Sw1500 kitbash from Atlas mp15dc origins


So there are lots of things I like in the realm of trains, but if I was to narrow things down regarding what to actually model, I’d lean toward the era starting with 2nd generation EMD products. Of the three common EMD 2nd gen switchers - sw1500, mp15dc, and mp15ac, I’m most interested in the sw1500. I’d rank the mp15ac second, and the mp15dc third. Atlas produces the mp15dc, presumably because they wanted to share product development across three scales, and they didn’t want to go up against the Athearn sw1500 in HO. So I would need to either scratchbuild or kitbash to get a high quality model of the sw1500.

Once I learned more about the prototype loco, and what was available as a starting point in O scale, I discovered I could kitbash an effective model by blending superstructure parts from the Atlas mp15dc with drive parts from the Atlas sw8/9/1200 (referred to as sw9 going forward for simplicitiy) series switcher. At its simplest, the kitbash could be achieved in three steps: shorten the frame, modify the hood, and swap drive trains with the sw9. I came to realize, right about the time I started cutting things apart, that there were additional issues that I didn’t foresee, but even then the simple kitbash would be straightforward.

At the other end of the spectrum, there are a lot of opportunities with the base model to improve detailing and accuracy. Of course, detailing can lead to a slippery slope, one which I currently sit comfortably at the bottom of. By the time I started this project, I’d decided to adopt p48 standards for track and wheels. I also didn’t foresee having a layout anytime in the immediate future. So I decided I would be a model builder instead, so I could stop sitting around planning things and start actually modeling something. Adding to these two conditions, I’m also very picky about accuracy and detail - yeah I’m one of those guys :roll_eyes:. As a model builder, with a finite project scope, and a disorder, it made sense to push the detail on this project to the furthest extent possible. I’m fortunate to have some capability with 3d modeling - actually it’s my best modeling skill at the moment - so I could create components from scratch as well as modify or utilize parts already available on the market.

Obviously, another modeler attempting a similar kitbash, or an sw1000, could pick any point on the spectrum to aim for. For the purposes of the initial project description, I will ignore all the extra work, and focus only on the bare minimum changes required for an sw1500. Once I have the basics outlined, I’ll describe how to go off the deep end.

Everything necessary to complete the kitbash has already been produced by Atlas, either as part of the mp15dc line, or as part of the sw9 line.

The mp15dc is essentially a stretched sw1500. It has a longer frame, different trucks, and a longer hood. But many of the styling cues and details are the same for both models, so all hood, cab, and frame parts from the mp15dc can be modified with minimum effort for the sw1500.

In a happy coincidence, the sw9 and sw1500 prototypes share the same bolster center dimensions and the same truck wheelbase, plus Atlas makes both AAR Type A and Flexicoil switcher truck sideframes used on the sw1500. In another coincidence, the Atlas sw9 and mp15dc share the same motor, motor mount, gearbox design, and gearbox mounting. The only distinctions between the two drives are truck wheelbase (AAR Type A/Flexicoil = 8’-0", Blomberg = 9’-0") and universal drive shaft components. Therefor the trucks and motor from an Atlas sw9 are a drop-in replacement in a modified mp15dc frame.

(1) Atlas mp15dc model - this can be either powered or unpowered, 2-rail or 3-rail, and really the drive components aren’t needed at all. From the frame up, including the fuel tank, will suffice. One thing to consider is that a 3-rail frame will have a larger coupler opening in the pilot. I don’t consider this to be an issue because it can filled in fairly easily (and will have to be filled in any case if a scale pocket is used). NOTE: there are three versions of the mp15dc model, each having a different condition for the hood directly in front of the cab. Two versions have boxy housings (square and tapered in Atlas parlance) sticking up from the hood for paper air filters. There is a third version without filter box (labeled “standard”) with a traditional sloped hood course. The standard version most closely resembles the sw1500 hood, but in reality the sloped section on the mp15dc is longer than the corresponding section on the sw1500, and will need to be replaced. Any of the three hoods will work, but the standard is probably still the best bet because the area forward of the sloped section most closely matches the desired profile. At a minimum, BNSF, BN, Frisco, Alaska models all have the standard hood. There may be others. The undecorated version currently offered for preorder from Atlas is also the standard hood version. However, nobody knows when these models will be produced, if ever, so it would be best to look for a donor through other sources.

