Dang. Now if you could only hunt down the guy in Chesapeake, Va. that has "K4PRR"....

[But, really, this is no different than the challenges steam and diesel guys face. Safety valve, driver tires, staybolts, or cylinder rings worn out? Guess you gotta give up and call it a day, eh? Nobody can get THAT stuff anymore, right?]

One huge difference: A competent machinist can make any of the above parts starting with a piece of metal. Or he can make a wooden or plastic pattern, have the part cast, and then machine it. I would like to see anyone make a replacement electronic integrated circuit in any railroad shop anywhere in the world. Very few electronic shops could do it either.

There are problems there, also... I seem to recall Andy Nold's Birney truck project had to go with cast steel parts, because the correct malleable iron was simply unobtainable. But these materials are at least understood and competent machinists can work with them, and have a good idea that the substitution won't be prone to failure.

The problem with turning an electronics hobbyist loose to revamp rail vehicle controls is one of understanding... They may come up with a working solution that may not adequately consider all potential failure modes. Meanwhile, the preservationist managing the project may (likely) not truly understand what the hobbyist has built. An example, just off the top of my head, is the choice of "sourcing" or "sinking" inputs for the logic. They will both work, but bad insulation on the wire to a sourcing input will blow a fuse somewhere; bad insulation on a sinking input will yield a false input.

Sandy: Don't think that I didn't check up on "K4PRR" before choosing the K3PRR callsign. That man has had the K4 call for a very long time and I wouldn't have even considered bothering him about it.

But I was surprised that no one had K3PRR. About a week after the FCC granted it to me I received an intense phone call from a chap in Altoona who felt that he'd had some claim on it. But his application went to the FCC several days after mine. First come, first served.

Or maybe just down the road from Cleveland at NORM, which would be more fitting.

There are several different levels of "electronic control".

You have discrete components, i.e. a bunch of resistors, capacitors, inductors and transistors soldered on a board. You can unsolder and replace components that have failed. Radio Shack and electronics houses have the parts and they are usually identified by a simple rating system (farads, Ohms, PNP or NPN, etc.) It is often easy to tell when one is burned out as the capacitor will swell on top and the resistor will scorch.

Next step is integrated circuits (ICs.) they are the little black rectangular components that look like centipedes. The analog ones are like the discrete components above but several are imprinted on one chip (usually transistors.) The digital ones have logic gates for very simple binary computations. They are usually mixed with discrete components. These are still pretty basic building blocks. They have a model number on them you can look up in a semiconductor reference and find out what is components are on it the and how the pins are wired. If you cannot find that model any more you can surely find one that does the same thing, but you might have to do some fancy wiring if the pinning is different. These will often have irregularities on the plastic case when they burn out but not always. Often they can be replaced with discrete components, if you can find them with the same ratings.

Next step is PLAs or Programmable Logic Arrays. The have the equivalent of several digital ICs in them and they are "programmed" to connect the leads of the digital components in a certain way (no actual code here.) It is on a chip larger than an IC (it actually IS a large IC). The programing may be factory only or field changeable. The trouble is, the same model of PLA may have been programmed in any one of thousands of different ways for different applications, so you can't just sub in the same model. If it comes from an identical board it may work. It is also getting quite a bit more complicated to tell if the component is even doing what it is supposed to, especially if you do not know how it was programmed.

Next step is a microcontroller. These are basically tiny basic computers, and hence all digital. You write code in some low level language like Assembly, compile it and use some programming device to put the code on the it. Physically it is an IC larger than a PLA. If it loses its code it takes specialty equipment to put the code back on if it is lost. Unless you have a cozy relationship with the manufacturer you won't have the code anyway. They are getting really hard if not impossible to replace the component on the board unless they are in a removable socket. Without electronics training in that model and specialized equipment it is virtually impossible to tell what it is doing or if it is doing what it is supposed to be doing.

Next step is the microprocessor. You are using one now to read this. They are expensive and complicated and you can forget about reverse engineering one unless you are a PHD with an electronics lab. Fortunately they are usually easy to replace with interchangeable sockets. There is usually no proprietary code on them as they load it from an outside source like ROM, Flash or magnetic disk. The code they do load is a whole nother ball game. Make a backup of it somehow or get a copy from the maker while they still have it. The nice thing here is that you might be able to swap it out the entire system for a different microprocessor based system coded by you to do what the old one did (Emulation) and some kind of interface converter. The trick is figuring out what the old one did, exactly. That requires the involvement of the manufacturer (of the device, not the microprocessor.) or a software company that deals in that sort of thing.

