SD35 Derails in Reverse when Moving Clockwise

Mark Ricci Apr 9, 2022

  1. C&O_MountainMan

    C&O_MountainMan TrainBoard Member

    265
    682
    12
    If the trucks are wired, as stated above, is it possible one has been inserted in an “over-rotated” position (but in some multiple of 360 degrees, so it still has the right orientation) such that the wires are bound up?
     
    Mark Ricci likes this.
  2. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    After reading the Model Railroading mag review and spookynet, Interesting excerpt from the SpookyNet SD35 review . ...especially the last sentence.

    "Performance on this new version is every bit the equal of its predecessors. Out of the box, I didn't have any problems with the wheel-blackening causing any sort of jitteriness. They run very quietly, throttle response is smooth, pickup is flawless, slow speed creep is "one-time-at-a-time" and the top-end speed is realistic (IE, not too fast). Pulling power is strong, with mine comfortably able to haul twenty-five assorted 50' freight cars through curves on level track (and probably a whole lot more than that). No problems with any of the wheels derailing (even on 9.75" radius curves). Overall, just really superb looking and running models.

    Maybe I got a "lemon" and most of the run performs as experienced by both reviewers??
     
  3. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    Additional Info...
    The following is a copy and paste from the scanned Atlas service description to further detail.

    Service Trip 1 210722
    1 - 40 003 727 n -scale SD-35 CNJ road# 2502 .Loco had a defective motor and p/u issues(from trucks) Motor was replaced as were trucks. Loco was tested, now performs as it should

    Service Trip 2 210806
    1 - 40 003 727 n - scale SD-35 Jersey Central road # 2502 .Loco was not making through switches on lower speed steps .Starting voltage was adjusted so loco would perform better at lower speed steps Loco was tested ,now performs as it should

    Service Trip 3 210830
    1 40 003 727 n -scale SD-35 CNJ road # 2502 Loco had stutering issue at low speeds .Derailment issue's in reverse.Loco was sleaned lubed ,and checked for proper operation .Loco was tested extensivley in reveerse at different speeds through 9 3/4turns ,were it performed as it was intended .Loco now performs as it should .Loco was going thrugh switches at speed steo2 ,withno issue's

    Service Trip 4 210401
    1-40 003 727 N SCALE SD35 JERSEY CENTRAL ROAD #2502 ESU SOUND.REPLACE TWO TRUCK WIRES ON ONE TRUCK. ONE WIRE WAS ALL MOST OFF BUT REO PLACE BOTH JUST TO MAKE SURE THE PICKUP IS GOOD. TEST OK. DID NOT HERE THE SOUND THAT THE CUSTOMER STATED WHEN TESTING THE LOCO.

    Its only after the last service call that double crossover issues (forward counterclockwise) are seemingly resolved. Forward clockwise doesn't really work on my layout so the remaining dxover derailment issue is not a problem at the present?? Its just the reverse thru 9.75
     
  4. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    Nice layout!

    If you drag your finger around the curves where the derailments occur, are any rail joints rougher than others that don't derail (dragging in the same direction of travel when the loco derails)? Drag your finger on top of and on the inside of the rails. This may be a combination of a finnicky loco with marginally poor rail joints. Rail joint tops and the outside of the curved rail joints can be smoothed with a flat file or stone. The inner side of a curved rail joint may need a rat-tail file, or a small slip stone with a curved edge.

    The transition between the 9.75R curve and the double crossover (when the crossover is thrown to cross), creates a severe S-curve which can be a problem for lots of equipment, and couplers between cars (especially if the railcars are long and/or have long overhangs beyond their trucks. If your equipment works well over it, you may be fine, but you may add rolling stock/locomotives that can't handle it later. I would add a right-hand single crossover in the near side of the loop (separated from the curve on the left end, to avoid another S-curve). Ideally you could replace your double crossover with a left-hand single crossover, but you could also just decide not to use the left-crossover path through the existing double crossover, and leave it in place.
     
    tehachapifan and Mark Ricci like this.
  5. bman

    bman TrainBoard Member

    489
    187
    22
    Are the trip pins still installed in the SD35's couplers? With that sharp a curve radius the trip pin will overhand the outside corner rail and can can an imperfect joint on a sharp curve if the coupler were to hang down too low. The pin can also catch on diverging rails on turnouts as well. I would think a truck problem would derail all the time. You state a 1 out of 5 so that tells me coupler pin. Hard to diagnose unless we can actually see where and how it's derailing. But that's my guess have had it happen to me in the past. Good luck

    Yes Atlas sound locos have wires attached to the trucks now. I have watched various models perform flawlessly around our club layout. Not saying a dud can't creep through but overall I haven't seen problems from what my fellow club members run.
     
    tehachapifan and Mark Ricci like this.
  6. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    Thanks everyone for ideas. Only had chance to have F3 pull and push 3 60' passenger cars forward and reverse, nothing derailed anywhere. Have a video of SD35 that may shed some light. Important to note that the SD35 is materially better than before. A number of things to checks out. After getting regular stuff done, will take a closer look at pins and other loco aspects closeup.