(1) Atlas sw9 model - actually the only parts needed from this model are the trucks, motor assembly, and any associated mounting hardware. Again, this can be 2-rail or 3-rail, the same parts are used in either version. NOTE: the motor assembly from the mp15dc cannot be used with the trucks from the sw9, even though it appears the same . The diameters of the attached telescoping drive shafts are not the same.

I’ll take a break for a while, then start back in describing the frame modifications. Stay tuned.



The frame for the Atlas mp15dc is actually a wonderful candidate for shortening (or lengthening for that matter). There are flanges along both sides of the frame, just inside the frame rails, running from the fuel tank to both end sheets. All of this flange area is free of interference with the swing of the trucks. The area formed by the flanges and side rails provides an ideal location for splice plates or bars, and the flanges can easily be drilled and tapped for screwing everything together. All connections will need to be mechanical, as soldering diecast won’t work. In the following photo, you can see the flanges marked in blue. Except for the groove in the triangular wedge at the top right corner of the frame, all the flange areas are original. The material is removed in the top right corner to accommodate the splice bar.

There are two primary areas where the mp15dc frame requires modification. Of course, the major operation is removal of material to shorten the frame. The second task is the removal, in whole or in part, of the diecast fuel tank mass. There is a third set of operations, but it is comprised mostly of optional work. There are various bits and bobs that serve no explicit purpose going forward, so they can be removed to clean up the frame or provide additional space for electronics installation.

I would actually recommend doing this step first, just to get the fuel tank out of the way for other operations. There is a large mass of diecast material in the center of the frame, that fills the cosmetic outer fuel tank casting and adds a good deal of weight to the frame. At least some of this material must be removed, as the factory fuel tank is about 2’-0" longer than the sw1500 tank. I decided to remove all of this material to the depth of the bottom of the frame rails, so that I had a clean slate for detailing the fuel tank and air reservoirs. But it is possible to only remove about 1’-0" from either end of the tank. The prototype tank is 5’-6" long, including end sheets, so enough material will need to be removed to fit inside that dimension. I think it’s simpler to remove the entire tank. It also turns out that removal of weight isn’t an issue, as the new modified frame actually weighs more than the original thanks to the additional brass.



Modifying the cosmetic fuel tank cover will be a topic for another day.

With the fuel tank cleaned up, work can begin on adjusting the length of the frame. Shortening the frame to the proper length for an sw1500 will require cutting and removing material in four places. This is partly because the bolster centers need to reduce from 24’-2" O.C. on the mp15dc to only 22’-0" O.C. for the sw1500. And it’s partly because the distance from the bolster to the end sheet also needs to be reduced, from 122" to 111". Note that two cuts are required for the bolster to bolster adjustment, as a single cut from the middle of the locomotive would affect the motor mount spacing.

The yellow lines indicate the approximate locations of each cut. The positions can vary somewhat, but I wanted to avoid cutting through the large chunks at the back of frame, and I wanted to avoid removing the low tabs at either end of the motor, that act to prevent the shell from deflecting inward.

The front end, after cutting, removal, and squaring:

A similar operation is performed on the back end of the frame, and additional material was removed from the bottom left hand corner to make room for the splice bar.

At this point, the frame can be spliced back together. I had originally intended to use 3/16" square bar stock for the splices, but 1/4" bar stock was on hand, so that size was used instead. My concern had been about seeing the bar stock behind the frame rail from a particular angle, but in practice this isn’t an issue. 3-48 screws were used, again because they were on hand. 2-56 screws would probably suffice. Everything is clamped together, drilled, countersunk, tapped, and then screwed together. A nice bonus here is that a single piece of bar stock can run from the fuel tank to the inside of the step well, achieving two splices at once, resulting in a very solid and robust frame.

