Coincidence?

Since I am recovering from surgery a great deal of today was spent thinking along the same lines. My mock-up doesn’t look anything like a steam blackhead so that has been the search word of the day. Here are some simple backheads for inspiration:

Perhaps shape could be similar to the photo on the left with the throttle as shown and the JB mounted on the right. Since I am enthralled by the pressure gauge, it may simply plug into the top of the throttle with a phono jack so as to be optional. Bell & whistle on the side? Brake handle(s) similar to the ProtoThrottle would be an acceptable substitute to a stand and the high tech stuff to program or acquire a loco might be behind the firebox door - idk. 

What if the throttle was double sided to have the display on the back with selection buttons there as well? 

A question for steam experts would be how an electric light was turned on or off. Is it tied to the generator?

Yes

Questions, fleshing out the concept...

Wasn't the whistle a pull cord? I'm thinking that would be a simple thing to rig. Just a momentary contact switch activated by the cord. Maybe coming out the bottom or side? I'm pretty much convinced my gauges will be the separate unit but that's just me. Since I'm not talking manufacture, these can be "one off" designs but it will probably be helpful to establish some common ground. "Givens and Druthers" will be different for each person.

For instance, can we establish the SMA 30 WIFI throttle as the starting point for the actual throttle tech? Are their alternatives?

I'm very interested in operating steam and water gauges even if they are just for show. It used to be possible to replace meter faces and I assume it still is. Today's graphic programs and printers should make authentic looks fairly easy. What drives the gauges? I already mentioned my idea, is there a better way?

Hope your recovery is going well Neil!

I was stationed in B1 engine room on a diesel ship in the Coast

Guard.  I'm doing this from memory dating back to 1966-1968 when I was discharged.  The only steam we had on the ship was for the boiler and evaporator in B2 engine room.  We had sight glasses for Fuel and water levels in the boiler & evaporator.  In fact we had a fire in B2 while on an Alaskan fishery patrol when a cylinder head on the diesel generator failed and shot flame at the diesel sight glass on the boiler melting the sight glass, and igniting the boiler fuel before the genset could be shut down.  All of the pressure gauges in both engine rooms used a bernal (spelling?) tube driving a gear to operate the needle.  If I remember correctly we had to calibrate every pressure gauge every month.  All temperature gauges used a thermo couple in the water lines to an electric gauge.  We also had exhaust temperature gauges that used a thermo couple in exhaust manifolds to electric gauges.

I would assume that the gauges on a steam engine would be similar.  Since the ship I was on was originally commissioned in 1940, nothing was digital.  I hope this is helpful to you guys.

The auto spell check here doesn't what a thermo couple is or whether it is one word or two, and automatically changes thermo to thermos.

How does the engineer run the train?

I wonder how many modelers even have a clue how the controls on a steam engine work and what the engineer does with the controls to make the engine go?

On one of the other steam engine throttle threads they have been debating what the Johnson bar does and how to model it.  Would any modelers actually have any idea what to do with it?

Would it actually simulate a steam engine or would it just be a conventional throttle that was in a grey case shaped like a backhead, with rivets on it.

@Dave

I've participated in that thread and it's part of the reason this one exist. Nothing against what those guys are trying to accomplish but besides the price range being beyond what I can reasonably afford, they are talking about more prototypical interaction than I think I'm going to want to deal with. My suggestion to them that they may or may not heed was to strike a balance between look and feel. I think I'm aiming more for look because of my desire to keep things relatively inexpensive by using existing throttle tech like Geoff Bunza's aforementioned WIFI throttle or something similar. Also this is much more about DIY and that thread is about an eventual manufactured item complete with custom coding, enclosure...ie "the whole nine yards". I might like to have some steam throttle like interaction between a Johnson bar and the throttle but I don't even know it it's possible within my budget and parameters. That's a brainstorming thread, this is a brainstorming thread but I believe the end goals to be entirely different and I'm certainly not trying to take away from what they are doing. Not that anyone said I was but I do want to make that clear. I'm kind of hoping some of the same folks commenting there will give some input here.

Pull vs Rotate.

Hi all

So far the 2 examples shown are activated by pulling on a rod running through the boiler that opens a valve in either the smokebox or steamdome that lets more steam through.