    Thank you. Pic was taken several weeks ago. Being a holiday week for me, decided to test and see how long to add removable stuff and look like this...

    IMG_2588.JPG

    Ironically, there appears to be more roughness on some of the 11.25 radii where no problems exist but you gave me a great clue. Noticed on a couple 9.75 that at junction, one rail does not line up perfectly horizontally and has a little play, sufficient to line up though unsure how to address without removing track, possibly remedied by a new rail joiner(s)??.

    Interesting. Have a couple Atlas gondolas (seasonal cars for my layout-don't run them normally) that would derail on some 9.75 when pushed in reverse by the SD35 but no derailments when pushed in reverse by F3. Have had a couple of trainman level cars pins with hanging in random places (switch areas primarily) Usually straighten out or flip car and then the different side has no issue. Mark underside front with pencil because cutting off pin is last resort.

    The primary reason Kato track was selected is the Double Crossover .. hate to give it up. Its just so cool! Combines so much in a small package. This is in many respects a first layout and knew that the Dxover is a tight fit and then connecting one leg to a #4 would be troublesome for longer locos and cars. To date, excluding the 60' cars, all rolling stock 40' or less. Definitely would do some things different especially with a layout that's 100% DCC controlled and has no mechanical switches needed for OPS. Technically no front or back, and any side can be against a wall. Started in DC for short time with a "physical" front. Wanted turnouts close by. If I were to do again and had DCC control from start, I'd either move the Dxover or the service loop Spur access to the other long side. Plus would never build another mammoth mountain on a 2x4 layout or do another snow scene.

    Curious, is there a relationship between wheel flange size and its impact on movement thru turnouts?
     
  7. NtheBasement

    NtheBasement TrainBoard Member

    427
    620
    22
    One suggestion for sussing this out is to run the loco very slow and watch closely.

    In the only similar situation that I had a leading wheel actually rose up off the track a foot before the derail point. It also only happened with one loco. I did not have good close access to both sides and ended up taking a video with my phone and running it frame by frame to see what was going on. It turned out the track was not flat at a joint on a curve and it had a slight bump, causing the derail at the next turnout. After that I checked every joint by laying a ruler across and looking for daylight between ruler and top of rail. I found a couple of other places and fixed those too.
     
    Mark Ricci likes this.
  8. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    Thank you for the input. Were you able to repair w/o removing track? If so, how did you do it? Been thinking... possibly setting rails in place and soldering joiner to both ends and let the solder hold rails in alignment??
     
  9. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    How did you end up aligning and securing rail ends?
     
  10. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    You didn't really think this would be your last layout, did you?!

    We always discover things while building and running the current layout, that we want to avoid/improve/incorporate in the next layout...
     
  11. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    :) No, but a little Leary of buying new locos. Batting 0 of 4 out of the box This layout is experimental in many ways.

    A small totally automated switching layout using a Kato Dxover is going to happen after move.

    A rough concept for larger and lighter layout without framing members but attached straight hardwood siding, also acting a veneer, to minimize or zero out pink foamboard flex. Light enough to be hung on a wall like a picture... maybe 30x60??

    I appreciate input from anyone who gets an opportunity to watch the following short videos. Important to note that double crossover creep performance is greatly improved since receiving SD35 back on 4/7

    There are 3 key questions after viewing..

    1-Do you think the 9.75 derailment issue would likely occur with most 6 axle locomotives?

    2-Are there ways to fix without track removal? If so, how?

    3-Assuming you are a newbie with 3 loco's, 17 rolling cars and only 1 loco has issues on the layout. Do you risk repair that requires track removal or major work or say, for this layout, restrict to 4 axles locos? At this point, everything else runs great except for the SD35.

    A-SD35 Derail 9 75 center Flat


    B-SD35 Derail 1st 9 75 MT Side


    C-60' Passenger Cars Reverse 9.75 No Derail


    Unsure if the Dxover click is related to the reverse problem??