Back end

Front end

From the top side, screws poke through into the recessed groove that the shell sits in. They were cut down flush with the bottom of the groove so that the shell would seat properly.

Once the screws are cut down, they can be backed out one at time and fixed with Loc-tite. I don’t anticipate an issue though, so I didn’t bother with this step.

The net result - the sw1500 frame is 4’-0" shorter than the mp15dc frame, which is evident in the last two photos.

These last few photos will deal with the handful of modifications that are either completely optional, or are only required pending other detailing and the electronics installation decisions.

Cab Sub-base - this material can be removed down to the top of the frame. All this volume under the cab sub-base can be used for lighting or connector boards, or for a keep alive.

Safety tread - I would consider milling off the safety tread, and then laminating new custom etch safety tread in its place. Doing so would cover the joints in the frame. It would also allow for drilling and tapping for insertion of brass plugs - cut off brass screws loc-tited or epoxied in place, which would allow for solder attachment of details like the air reservoirs, ground lights, cabling, piping, and air dryers.

Miscellaneous parts

  • in yellow, the mounting tabs for the plastic factory air reservoirs, if replacing
  • in blue, the jacking pads, if replacing
  • in red, some sort of filter or dryer, if replacing

Miscellaneous parts

  • in green, these tabs are where the hood used to screw on, they can be removed or left for bearing and attachment of an electronics shelf
  • in yellow, the optical sensor mount for the flywheel tach, only needed for tmcc ops :roll_eyes:
  • in red, board mounting lugs, can be removed but useful for electronics shelf mounting at the front end.

That’s it for now. I’ll address the fuel tank modifications tomorrow, then it’s on to the shell.


Jim! Awesome!

Very elegant approach that is repeatable and achievable! Keep them coming!


Thanks Rick. It’s actually fairly simple to cut and splice the frame. It’s more an issue of being methodical in all the steps, four times over.

There are some clarifications or corrections to make before I move on to the fuel tank. The first concerns the large chunks of diecast material at the rear of the frame. In reality, all of this material in the yellow box can be removed, with the only consequence being a further reduction in weight.

I had chosen to not remove any of this, to retain weight, and I didn’t see any other explicit reason to remove it at the time. In looking at the situation now, and seeing how much of the hood volume is consumed by the speaker and decoder, I can see that it might be a reasonable trade-off of weight for extra real estate. There is plenty of clearance with the rear truck. The additional space under the sub-base is about 1"L x 2.5"W x .625"H. It’s quite a bit of useful room.

The second area that I missed earlier is at the front of the frame. There is a rib of material that will need to be removed to make room for the splice bar. It should also be noted that there are any number of ejector pin marks around the frame that should be removed to allow the splice bar to seat flush against the flange. A couple of them can be seen in this photo. The marks in the middle of the frame don’t need to be removed as they don’t affect anything. But marks along the flanges should be addressed.

Lastly, the cab sub-base is held to the frame by four screws. There are four threaded inserts on the underside of the sub-base, like the ones in the yellow boxes below. There should be two more in the green boxes, but I unfortunately just cut them off recently. I will glue them back in now that I’ve had an epiphany/face palm/duh! moment.

If the frame is shortened as I’ve done it, the pair in green are pushed forward and interfere with the outermost areas of the rear bolster (which is why I cut them off). Looking at the situation now, it’s clear that I can remove some of the bolster leg material and narrow it enough to clear the threaded insert. Drill and tap for the screws. I need to check but I believe they are #2-56 screws. But they may be metric so I need to find that out.

That’s it for corrections, until I figure out something else I’ve missed or haven’t thought of yet :thinking:.


I would continue, but I’m being obstructed!