Point is, there are options on electronically controlled equipment. However, you are NOT going to make this stuff from a hunk of silicone dioxide or write the code from scratch yourself. You have to use spare, surplus or substitute parts. If you get it wrong, it is not going to make a noise to tell you so before it burns itself or something else up. And most importantly it is going to require an entirely different skill set than what most museums posses today.

All that having been said, I have no interest in restoring electronic controls. I support computers 40+ hours a week and that is quite enough. If it were me I'd backdate equipment to operate on older, more repairable technologies, even if it operates less optimally. After all, how many museums push their equipment to the edge anyway? I'd rather drive a car with a distributor and carburetor that I can fix and tweak with basic tools rather than inserting one of a hundred throw-away parts I might or might not be able to source (and you never seem to know which one is faulty.) When modeling I usually shun anything with electronics in it, favoring sixty year old mechanical devices and electrical principals. Just seems to have more longevity in it.

I don't go to the local electronics wizards to get staybolts or fix air brake valves, either.

I'm sure that, fifty years ago, there were plenty of professional railroaders that said "Ha! No WAY those backyard wannabe rail barons will be able to do the work required to keep those coaches/trolleys/steamers running for years! As soon as something breaks on that thing, they'll have to call it a day or have horses pull it!"

Backyard mechanics these days aren't adjusting carburetors (well, some of them are). They're dropping in new onboard computers and reprogramming the onboard diagnostics systems to "soup up" their sports cars.

Yes, it's entirely possible that after only a couple years, a new owner of a San Diego Red car might not be able to operate it, and they'll be stuck with a "hangar queen." But it's not a foregone conclusion. And some of the field have been dealing with this kind of "parts obsolescence" thing since the first attempt to regularly run a "preserved" diesel locomotive.

ebtrr names some issues that you'll face dealing with legacy equipment, but he also says some fairly scary (and inaccurate) things about the challenge of creating ones own modernizations. PLA's are obsolete and microprocessor control is unneeded - a common, off the shelf, open-source, $30 Arduino contains all the computing power we'll ever need or use. Add a WiFi module so your stuff can hop onto the shop's WiFi and dump its diagnostics every night to a web site.

You don't need assembler, they're making this stuff in ever higher level languages that anyone with an interest can learn and write in, in a couple of days of actual interest. Easily. CHILDREN do these sorts of things as throwaway projects for fun.


A similar miss of the point. Nobody's going to fab a replacement for a single IC, and even if they DID work in Silicon Valley at one of the chip manufactories, the processes and chemicals would be different (due to heavy metals, chemical restrictions etc.) and it would not be wrong at the molecular level and not an authentic reproduction. Nobody's going to do that. They're going to replace a module or a system, which is a great deal easier since it has fewer and simpler inputs and outputs.

If historic fabric is paramount over operational practicalities, then you need to make it a static display. But then, you need to make everything a static display because you also replace wood with different species non-old-growth, epoxy-dip motors, weld on steam engines, etc. etc. You always have to make compromises for it to move.


Your electronics guy should be internal to the organization. There are plenty of electronic geeks who are railfans. Electronics requires intelligence which also suits being a railfan or preservationist. If you don't have any electronics guys in your group, you may have a recruitment problem, which may be directly related to attitudes. Nobody on this thread has displayed an actual working knowledge of the state of the art. I doubt many of you would be willing to actually engage your brain cells long enough to build a Joule Thief let alone a Lego Mindstorms kit. Your teenage kids could do circles around you in this field, just because they are not impeded by mind-sets.

As far as failure modes, please recognize that design is an iterative process. That's what fuses and meters are for. Also what design standards and punch lists are for. And testing. An electronics project is not above being managed in the normal way.


The answer is once you have a working design, you make 20 of them, test them all and store them for future use. Obsolescence is not a concern until you have consumed the supply, and if you're blowing 20 of them, the design needs revisiting anyway. In this day and age, electronic circuits ought to self-protect.


Wring hands all you like. At the end of the day, the choices are two:
1. A cold dead static display faithful to historic fabric; or
2. Living history, delighting and educating the public.

As you yourself point out right here, electronics are renewal parts, just like wheel tread. They don't go obsolete just for museums. Looking at preserving BART A/B stock, there's not a car in the system that has original electronics. And I certainly hope they've had electronics upgrades since the early 90s rebuilds. The Boeing LRV's lived an embarrassingly short time and they too got major technology upgrades. Near their retirement they literally ripped out a whole row of seats to install a control box. Now what do we do about that???