However there may have been far more steam locomotives worldwide that had a lever that rotated the rod. 

Small industrial locomotive (right-hand drive)

On a mainline express loco (left-hand driver's position, Johnson bar/Reverser replaced with screw wheel ). Sorry for the camera shake but It was all I could do to get a photo under a cab roof that cut out most of the ambient lighting,

A pull-type on a US-built Davenport - Note the tri-cock set up as well - another give-away of a US-built locomotive. (right-hand driver's position)

The rotate-type may be far more simpler to build than a pull-type regulator in model form.  

One possible layout

When I was not long out of short pants, my local club had invested in some electronic DC controllers. These were transistor based, and also used resistors and electrolytic capacitor to simulate inertia (based on a hone-built version using a simple circuit). Setting the brake (RH lever) to zero (swing to horizontal position), one would open up the throttle to the desired “speed” and after a brief pause as the volts climbed up the loco would start moving and gracefully increase speed up to the setting. If you cut the regulator, the train would slowly decrease speed. The “brake” actually controlled the discharge of the capacitor which fed into the base of the driving transistor, so applying it increased the rate and putting it down to full would stop a train in 4 or 5 feet from 100% throttle, if the latter was simultaneously reduced to zero.

In practice, we tended to “drive on the brake”, using the regulator to set a speed, and then release the brake to varying degrees. Setting it at ¾ meant a slow start and steady increase in speed, whereas releasing it completely led to a rapid acceleration. In truth, the regulator was acting as a throttle on the volts being sent out to the track, and the brake was a load simulator which affected the charge and discharge rates - I suspect that full brake provided very little resistance to discharging the electrolytic capacitor, and it simply lost its charge more quickly that it could be replaced, leading to a rapid stop from a running position, or no movement from stationary.

What I liked about it was the ergonomics. I would now align the direction lever to a forward (up) and reverse (down), but that’s only because I gave become used to DCC and radio control, where left-right is no longer an appropriate choice. Notice also that the simulation could be switched off completely: very useful for testing and sometimes yard shunting. With some change and thinking, it might be possible to tailor this to what you want. It could also be useful for part of the steam Proto-throttle discussion, but there are too many agendas running in that thread!

Thanks

All of this is exactly the kind of "food for thought" I was hoping on getting. From pictures of backheads to that description of a hardware solution. I can follow what's happening in that type of circuit but if it's possible to transfer it to the three wire pot inputs of the WIFI throttle or if it's more easily doable in software these days is not easy for me to discern. I can still follow a circuit diagram and build most anything from one but I must admit to being quite rusty on my electronic theory and I never was much at design and engineering. My software skills are worse. I can build PC's from cases, mobo's and cards but programming is another matter entirely. That's the guy you're dealing with here so better minds than mine are needed!

Hmmm...

Simon's idea looks to be a good one, and one that could be "good enough" for me with just the following actual controls:

Throttle lever as per above.

Adding an adjustable momentum potentiometer w/knob.

Brake lever as per above.

Adding a "Johnson Bar" lever as above BUT this one is keyed to the sound decoder's chuffs to increase volume and "cut-off" of the chuff sound. (I think such a feature is possible with certain TS2 and TCS steam decoders?)

This would offer enough complexity to give a feeling of heft when operating, and with the independent control of the chuff feature, moving the "Johnson Bar" farther to its extremes produces no appreciable chuff (drifting), light chuff (as if running light), medium chuff (cruising), all the way to a laboring exhaust sound (starting a train, working a grade/etc) when such sounds would be appropriate without regard to the momentum setting.

FWIW.

Andre

Great thread

Guys,

its great to see this tread fire up.   I will follow it with a great deal of interest.