On to the fuel tank. I’ll start off by saying that there’s more than one way to handle the fuel tank, and I’m not even sure which way is best or what my final choice will be. So I’ll outline what I’ve done so far, and talk about alternatives that I’m considering.

The fuel tank on the mp15dc is comprised of two parts: the diecast weigth that is integral to the frame, and the diecast cosmetic “skin” that is attached over the weight. The internal weight is shown with the frame photos above. The outer skin is shown below.

The existing tank isn’t the right length, or quite the right width, or quite the right profile. The profile is close enough though that I decided to go ahead with the stock tank. The length and width were corrected.

The internal weight has been removed, so the outer piece is all that’s left at this point. As it stands, the mp15dc fuel tank is about a scale 92" long and a scale 124" wide. It should be 66" long for the sw1500, and 120" wide for all later EMD locos. At a minimum, the tank needs to be shortened. I’m picky enough that I also wanted to narrow the tank. The extra 4" isn’t much at scale. I was more concerned with the visual disparity that the eye picks up, namely the tank extending past the face of the frame and air tanks above.

To shorten the tank, I removed enough from each end, including the end sheets, so that 66" remained. I didn’t want to have a joint in the middle of the tank, and have to deal with finishing it convincingly, or risk misalignment. I also wanted to represent the way that the end sheet is about 1/2" larger than the tank profile. This lip is readily apparent on the prototype. I wasn’t confident about laminating a new end sheet to the diecast tank. The plan instead was to fabricate new end sheets from styrene with a lip that would fit inside the tank, with appropriate details added.

This method has a distinct pro and con. There wouldn’t be a seam in the middle of the tank, but removing the ends allowed the tank casting to stress relieve slightly. The tank is now slightly asymmetrical, which isn’t readily apparent to the naked eye, out of context. The difference is somewhat discernible where the fuel tank runs parallel to the air reservoir, the gap isn’t even (technically the reservoir is hung with a slope, so the gap shouldn’t be even, but that is a whole different thing to worry about), but this can be addressed with shims. This does somewhat complicate the process of laying out the new end sheets in styrene. They’re still a work in progress.

To narrow the tank, a cut was taken down the middle lengthwise, and enough material removed to bring the overall width to 120". It’s a very small amount of material removed, and the operation could very easily be deemed unnecessary or unworthy. At any rate, a brass splice plate was fitted, everything was drilled and tapped. Exposed screw ends were dressed flush with the bottom of the tank. As with the frame splices, the screws could be backed out one at a time and set with Loc-tite or epoxy. There are an additional two holes in the center, drilled and tapped for attachment to the mounting bracket.

Mounting the revised tank, and air reservoirs, is a bit tricky. The original tank screwed into the frame clear of either end of the motor “compartment”. There aren’t a lot of places to attach the modified tank without interfering with the motor, and especially without end sheets with their proximity to the frame. Ultimately, a “U” shaped bracket was used. It is attached with screws that pass through the frame rails horizontally, from inside to outside. The same screws are drilled and tapped into the back side of the air reservoirs. The bracket is clamped between the reservoir and the bracket. New holes were drilled in the tank for the fuel filler pipes, and a 3/16" hole was drilled in the right side of the tank for the inset fuel gauge. The fuel tank is then screwed to the bracket.

It all works well enough, and there aren’t a lot of other options anyway, but I am going to revisit the whole arrangement. The obvious issue is that the reservoirs, bracket, and fuel tank cannot be removed without taking the motor out. The motor is easy to remove, but of course first the electronics shelf needs to be removed. Another issue is that the volume of the fuel tank is taken up by the bracket. If the tank could be hung from its ends, the bracket could be removed and the space utilized for electronics.

I’d like to find a solution where the tank is supported from the end sheets, attached directly to the frame somehow. I will also consider 3d printing or etching the end sheets, mounting brackets, and other details. The last option is to 3d print the entire tank. This is the best solution from the perspective of accuracy and simplicity. I could very effectively represent the way the prototype tank is formed on a brake, rather than rolled, with facets every two to three inches.

