Model Railroading > 28 Step Speed Matching (Part 2) – Sneak-Up On It

Needless to say, the accuracy of the speedometer is critical.  I had used the AccuTrack Speedometer for 3-Point Speed Matching of the pair of engines I ran in January.  After tweaking the CVs, the engine needs to make a lap on the test loop, reach the speed of the selected throttle setting, and run thru the speedometer. The process of Speed Matching is making a series of adjustments to the CVs (that affect the Speed Curve), and then measuring the resulting speed.  Each adjustment gets closer and closer to a match, for that throttle setting.  Usually, the speed is adjusted upwards until the speed over-shoots the target MPH.  Then the speed is incrementally decreased slightly, until the desired speed is achieved.  This process can be VERY time consuming.  Much of the time is spend waiting for the engine to go around the test loop, and pass-thru the AccuTrack Speedometer.
 
Let’s Roll
When TO Member djh4d (Dave) posted this about the AO Rail Speedometer:
     https://www.trainorders.com/discussion/read.php?3,6006022,6006141#msg-6006141
I immediately saw the value of being able to have a locomotive test stand with a roller-based speedometer.  The Speed Steps could be altered, then the throttle can be set to that step value to find the corresponding MPH, in seconds.  (The AccuTrack requires the engine to run-thru the AccuTrack tunnel, then run the entire loop for another pass.  Or, the engine can be reversed back-thru to a new staring point, for the next MPH test cycle.)  Knowing how time consuming it was to deal with just 3 CVs for the 3-Point Method of Speed Matching, having to deal with the 28 CVs of the 28-Step Method was intimidating.  I had 8 engines to Speed Match.  With the AccuTrack, it would take until Christmas to get it done!  :-O
 
With the AO Rail Speedometer, the cycle of adjusting the CV and determining the MPH is significantly reduced.  The posting from Dave could not have come at a better time.  Thanks Dave!
 
The Golden Engine
This is the designation for the engine with the baseline speed all other engines need to match.  The choice of this engine is important.  It should be an engine with reliable mechanical and speed characteristics.  (No jerky movements, or speed lags.)  This means that an old favorite engine (with too much run time) may not be a good candidate.  It also should be consistent in its Forward and Reverse speed curves.  Unfortunately, I discovered the SP 8556 (that I chose as the Golden Engine) began to exhibit some odd behavior after the 3rd engine was Speed Matched against it.  The forward and reverse directions had different speed curves.  The work-around was to make sure it was the lead engine and only facing forward.  Each test engine was spun 180-degrees to Speed Match its reverse direction, since merely reversing the consist would impart the bogus Speed Curve of SP 8556 to the test engine.
 
Squeaky Clean
It goes without saying that during this process the wheels need to be cleaned regularly, as does the test track.  As an engine is being broken-in with laps around the test track, have it pull a track cleaning car.  This not only keeps the track clean, but reduces the amount of cleaning required for the wheels.  AO Rail has an accessory for their test stand that is a Wheel Cleaner. 
       youtube.com/watch?v=hZQOnQ7Jmz4&t=1s
It basically clamps a section of a paper towel across the rails.  Saturate the paper towel with IPA (Isopropyl Alcohol).  One truck is placed on the usual roller assemblies, with the other truck placed on the paper towel.  When power is applied, the wheels spin on the paper towel, rubbing-off the gunk.  Slide the paper towel jig back and forth (along the track) to expose new/clean paper towel surface.  If the skid marks are pretty dark, unclamp the paper towel and move the soiled paper towel slightly, spray and repeat.  This is much more convenient than my old method of draping a saturated paper towel over a section of track, and trying to hold-back the engine while I slid the paper towel towards me (while trying not to rip the paper towel or derail the engine).  It is tempting to skip this wheel cleaning chore, but that is asking for a frowny-face later when the engines are not quite in-sync when negotiating a horseshoe curve.  
 
Lock-Step
If the engines being Speed Matched will not be in helper service, getting them “close enough” works.  The string of cars only interacts with the coupler of the trailing unit.  Any push-pull between engines is resolved within the consist.  There will be more wear and tear on the engines (because they are fighting each other), but that is transparent the string of cars.  However, in the Speed Matching game for helpers, inches matter.  Consider the Golden Engine placed 1” away from the test engine, and they make a trip around the 20’ test track loop.  If they are now 2” apart, that will compound the farther they travel.  On a long string of cars (like the 36 cars in front of the mid-train helpers on the Oil Cans) there is about 3” of slack.  This allows for some forgiveness when coordinating the speed of the helpers with the head-end.  However, if the helpers are lagging 1” every 20’, it doesn’t take long for the slack to be stretched-out.  If the cars are on a straight section of track, the helpers will be dragged along.  But eventually a curve is encountered… string-line.  This is exacerbated by the double-shelf couplers, which (like the prototype) will NOT uncouple during a derailment.  The cars tip-over like dominoes.  If I’m lucky, the curve is on a flat surface.  If not, the curve has a steep slope on one (or both) sides of the roadbed.  How many details are damaged as a result of the derailment?  To add insult to injury, much of the layout is in-view of the public.  Scrambling to find a footstool to be able to reach the cars feels like a walk of shame.  :-(
 