Mark

Things to notice

Hi all

Some things (generalisations) to keep in mind:

• the bigger the locomotive the shorter the length of the gauge glass. Check out the photo of the UK mainline express locomotive.
• backheads may be fitted with backhead injectors - see the UK express loco and first photo of Hudswell Clarke loco "Fiji" These backhead injectors were more common on UK and Continental locomotives
• For the more usual auto-positive injectors used in the US, there would only be a steam valve for an injector on each side of the loco. The water valve for the injector would usually be somewhere on the water feed pipes coming from the tender 
• tri-cocks were more of a North American design feature. UK designs went with 2 gauge glasses pretty early on after a couple of boiler explosions on locomotives equipped with a single gauge-glass.
• The UK and European railway systems were a lot more regulated earlier than in the US.
• Braking systems in the cab - Air brakes vs Vacuum brakes - UK prototypes stayed with vacuum brakes.  Check out the mainline express photo for an example of a vacuum brake control valve.  Westinghouse had the braking monopoly when it came to the US/Canada. Australia had a mix of systems and sometimes even both Westinghouse and Vacuum on the one wagon. I'm unsure about who ran what braking systems in Europe. Industrial locomotive sometimes were fitted with a steam brake that applied boiler pressure steam to the brake cylinder of the locomotive.The cab valves for each braking system look significantly different.  
• For US prototypes, anything with a grate area bigger than 48 square feet had to be mechanically stoked,  These stoker controls were also in the cab. This didn't apply to UK and European prototypes. Even some of their bigger locomotive were hand-fired, but even the UK's largest 2-10-0 "Evening Star" only had a grate area of about 40 square feet. 

That ought to do for this post. 

Another design!

This used to be made, but the company went south:

It might not be clear, but there is a reverser/Johnson in between the rotary dials (used for selecting locos, changing turnouts, etc, as per CVP’s “Easy DCC”).

This does lack a brake lever, so that would need to be added - personally, being right handed I like to have the controller in my left hand so that my right hand is free to uncouple and operate manual turnouts (reflects my prototypes of interest) so I would move the throttle to the left slightly, and have the brake arranged vertically, to the right of it. I would also add a sprung lever for sounding the whistle.

Re the reverser and inertia.

Several ideas here. On the prototype, the regulator controls how much steam is released, and the reverser controls how it is used, i.e. for what percentage of the piston stroke is power applied. These are analogous to setting the volts and the width of the pulse in the control of the motor - normally, it is full volts applied for portions of that pulse: if we were working purely electronically, we could use the circuit to do exactly that, and the exhaust sound could be adjusted to match, so that each beat varies in length and also volume, creating a “bark” when both regulator and reverser are fully opened, but cutting back as the train gets underway with very little sound when the reverser is notched back when coasting at speed with just sufficient steam to maintain velocity. Until the train has been brought to a standstill, putting the reverser into reverse has no more effect than putting it to zero, to cut off steam, and also virtually all exhaust sound.

If working digitally, then we can be more sophisticated and use the regulator to set the maximum speed, say 50/128 steps, and the reverser can be used to set a rate of acceleration - fastest for full forward/reverse, slower for less. (If you wish, these could also be used to simulate use of steam, so that leaving the loco in full gear at full speed uses up more steam than can be produced!) But the rate of acceleration also needs to be tempered by three more factors:

1.) The tractive effort required at the drawbar to move the train on level track, usually referred by railwaymen in the U.K. at least with reference to a load factor based on a common unit, modified to suit specifics, e.g. “train of 6 with a load factor of 8” - six for the length, 8 for the load;

2.) The effect of gradients on the load: can the train get up hills unaided?

3.) The power (tractive effort at drawbar) of the engine.

Various ways to express this, but for our purposes we might consider expressing it in terms the load factor, e.g. 10.

Taking these in reverse order, the power classification can be entered as a CV for the loco, and if consisting, these are added together to provide a total power for the consist. For banking, there would need to be a way to temporarily combine the loads of both the train engine and the banker whilst retaining separate control. The measurement of back emf (BEMF) of the motor could be used to derive this, so if you don’t advance the throttle and/or reverser to compensate, the train slows down. At the moment, BEMF seems to be used to do the compensation for you, or gets switched off. I am suggesting something slightly more sophisticated...

Finally, a CV that is adjusted “on the fly”, to enter the load of the train: this is applied when the loco/consist is coupled to its train, and released when the loco/consist is detached.

Then we bring in the braking factor: more of an issue in the U.K. where trains ran without continuous automatic brakes a lot longer than elsewhere, but even so, a train driver has separate controls for engine brake and train brake. 

This can be done with two levers, although I suggest that it might be simpler to only “apply” the loco brake for the first 25% of of the lever movement, then to gradually apply both. All this does is change the rate of deceleration applied to the speed of the train. We could also enter a CV on the fly to reflect the proportion of vehicles with power brakes, or connected power brakes (e.g disconnected when switching), such that when there are no power brakes connected to the engine, deceleration is much increased.