But 3d printing larger pieces can be expensive, the finish quality needs to be addressed, and the plastic doesn’t offer anything in the way of weight. I wouldn’t be surprised if I settle on some sort of mixed media solution with etching and 3d printing, but I haven’t come to any conclusions yet.

One last thing to note is that the motor hangs down below the bottom of the frame. It does on the mp15dc as well, but the fuel tank is long enough that the motor is contained inside of it. Whichever solution is picked for the end sheets, it will have to have a notch along the top edge to accommodate the motor. Unfortunately, this is right in the area of the fuel tank mounting bracket, so that detail will be somewhat compromised. But there is no option other than fabricating a custom drive from scratch.

That should do it for modifications from the walkway down. Next I’ll address the hood.


Hey Jim,
How is the rail head to deck (walkway) height? The prototype 1500/1000s are around 64.5 inches. I’m not sure about the MPs, but I do know the MPs hood was about a foot higher. Does switching trucks effect the model’s deck height?


The prototype top of frame height for the 1000/1500 is indeed specified as 64.5" above the rail head. There is a mp15ac clearance diagram on the UtahRails site that indicates the same dimension. So I would say they are equivalent.

I don’t think there is any difference in deck height based on which truck is used. If there is, it isn’t appreciable. When I compared the trucks, they spec out the same, except the Blomberg truck is 1’-0" longer, it has 4 idler gears instead of two, and it has smaller axle gears. AAR A on the left, Blomberg on the right. There aren’t any gears in the 8’-0" truck, which is why the holes with bushings are empty at the moment. 40" P48 wheels in both trucks. In my opinion, the height for the top of the gearbox and kingpin are the same for both models.

As far as hood height is concerned, I don’t think there is any difference between the sw’s and the mp’s, at least not based on information I have. Looking at both drawings and photos says to me that they are the same. I sure hope they are, otherwise I’m really gonna make a mess of this project :grimacing::flushed::sob:.


Wow. That is a lot of amazing work. I can’t wait to see the (un)finished model. What road will it be?


Hi Jay,
Thanks! I can’t wait to see it either :roll_eyes:! It’s been in the works for a while now, and in my head for even longer. I’m not the fastest modeler because I’m pretty new to the execution phase of modeling, as opposed to the scheming and planning phases. So I have to learn about almost everything before I try it for the first time.

The funny thing is most of the frame work was done by someone else. I lucked into meeting a fellow local modeler that is a retired machinist and a p48 dabbler. He offered to do the machine work for me, which was greatly appreciated.

The unit will be based on an SP prototype, either in traditional SP paint or as repainted into UP livery. I haven’t fully decided on an era yet, but I’m sort of leaning toward the late 90’s or early 00’s. I like the quirky SP light package detailing, even more so once some of the lights have been removed. And I like the high number boards. SP had a lot of sw1500’s, but I will be restricted to a fairly tight range of numbers to choose from unless I decided to modify the pilot details and the hood door directly forward of the hand brake. The pilot end sheet profile where the bottom step is attached varies by phase, and the door in front of the hand brake went from half height to full height somewhere along the line as well. I will have to double check, but I think I will end up in the high 2500’s/low 2600’s if SP, and the corresponding renumber if in UP paint.

I still have lots and lots of work to do, but I feel like I’m right at a threshold where suddenly quite a bit will happen quickly. Fingers are crossed, anyway.



Glad to see Jay posting! He’ll never admit it, but he is a data machine!

So I did some searching in my basement and found some MP/SW drawings in the 1980 Car and Locomotive Cyclopedia, the MR Diesel Locomotive Cyclopedia, and a photo copy of an old Diesel Era article. It looks like the SW1000, SW1500, and non filter version (phase-I ?) MP15DC all have the same hood height around 12’6" and cab height of 15’.

