Model railroading can take one to unexpected places.  I have discovered this because I’m doing things that I really didn’t expect to do or even thought of when I started.

I would like to say that I am definitely not a model railroading expert.  I am learning along the way , very much as you are doing too.  This series of posts will document my pursuit of smoother running steam locomotive models…..

Last Sunday, I was all set to reveal, in the-then next post, my latest steam locomotive in the Fillmore power pool: NYC Class L3a 4-8-2 Mohawk .  Previously tested prior to painting, decals, and weathering, I thought she looked and sounded great, and was running very well.  The last thing to do, after quite a long conversion (I’ll do a proper post on this engine later), was to make some final runs from staging through the roundhouse inbound leads, the turntable and roundhouse, the ready tracks, and back into staging.  A full dress rehearsal.  Well….

When making slow runs (actually every run is a slow run here, with speeds between 1% and 5% throttle setting), I started to notice a cyclical hesitation most noticeably in the forward direction.  It isn’t a large amount of hesitation (it happens in the same place of driver rotation), but it really started to bug the heck out of me!

After some of my typical cursing (I have many flaws), I decided to see if I could find the cause.  A big part of trouble-shooting is to carefully observe the locomotive in motion, watching not only the obvious area where the hesitation is, but also along the entire running gear and on both sides.  In short, it calls for patience and I spent much time just running and watching.  A hesitation or jerky motion falls, in my view, into two general categories: mechanical or electrical (the DCC decoder).  Since the problem was clearly cyclical, happening at the same place of rotation at every rotation, at slow speeds, gradually disappearing as throttle was increased, I decided that this was a mechanical fault.

The following is a rough log of how I approached the trouble-shooting portion of this project.  NFF = no fault found, meaning that I did a verify retest of the engine.

  • checked to see that I had re-assembled the running gear correctly.  NNF
  • checked the main rods, side rods, eccentric rods, cross-heads for play/clearance. NFF
  • found the hex screws securing the side rods to the drivers were quite loose – I had this problem before on another MTH Mohawk.  It appears that they may not have been tightened at the factory. I tightened them carefully. NFF
  • checked for latteral motion in the drivers. NFF
  • checked for anything bent or not looking as it should. NFF
  • I have heard that a new engine should be run-in and I generally do this as standard operating procedure.  However, I missed doing this at the DCC proving stage, so I put her up on stationary rollers and ran her at 35% speed forward for 30 minutes, 35% in reverse for 30 minutes (my “normal” run-in), plus 35% forward again for an additional hour. NFF – the hesitation was as before, much to my dismay.

I decided to make a more detailed examination of the running gear.  In the picture below, the boiler/cab is removed.  This is an MTH model of die cast metal construction.  One of my pet peeves in model railroading is traction tires – I don’t like them, so as part of the prep procedure for getting an engine ready is replacing the traction tire driver with an all metal one.  Surprisingly, I had to add a lot of weight over the drivers to get traction: 5-1/4 ounces!  Fortunately, the die cast construction left a lot of room above the drivers inside the boiler and lead weights are stacked like sandbags.


I have an MTH 4-6-4 Hudson where I replaced the traction tire and it operates just fine so it was a surprise to add so much weight.  Then again, there are more drivers to distribute the weight on the rails, so less pressure per wheel (just like the real thing).

On my soft-foam cradle, I performed the following checks.

  • checked more closely for clearances and play.  NNF
  • removed the motor and spun it slowly by hand.  NNF
  • removed the gearbox, dismantled and cleaned it, checked the worm and spur gears visually, assembled and lubricated, and spun slowly. NNF

I thought it was unlikely that the motor or gear box was at fault since the motor and gears move many more rotations to make one complete driver rotation.

  • with the motor removed I ran by hand the frame with drivers and valve gear connected on a piece of 18″ flex track.  I thought I detected some resistance when pushing this assembly back and forth, but not conclusively as I am sure I was applying a force greater than the motor so could not get a good feel. NNF
  • The drivers did not wobble and the spur gear on the main driver did not either.  NNF
  • removed all of the side rods and checked for burrs, ovalized holes, straightness, and measured centers for consistency.  NNF
  • with the gearbox and motor remounted, the main and eccentric rods disconnected and taped up to the running boards, clear of the drivers, I ran the engine without any side rods on my 6′ DCC programming track.  With only the main driver working, I had to add 12oz more of weight to get traction on that wheel set.  Even so, traction was marginal but I did not observe any hesitation!  A-haaa…
  • I connected the side rods between drivers 2 & 3 (3 is driven by the motor).  I still had to have the 12oz additional weight, but traction was improved.  There was no hesitation – interesting!
  • I connected the main and eccentric rods to the No 2 driver and retested.  No hesitation.  Getting somewhere now???
  • I re-assembled the side rod between No 1 and 2 drivers and tested.  Traction was much improved.  There was no apparent hesitation – okay what about the last rods?

Hoping that dismounting the side rods and the cleaning with careful re-assembling would bring a cure, but no, the hesitation is still there!

I have heard of an old model railroading trick to oval-ize the rod holes (by filing) to gain more clearance and relieve any binding (one set of side rods cleared horizontally, the other side side rods vertically), but don’t have the guts to try it.  I don’t really buy into that anyway.

But all this was very worthwhile.  I believe I am close to finding the fault.  To me, at least, it’s pointing to a quartering problem on the drivers.  With an eight-coupled locomotive, any inaccuracy in the position of the rod crank pins can build up.  If the driving wheel sets are not assembled accurately with respect to their 90-degree crank pin positioning (quartering), then a lopey or jerky motion could occur because the side rods are fighting each other during rotation.

A take-away:  It is evident that the side rods actually do perform work on a model steam locomotive.  They distribute the main driver’s motion to the other drivers where they contribute to traction and power transmission.  It is the opposite of the real thing, but none-the-less important.

A curious thing: when the assembled locomotive was running on rollers, the hesitation was not noticeable and I chalked that up to the rollers not providing much resistance/friction.

I do not have the tools to check the quartering, and if found to be off, to remove a driver and press it into an accurate location.  I contacted Northwest Short Lines (nwsl.com) and Dave Rygmyr very patiently answered my e-mail questions with very detailed replies.  I purchased the necessary tools and when they arrive, I will describe them in a future post.

’til then NYC 3004 will rest on a garden track at Fillmore Avenue Roundhouse.