This requires one extra CV to be entered with the train consist information: to reflect the brake effectiveness, so I might enter a load factor of 8, brake factor of 8 (fully air braked train), load of 8, brake factor of 0 (switching/unfitted train) or load factor of 8, brake factor of 3, (partially fitted train, or a reflection of the fact that vacuum brakes are less powerful than air brakes).

We don’t then need to vary the inertia simulation with a potentiometer: it is all done for us.

Many real roads had their own power classifications for engines, which looked at various features including factor of adhesion and tractive effort. This could be reflected in a simple software table used when setting CVs. For example, I might use “class 1” for loads up to 10, “class 2” for loads from 11 to 16, “class 3” for loads up to 25. Anything above 25 then requires more than one loco, or some more powerful locos to be acquired - just like real railways.

Incidentally, not all U.K. railways used vacuum brakes. Some did opt for the Westinghouse air brake system. (And as a consequence, many locos and stock gas to be “dual fitted” for running on different lines or with equipment on through trains. This was one or the other on a train.) I mention this as for Westinghouse brakes, I would like to see a minimum of 3 settings for the sound: off, “run” and “recharge”.

That little (?) lot would satisfy me, but there is no reason why more couldn’t be programmed into a software interface for those so inclined as to add features for the water level and feed, frequency if firing, etc.

Another typically long post from me. Sorry, but sometimes ideas cannot be reduced to a “ten word soundbite”.

throttle vs Johnson bar

i had had the impression that there was a throttle and reverser/Johnson bar.   But from another thread on horsepower v. acceleration, i realize that there is no separate throttle.

i'm told the Johnson bar determines "cutoff".   More generally, it determines for how much of the stroke steam is allowed to enter the pistons.    The longer steam is allowed to enter, the greater the sustained force and horsepower applied to the drivers.  (similar to how fuel intake is determined by how long injectors are open on a constant pressure automotive fuel system).

but that same steam on the one side of the pistons needs to be pushed out when the piston is forced in the opposite direction.   So there's a need to limit the amount of steam filling the piston at higher speeds when accelerating.    And of course, there's no need to for excessive force when maintaining speed, cutoff will be shorter.

if the above is correct,  a Johnson bar on a steam Proto throttle would behave much differently than a simple speed control knob on a model railroad throttle.   You might turn it up all the way to get the loco moving, have to turn it down to maintain acceleration as speed increases and finally turn it further down to simply maintain speed to mimic the horsepower needs at a train gets up to speed.

it might be worth having separate threads for discussing what a steam throttle looks like v.  what the electronics do to simulate the behavior of a steam engine including the water level in the boiler, firebox heat production and the anticipation for steam requirements

My impression as well

A "throttle" and a Johnson bar working in concert to move the locomotive. If that's not the case then I suppose  I need to do more research but the unit I propose doesn't need to go too far down that rabbit hole anyway as I fear it adds to complexity and cost. If we can produce a reasonable facsimile of the Johnson bar and/or throttle effect without a lot of complexity then I'm all for it. For my version of this at least, whatever is done needs to end in a three wire pot connection to the WIFI throttle. Overly complex circuitry or more than a very limited amount of custom coding are non starters in my opinion. In fact, I would prefer no writing of additional code at all unless someone in the know is willing to do it for free and make it available to everyone because I certainly don't know how. It would also have to be using an off the shelf item such as an Arduino. Doing that sort of work for free would be a lot to ask of anyone so a feasible hardware solution is better if it's even possible.

Johnson Bar

CORRECT. The throttle and Johnson Bar work IN CONCERT.

The Throttle/Regulator determines HOW MUCH steam goes to the cylinder valving.

The Johnson Bar/Reverser determines HOW LONG (cut-off) said steam is passed to the piston.

MORE length of cut-off equals MORE steam passes through the valve to the piston, thus can produce more power. The two, throttle and Johnson Bar, MUST be used in concert to coerce the best performance from a steam engine.

FOR MODELING PURPOSES:

Seems to me that trying to digitally incorporate the interaction between the throttle and Johnson Bar would be quite complex. Thus, I would be inclined to only represent the AUDITORY differences brought about by throttle/Johnson Bar settings, load, labor, etc, and use digital momentum to represent weight/load behind the engine.