EMD made some alternate models like the SW1001 and MP15AC that addressed clearance issues and have varying dimensions and I believe this is the source of some confusion. It appears that a lot of drawings and data sheets on the web might be based on the rather more common EMD published literature advertising the factory alterations to the standard designs vs the actual less commonly found base model data/drawings. There are a couple web base fan sights that have data list, or drawings. I have found quite a few errors I some of these, and others efforts seem fairly well researched and executed.

FWIW, the Trainiax drawings site seems to have some pretty accurate png file drawings in the 1:35 and 1:16 scale sizes. Even so, I’m leery (until confirmed through photo comparison) due to the what I’ll call the “silhouette effect”. Basically, from a casual glance, most people see the whole locomotive to be correct if the silhouette is correct. But a closer look or a better understanding of the mechanics involved, will show an amazing variety of differences. Model railroaders and manufacturers use the “phase” naming convention to help communicate these differences. The silhouette effect probably originated during the steam to diesel transition era and it’s moto is: “all diesels look alike”!


I have EMD drawings which spec the top of loco frame at 64.5", and top of the hood frame (EMD hoods are essentially cages sheathed in doors or sheet metal) at 7’-2" above the top of loco frame. So that would put the top of hood at 150.5" above the rail. This jibes with the information found in the mp15ac clearance diagram on the UtahRails site.

Regarding fan sites, Trainiax is surprisingly accurate and comprehensive considering that he is often deriving data from photos as opposed to drawings. I have found some things there that don’t quite add up when compared to other references, but the big picture info is pretty solid. And there is a lot of good phase information there too. Conversely, I avoid www.thedieselshop.us as a reference. There are so many mistakes and inaccuracies that I’ve noted that I do not trust any of the data to be correct.


In response to a comment from Bob elsewhere, I decided to double check the heights of the two different gearboxes. The photo I posted above actually doesn’t do a very good job of showing how similar the trucks are, probably due to the camera angle.

Compared side by side, p48 wheels all around, measured to the top of the small ridge adjacent to the kingpin, there is an average difference between the AAR type A and the Blomberg of +/-.002". Results vary among the samples as there are often slight differences between the halves of both types of gearboxes. I’m willing to consider the .002" disparity as a wash for this project.

Perhaps more importantly, I checked the deck height with the trucks installed. Accounting for p48 tread depth, I measure the deck height with AAR type A trucks at +/-65.75", which is a little more than an inch over the specified dimension. I’m willing to do lots of dysfunctional things in the name of accuracy, like detailing things that will never be seen, but for the sake of simplicity I’m willing to accept the discrepancy in finished deck height on this model.


Glad to hear that the gearbox height was just an optical delusion. :grin:

Great build, by the way! And there is no way that any rivet counter’s so-called calibrated eyeball can see a 0.021" discrepancy in the overall deck height.


Every reference I can find lists the MP15DC and SW1500/1000 hood as the same 12’ 8" height above the top of rail. The MP15AC is listed at 13’ 1 1/2".
But there are photos that sure make the MP15DC LOOK taller than the SW1500. Thought it might be different cab height, but they are all the same as well. Top of the cab is 15’ for all of them.


Hmmm, I wonder if the extra 5.5" is accounting for the radiator on top of the mp15ac? The only reference I have at the moment for the mp15ac is from the UtahRails site, and it suggests that the height behind the radiator is the same as the other switchers. But I realized that I can’t compare pictures of the front ends and proportions involving the radiator fans since the mp15ac doesn’t have the usual front fan :thinking:!

The good news is it sounds like I can go ahead with the mp15dc hood for my model. Phew!


Here’s the quick down and dirty on the hood modifications for the sw1500. There are lots of other improvements that can be made to the loco superstructure if so inclined, but I’ll talk about those things later. For now I’ll detail the bare minimum changes to go from the mp15dc to the sw1500. Changes for the sw1000 will be similar but different.