Calibration Validation                                                                                                                                                              
This engine now needs to run in Reverse at the same speeds.  In order to confirm the new roller speedometer is consistent in both directions, the locomotive was first facing to the Left, MPH values taken, then spun 180-degrees, and then facing to the Right, and MPH values taken again.
The Reverse Trim (CV95) was kept as it was from the 3-Point Method. 
 
Note that the 2nd column “CV Value” was assigned by LokProgrammer when I chose the Linear graph option.  The voltage going to the engine is divided into 255 increments.  To form a Speed Curve, each CV gets a range of these increments.  For the 2nd engine to be Speed Matched, this column will be modified (by moving the graph points up and down) until the MPH matches the 1st engine.
 
The Forward MPH was measured with the engine facing Left, and then again facing Right.  This was to confirm the AO Speedometer was giving consistent readings, regardless of the direction it was facing. 
 
These are the results for the Golden Engine, SP 8556:
 
Throttle          CV           Forward MPH          Reverse MPH        
Setting         Value         Left       Right         Left       Right              
    ----            -----        -----      ----           -----       -----            
      1                1            0          0               0            0                    Start-Up (CV2) value: 1               
      2                9           1.6       1.6            1.7        1.6                  Top-End (CV5) value:  137             
      3              18           3.4       3.5            3.5        3.5                 Forward Trim (CV66):  0
      4              28           4.9       4.6            4.3        5                     Reverse Trim (CV95):  130
      5              37           6           7               6         7                                   
      6              47           9           9               9        10                               Road Number: SP 8556                  
      7              56           11         11           11         11                                     
      8              66           13         13           13         13                                      
      9              75           15         15           15         15                                      
    10              85           17         17           17         17                                     
    11              94           19         19           19         19                                       
    12            103           20         20           20         21                                     
    13            113           22         22           22         22                                       
    14            122           24         24           24         24                                       
    15            132           26         26           26         27                                        
    16            141           27         27           29         29                                         
    17            151           29         29           30         30                                         
    18            160           32         32           32         32                                         
    19            170           33         33           34         34                                         
    20            179           35         35           36         36                                         
    21            188           37         37           38         38                                         
    22            198           39         39           40         40                                        
    23            207           41         41           41         42                                        
    24            217           42         42           43         44                                        
    25            226           44         44           45         45                                        
    26            236           46         46           47         47                                        
    27            245           48         48           49         49                                        
    28            255           50         50           50         51                                         
 
The good news is the roller speedometer is very consistent in both directions.  It was rarely off, and only by one MPH if that.  This means I do NOT have to spin the locomotive to get an accurate MPH reading (to keep the roller spinning in the same direction).
 
The bad news is the Reverse Trim (CV95) appears to be elevating the speed in the Reverse direction a tad.  It behaves until Step 14, but then starts to differ by 1 (sometimes 2) MPH above the Forward MPH. 
 



Edited 3 time(s). Last edit at 08/03/25 08:22 by tmotor.

---

CV95 was reduced from 130 to 129.  Here are the results:
 
Throttle        Forward     Reverse     
Setting            MPH          MPH 
    ----              -----          -----          
      1                 0              0               Start-Up (CV2) value: 1               
      2                1.6           1.6            Top-End (CV5) value:  127             
      3                3.4           3.3           Forward Trim (CV66):  0
      4                4.9           4.7            Reverse Trim (CV95):  129
      5                6              6              
      6                9              9                           Road Number: SP 8556                  
      7              11             11          
      8              13             12          
      9              15             14          
    10              17             16          
    11              19             18          
    12              20             19          
    13              22             22          
    14              24             23          
    15              26             25            
    16              27             27          
    17              29             29          
    18              32             30          
    19              33             32          
    20              35             34          
    21              37             36          
    22              39             38          
    23              41             39          
    24              42             41          
    25              44             43          
    26              46             45          
    27              48             47          
    28              50             48          
 
Just reducing CV95 by one notch got the low-end more even, but from Step 8 onward, the speeds were too slow by 1 or 2 MPH.  This confirms the CV95 value of 130 as being legit. I would rather have the engine pushing a bit more, instead of dragging.
SP 8556 will be the Golden Engine that is used as the baseline for Speed Matching the rest of the engines in the consist.  To give each new engine a starting-point, the decoder file will be copied from SP 8556 and written onto the new engine.   It will then be modified until the MPH for each Step matches.   
 
Dave