And even then, much of what fails in electronics is the same old thing: capacitors. Ask any old "tube man" about that. Now, you couldn't tell the difference between a 1976 capacitor and a 2011 capacitor unless you were in the capacitor business or have the unit loose so you can read the tiny numbers on the case. So right there we can fix a lot of stuff. Scorched resistors are another thing I see a lot of. Other than that, silicon electronics do not fail that often unless the capacitor failure killed the silicon, or somebody mishandled or misapplied it.

And that's the enemy here. Ignorance. Take the ever-complex GE 44-ton voltage regulator, which must balance up to two auxiliary generators. There's the usual old clackity-clack regulator with a bias winding included to balance them, plus a reverse current relay... times two... on the unit, there are probably 15 adjustments many of which affect each other. But each module comes off, and its 2-3 adjustments can easily be made on a bench. By which I mean a (non-metal!) desk with "bench power supplies", some meters, soldering irons, a stock of wire, resistors, capacitors and odd bits, maybe an oscilloscope. You'd run diagnostics and do repairs there. It's a simple thing, museums just don't have them. And so they lack the facility for people to develop the skills to work with this stuff. Old or new.

Gee, maybe it's time to actually develop courseware on this stuff.

I know several electronics-wizard railfans who would be godsend for anyone dealing with this stuff. We routinely run into each other at hamfests, train shows, computer shows, trackside, online, and the like. The geekery is pervasive--I see them at the same beer fests, the same music fests (working the sound gear), the same sci-fi/anime fests, etc.

All of them have, at some point, been "turned off" by their "local" rail museums, preservation outfits, etc. I have to cash in personal favors to get help from them for the rail preservationists.

We have a problem. But we already knew that.

I've popped this response out into its own thread.
viewtopic.php?f=1&t=34887

Of course PLAs are obsolete. So is solid state, DC traction and steam locomotives. I'm not proposing it as a solution, rather as a potential problem if it was found failed in a piece of equipment. But, whatever. Obviously your electronics knowledge is more up-to-date than mine so I will happily defer to you on all electronically controlled rail equipment - as I said I want no part of it.

I've read the posts to the initial post with considerable interest. I've found all of the comments to be at least interesting and thought-provoking.

I'll try and offer feedback on a few comments and questions:

- Regarding the questions about size and turning radius, the U2 can handle 82' radius curves. The tightest curve at the Rockhill Trolley Museum is 90', which has been successfully navigated in years past by the Liberty Liner Independence Hall and DRJC #1009. Width is not a problem either, as the museum regularly runs the 9' wide former PSCT/NJT PCC Car #6, which is wider than a U2.

- There was a lot of thought and planning that went into (and is still going into) this acquisition because it is so very different from anything else in the collection. One of the factors considered was the technology; while more advanced than a PCC, is not so advanced as some of the most modern LRVs. The sustainability of this vehicle into the future has been a major part of the planning process from spare parts to technical know-how both within our organization and outside. Part of our maintenance staff includes volunteers whose "day job" is maintenance of LRVs of the same vintage. We do also have some in-house knowledge of solid-state electronics; for example a SEPTA GOH PCC car has a solid state voltage regulating system that we have done some preliminary reverse engineering on (through an individual who is not a railfan but who works in the electronics field). To infer we have it all figured out would be grossly incorrect. But we continue to look into ways to ensure this vehicle can be maintained as a future operating exhibit. I know our museum went through a similar knowledge gap decades ago when its first PCC car was acquired. We have come a long way in this regard and I am optimistic we will do the same with the U2 in due time.

- At least one post discussed how infrequently some of our heavier and high performance cars are operated. Granted, vehicles like PCC car #6, P&W Cars #10 and #402 are not part of the "bread and butter" cars in our operating fleet, but these cars are operated more than some may think, particularly #6. #205 has gotten some exercise as of late for Operator Training and #402 gets used periodically to do what it was built for - pull heavy loads. The U2 will likely not be an everyday vehicle, but will likely see similar frequency of use as these vehicles.

- Liberty Liner Independence Hall got mentioned several times. This train is a power hog, but has seen some use. In fact, it may see more use soon thanks to some recent progress. You can read for yourself at: http://www.rockhillcarprojects.blogspot.com/.

Best wishes,

Matt Nawn
First Vice President
Railways To Yesterday, Inc.