In a nutshell, all the changes happen between the two red lines. Everything fore or aft of the red lines can remain intact. So lets focus on the louvered doors and the hood transition to the cab front.

We need to go from this:

To this:

And from this:

To this:

Don’t mind the filter box on the last couple of photos. 99% of the work is about the doors, latches, louvers, and grabs.

The first step is to remove the low hood section directly in front of the cab. Use care when removing this section as several parts of it can be reused on the sw1500. On the right side of the loco, the section with the grab iron holes can can be salvaged and re-positioned between louvered doors. It has the advantage of already have the nut and bolt detail castings. On the left side of the loco, I want to save the stepped area directly in front of the cab in hopes that I can tack it onto the revised hood to fit with the cab steps. The door in this section isn’t of much use, but the latches can potentially be reused.

Speaking of the latches, it’s worth noting that PSC makes knucklebusters in both brass and plastic. If scratchbuilding car sides or installing the odd latch, these parts could be the way to go. I’m going to try to salvage and re-position the Atlas latches because I am concerned that the PSC version doesn’t match the existing version. I’d like all latches to come from a single source.

Lastly, the peaked roof of this hood section should match the profile of the rest of the hood. If modeling an sw1000 with the shorter radiator, or stretching an mp15dc into an mp15ac, this leftover hood section could be quite useful.

I’m not going to get too detailed on how to modify the rest of the hood. It’s pretty straightforward, based on the photos. Louvers, latches, hinges, and grabs all need to move around or disappear in one way or another. For now, I’m going to try to cut and paste these parts if possible, while keeping the general integrity of the hood intact. So I will cut out a latch from this area and insert it on a different door. Another option would be to try and cut out doors in whole and switch them around or flip them. Yet another option would be to shave all the surface detail off and use styrene quarter round to form new louvers. Everything has crossed my mind. I’m in the process of modifying this area of the hood right now. I will post results when I have some. Time to buy some filler!

The sloped section directly from high to low hood needs to be removed. While it generally looks ok, it is in fact about 5" too long. I eventually created a 3d model of this part for printing, but initially I was going to build the section up out of styrene sheet and quarter round. The 3d model is much simpler when dealing with the miters and compound curves involved in this part. This isn’t the best photo, without primer. But it’s the best I have at the moment. The hood transition includes the access panels on top, bolt detail, and a lip at the rear that fits inside the opening on the face of the cab.

Not a lot of words or photos for this portion of the write-up. That’s partly because much of the work is only theoretical at this point. Once I get things mocked up, I will post additional images.


EMD switcher dimensions…



The Hosam site has been a go-to for me for years now, especially regarding truck centers and wheelbases. It’s partly responsible for the idea to transplant the sw9 drive into the sw1500.

Corroborating what the Hosam site suggests, I found two different RMC articles concerning the mp15ac, both with drawings. While neither has the top of hood dimensioned, they both indicate graphically that the height from top of doors to top of hood is greater than on corresponding sw1500 drawings. So the mp15ac is apparently indeed a slightly different beast. Useful to know for when I tackle that project down the road.

I did notice while revisiting the Hosam site that the walkway height for the sw9 is something like 9.5" lower than for the sw1500. So it would make sense that the bolster on the Atlas sw9 is taller than on the mp15dc, lowering the walkway height in relation to the trucks and top of rail.


Well, it turns out that modifications to the shell to rearrange the louvered doors will not be as simple as I had originally thought. So that photo I posted above with the arrows saying everything to the left and right of the louvered doors can remain as is, should be ignored.

It turns out that there isn’t quite enough room to fit the louvered doors and blank panel. This state of affairs confuses me, because I could swear that I checked this condition back during the discovery phase of planning. Based on the information I had at the time, I expected to gain the width of the space between the two pairs of doors in question, and concluded that this was sufficient. In practice, I clearly miscalculated. Based on measurements taken with the hood, cab, and sub-base fixed in place on the shortened frame, I will need an additional +/-.150" to fit the final four panels in place.