FWIW: I was a paid employee that worked extra for a tourist outfit that ran steam engines and I ran steam often for several years back in the mid-late 1980s, thus I'm quite familiar with the functions of said steam engines as well as how a steam engine sounds under various conditions. My above explanation is simplified, but at least gives you an idea.

Andre

Andre

If you need to control AUDITORY only differences then in my opinion you need decoders that can separate sound from speed as an option.

Mark

Mark...

I figure you are correct, but I was in reference to a steam throttle with levers as in the above example.

I intend to use TS2 decoders in my steam engines and manual notch using my NCE handheld when I can. But it would fun to have a hand held with levers and sound/etc!

Andre

Steam locomotives and DCC Decoders

Everytime I look at the idea of a "steam" throttle, the same thing comes to mind - DCC decoders aren't actually built to do steam, at least not yet. 

Since controlling the speed (and direction gets mixed up in there too) seems to be a function of at least two controls, the "throttle" paradigm that DCC uses to convert AC into graduated DC voltage for the motor just doesn't quite work - somehow you need at least one other variable control, like the throttle, that would influence the speed.  And this really can't do as a CV, since I don't know about you, but the thought of my command station having to constantly re-write a CV on the decoder gives me the willies. 

Jeff

Thinking about getting a steam loco to act like a steam loco makes your head hurt.  But in my opinion it is possible with a lot of effort.

My head hurts now.  Hahahaha

Mark

A few random thoughts for the mix

So... what specific behavior do you want, within the constraints of DCC reality? If you can succinctly describe this, you can likely build it.
Have fun!
Best regards,
Geoff Bunza

Random thoughts

First of all, I'm happy to see Geoff show up over here! It's the WIFI throttle that gave me a lot of the impetus to start this conversation. As already stated, the discussion of the steam proto throttle seems to be getting more complex and expensive than what I can think about so a DIY solution that is customizable like Geoff has brought us is music to my ears. Nice to know it's capable of doing more if someone has the knowledge. Welcome Geoff and thank you!

I'm already about 99% convinced that my steam look throttle will be two parts so I may go radical when I get around to building it. I'm talking a stationary panel (back head) that will include a steam pressure gauge, a water sight glass and a firebox! All simulated of course. My eccentric creative mind is already brimming over with ideas on how to accomplish it.

Of course, as crazy as I can be, limits have to be set. Just as I recently posted on the other discussion, at some point there has to be a can and can't do/have.

Also, on the other thread there was the suggestion of 3D printing certain parts by Scott of Proto Throttle fame. I don't know why I hadn't thought of that but I hadn't. Once again my mind brims over with possibilities. Oops, there are those pesky limits  I just talked about!

At the risk of repeating myself...

Yes, decoders are not currently really designed for steam, but, it is possible to work around it.

Use the throttle to set the speed, and the reverser to dynamically adjust the “inertia simulation” for acceleration: fully open, decrease the value in CV03, notch batch on the reverser, increase CV03. Optionally, combine this with a bit of extra information stored in the control system, reflecting the load (in terms of axles, tons, or what have you) and the power rating (how many axles, tons, etc the loco can pull) and a very simple program will adjust and constrain CV03 and CV04 (a load close the maximum load for the loco will require more time to slow down as well as accelerate). It can be made more or less complicated, and combined with dynamic exhaust management via back emf to adjust the round, can easily be made to behave similarly to a real loco.

At the risk of repeating – Some Big Issues

Requires a not so simple program that is specific to each type of decoder(s) and the CV settings and ranges (Max / Min, endpoints, etc.) of said decoder(s), and programming on the main (POM) which also presents lots of additional problems. You could upload some of this information from a JMRI Engine Roster, but a lot of parameters would still be missing.

Since all of this dynamic exhaust management happens in the decoder, how does the throttle know what the decoder is doing. Humm! All those messy or missing CVs and parameters. I have synchronized exhaust (chuff) sound on most of my steam decoders, so the only method of having sound similar to real loco[s] would be to also change the chuff timing and sound CV’s by POM. Again the throttle would have to know the type of decoder(s) and the CV settings and ranges for each decoder(s). Definitely not simple programming and throttle support / management.