switch as there was not an original diagram.  I was real smart in those days!

Charlie

Thanks, Charlie! I haven't had the chance to work all the way through your wiring diagram yet, but I think I see how you went about setting it all up, and how it's intended to work. It's amazing how all of us over the years have come up with very similar solutions to a similar set of problems and situations.

In fact, having solved the task I started out to explore (emulating the Marx "Twin Train" set functionality on a simple figure-8 track), and extended that solution to a closer version of the embedded figure-8 pattern of my layout, I'm now faced with a further complication.

Specifically, I noticed that, with the longer "loops" of the test bed embedded figure-8, it is possible for one train to catch up to the one ahead of it before the leading train enters the control section and triggers the block section to stop the following train. In other words, the following train can get too far into (or even beyond) the block section before the block section is de-powered, allowing the second train to continue to catch the first, even if there's not an immediate crossing collision at the intersection. Since there's no way to adjust the differential speeds of the trains electrically, my interim solution was to try different consists for the trains, adding cars and shifting them from train to train until the speed differential was minimized, but not entirely eliminated.

This is far from ideal, since even if successful, it would limit the equipment I could run on the loop, and careful tinkering would be needed every time the track is used (and probably from time to time even when in use). Moreover, the actual loops are considerably longer on the layout than on my test bed, so the effect of *any* differential would be magnified in operation.

So, my thoughts have turned to adding a more conventional block control system, much like you've outlined, to the crossing control system I've been working on.  Basically, my thought is to implement a trailing control block, one that will maintain an adequate separation between the trains even if there's a much larger speed differential. The only alternative I've come up with is to dramatically enlarge the length of the control section of the current design, resulting in an earlier triggering of the block section, and "closing the door" on the approaching train. I'm trying to think through the possible permutations that would have to be anticipated and controlled, before I decide which way to go. In any event, thanks for the additional information, Charlie!