Where did the extra space get to? In doing some forensics, I found that there is some stretch in the width of the doors. Each of the six doors underneath the exhaust manifold hatch are a scale .375" to .500" too wide, as are the louvered doors. It also appears that the exhaust manifold hatch itself is 6" too long. Lastly, the flat area of the sub-base behind the cab, where the sand fillers and battery box lids are located, seems to be too deep, which pushes the cab front forward by several scale inches.

I see two solutions at this point, both of which are manageable but will involve a great many more steps to implement. I’m not particularly excited about either choice, as I had hoped to keep things simple (although keeping things simple went out the window a long time ago anyway).

Solution #1 is to take a trace of material out of each door, and then splice the pieces back together. This would involve a minimum of 6 doors per side, and possibly as many as 9 per side. I can recapture the length lost to stretch by removing a saw kerf’s width from each door. This will undoubtedly be a tedious exercise, with plenty of opportunities for error. This approach has the advantage of retaining the latch and louver detailing from the original shell. It has the disadvantage of introducing a lot of joints and potential weak areas in the shell.

Solution #2 would be to bin the hood sides and scratchbuild replacements with sheet styrene, PSC latch detail parts, and Archer resin louver decals. This method has some appeal as I could do the whole side with one backing piece, reducing joints and improving strength. It will be a wash with solution #1 in terms of cutting, as I would need to open holes for all the latches. One distinct advantage with building new sides would be in the use of a thinner backing sheet versus the .085" thick Atlas shell. Even going to .060" styrene would gain .050 which would greatly facilitate installation of a Tang Band 1931s module.

I think that the first tack I will take is to see what I can manage with the existing shell. I already have it in hand, and I really want to avoid the condition where the finished model is comprised of so many new parts that I should have just started from scratch. I already kind of know that anyway, but I’m clinging to my original premise for as long as I can on this build. I’m also already feeling quite a bit more confident in my skills even with the minimum amount of experience I’ve gained so far. With care, patience, and the right tools, I think I can pull this off. Besides, the worst-case scenario is that things don’t work out properly and I resort to building the sides as per #2.

I do think that this experience indicates that scratchbuilding hood sides would likely make more sense for a model like the sw1000, where there is an even greater variation from the Atlas shell door arrangement.

P.S… It looks like Archer offers an O scale knucklebuster decal. This is very tempting as it does away with the need to cut a hole for each latch. The disadvantage is that there isn’t any depth representing the cavity behind the handle. The latter is a bug-a-boo for me, so I probably won’t pursue this option.


It’s been several weeks since the last post on this project. I handed off the frame for a quick bout of additional machine work, to remove the extra material below the cab floor and add a pair of holes for attachment of the sub-base.

It would have been ideal to take all surfaces down flush with the top of frame, but there was concern that any inconsistency or deformation in the frame casting might lead to some areas being flush while others cut below the frame surface. So it was left a bit proud. It won’t be an issue when fitting some electronics componentry under the cab.

While the frame was out for work, I did some further study on the Atlas hood, cab, and sub-base to double check my conclusions in the last post. Upon further review, it now appears that things are not as problematic as my last post suggested. According to my new math, which is yet to be proven out, I should be able to gain any needed length by removing the space between the two pairs of louvered doors, combined with reducing the depth of the horizontal face of the sub-base behind the cab. Hopefully this will suffice, because cutting and splicing the louvered doors, while feasible, would involve numerous opportunities for failure.

One other issue I noticed was that when the sub-base is screwed to the frame, the back face isn’t quite flush with the face of the step well. It’s off by a very small amount, but with tolerances being so tight, every little bit helps. Plus it doesn’t look quite right. The solution was to take a few swipes per hole with a small round file to elongate the rear pair of holes.

I think that the conclusion reached previously regarding scratchbuilding hood sides for the sw1000 may still hold true.

All for now. I’m going to get back onto the shell modifications in the next day or two, and will report the results once I have them.