    D-SD35 SD35 Occasional Click thru Dxover Heard with Sound


    E-F3 No Click ever thru Dxover Heard No Sound - Smooth


    Thanks to all for the help...
     
  12. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    In the first two videos, it looks like the outer wheel's flange is either catching the exposed end (corner) of the mis-aligned next outer rail, or the outer wheel's flange riding up on the top of a raised rail-joiner, from where it is much easier to climb over the outer rail, especially if there's also an exposed rail corner to catch.

    1. I would try to file/stone that rail edge (inside edge & top of outer rail) even with the mating rail. Go slow, removing a little material at a time until the joint is smooth. This can be done with a dremel, but go very slowly with it, you don't want to remove more rail material than necessary.
    2. It is also possible that the caulk used to fasten the track to the layout was forced inside the coupler pocket when you pushed down on the track to seat it in place. The caulk could have then pushed the metal railjoiner upwards, and the wheel flange is riding up on top of the unijoiner, which makes it easer to climb over the outer rail. Compare the distance between rail top and unijoiner metal top on the derailing and non-derailing joints around the layout.
    If you can take a closeup picture of the inside of the outer rail joint where it is derailing, and an outer rail joint that is not giving you any trouble, we could compare the relative height of the metal insert in the unijoiners, compared to the rail top.
     
    Mark Ricci likes this.
  13. tehachapifan

    tehachapifan TrainBoard Member

    1,859
    859
    46
    I agree with Big Jake in that it appears the outer wheel might be climbing up on the rail joiner. What track is this and what rail joiners are you using?
     
    Mark Ricci likes this.
  14. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    I think it is Unitrack with Unijoiners; at least that's what brand the double crossover and switches are (I can see their little slide levers in the side of the roadbed).

    If indeed the caulk used to adhere the Unitrack to the layout base has pushed the metal insert of the Unijoiner upwards, this may be a good lesson learned about how (not) to glue Unitrack down to the base. Don't do it near the track joints, or at least not with a big glob of caulk.

    There are two or three hollow pedestals on the underside of Unitrack ground level tracks' roadbed (except on switches, crossovers, etc. which have a sheet metal cover on the bottom.) The hollow pedestals are normally used to secure the Unitrack in the Viaduct housings with small screws. You can use them as guides to drill from the bottom, up through the roadbed if you want to use track nails to secure the Unitrack to the base, since that would support the middle of the roadbed while you nail it in place.

    I would use a blob of caulk/adhesive on those pedestals to glue the Unitrack to the base. Some users run a small bead of adhesive along the edge of the roadbed, filling the little groove between the outside edge of the roadbed and the layout base. Others that add ballast around their Unitrack roadbed, just use the balast/glue to hold the Unitrack in place.
     
    Mark Ricci likes this.
  15. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    Yes, its all Unitrack. Thought an attached roadbed, and its higher rigidity, would be better suited for a transportable layout.

    Didn't think used that much caulk as track section was flat and just smeared some caulk along the side of the roadbed to foam. It is interesting that you mentioned nails. Never wanted to use glue or caulk. About a week after buying the track accidently discovered those hollow pedestals and used a pin vise to open to top. The old Atlas HO track nails fit perfect. Curiously, went on Kato site and looked at all the cards and never found any mention of those pedestals and/or the option to use nails or screws. At that time, only had 1/4 cork on top of 1/4 plywood and so the nails worked perfect. Nailed some of the track down-nice and snug. Then decided on 5/8 foam on top and that killed use of the nails. Could not find anything that gauge long enough.. There is no caulk on the roadbed of any turnouts. Non committal in the event track would be ripped up in future, if the whole layout could not be moved, and would make removing the dxover and turnouts easier. White ceiling paint is the binder for ballast, which is Woodland snowflake, and does a great job adhering the track without distorting underlayment like caulk.

    Below are 3 angles of one of the 9.75 junctions shown in the video.

    IMG_2621 9.75 Junction 1 Top.jpg IMG_2622 9.75 Junction 1 Side.jpg IMG_2623 9.75 Junction 1 Side.jpg ....


    Thank you for the help!
     
  16. zophia

    zophia TrainBoard Member

    45
    69
    5
    On the first picture. The outside rail second tie from rail jointer on left is out of gauge.
     
    Mark Ricci and mtntrainman like this.
  17. NtheBasement

    NtheBasement TrainBoard Member

    427
    620
    22
    I use flex instead of sectional. I was able to level the joint by ungluing one of the tracks and shimming it up with card 5 inches or so from the joint. The real problem was at a seam in the underlying plywood.
     
  18. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    Mark,

    If you look at your track joints that cause fewer/no problems, and compare them with joints that usually cause problems, and note any differences, that may help. Pay attention to gaps or differences in position of the two joined rail ends (uneven heights, gaps, as well as lateral positions) relative to each other.

    Unfortunately, in the first two pictures, the camera found higher detail areas to focus on, that were away from the rails, leaving the rails themselves out of focus enough to make them difficult to interpret. The third picture is in much better focus, but from directly overhead, we cannot tell how high the unijoiner is sitting.

    Can you try to take more pictures from the first two angles, but somehow get the camera to focus on the rails? Maybe with additional lighting from a low angle to accentuate the track details (and perhaps entice the camera to focus on them)? Maybe even use some black construction paper cut to mask off the areas in the picture beside the tracks?

    3 axle trucks are usually less forgiving of track irregularities and short radii. The longer wheelbase between the outer axles on the truck, combined with the 3rd axle in the middle, limits their ability to negotiate tight radius curves unless everything is close to perfect. Extreme measures might include removing the flanges from the middle axle of each truck, allowing it to function more like a long-wheelbase, 4-axle truck. In a tight curve, the wheel flange on the middle axle is pressing against the inside rail, while wheel flanges of the outer two axles are pressing against the outer rail. Without that middle axle flange to push on the inner rail, the pressure of the outer axles' outer wheel flanges on the inside of the outer rail is reduced, thereby reducing their likelihood to catch on less-than-perfect rail joints.

    In the meantime, you can alter your operations to avoid running the SD35 on the inner loop. Prototypically, 6-axle locomotives are often reserved for longer haul road duty, which could use the outer loop (with their larger radius curves) on your layout. You could assemble a long haul train in the yard with a 4-axle loco, and use that loco (or yet another one, if you want to add a little more interest) to pull the train out of the yard, around the inner track, and (at least partly) through the double crossover to the outer tracks, where it can drop the cars off for the SD35 road loco to take over for the road trip. Likewise, the reverse is true when the long-haul, 6-axle loco arrives on the outer track with a fresh train to be broken down by a 4 axle loco on the inner loop and in the yard.

    In fact, many older industrial sidings are not built to handle the weight of a large road loco, so they either have to be switched with a lighter, local 4-axle loco, or additional railcar(s) behind the road loco are used as a "long handle" to access and position the freight cars being switched, without the heavy road loco setting foot on the light industrial site's trackage. These additional operational details can add more interest to your normal operations.

    I'm not suggesting these ideas as a permanent "fix", but as an interesting way to enjoy your layout, with all your locos, while we continue to figure out what's wrong and how best to fix it.
     
    Mark Ricci and Hardcoaler like this.
  19. Mark Ricci

    Mark Ricci TrainBoard Member

    480
    649
    13
    Thank you for the great ideas. Yesterday, bought some Kato 9.75 and setup test track..

    IMG_2640 SD35 9.75 Test Layout.JPG

    It performed w/o derailment. Guess expected that.. Needless to say, really need to "FIRE" the track layout designers and layers.



    This seems to clear the loco and now wondering if the severe derailment when running the rarely used double crossover clockwise forward (inner to outer) is related to layout. Need to buy a double cross over for the planned automation layout so will duplicate the dual ovals and xover. Definitely don't want to go back to Atlas until the results. Since I'm moving soon, really wanted to determine if any issues are specific to the SD35. Going to wait attempted track repair until after move when layout can be worked on with the 46" legs.

    Like to always park the loco's for the evening in the spurs so the SD35 has trouble with it. However, when at hobby shop saw a box of used locos about to be priced, and there it was, an Atlas 2017 generation VO 1000. This will take the place of the SD35 for awhile. It runs beautifully on the layout. But gotta say, fitting 4 loco's with only a few cars on the layout and moving them around.. challenging, isn't DCC control Awesome. This is really a close encounter..



    Thanks for such valuable input and feedback.
     
    BigJake, Hardcoaler and MK like this.
  20. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    Mark,

    Well, now that the source of the problem is known, you can decide how best to correct it. Glad you can move forward, even if not immediately.

    Just remember, real life interferes with 1:1 railroads too.
     
    Mark Ricci likes this.

Share This Page