Speed Matching Locomotives


D&J RailRoad

Professor of HO
I've done quite a bit of speed matching locomotives over my time and was thinking, there has to be a better/easier way.
I'm sure technology is there to design and fabricate something that can do this.
Prolly along the lines of a locomotive test stand where the loco is mounted on rollers so the wheels can turn while on the stand.
A speed matching device would have two stands with rollers.
The first one, we'll call the base unit, holds the loco that you want all other locomotives to match to.
The second stand holds the locomotive that you want to adjust the speed of to match the loco in the base unit.
Both stands are electronically linked for comparison of the speed of rotation of one of the wheel rollers on each stand.
The second stand is linked to a computer that is running JMRI decoder pro or the ESU LokProgrammer. The speed table tab is activated so the speed of the second loco can be adjusted to the base loco.
The speed steps are set to 128 and the programming is activated. The computer starts with speed step one and compares the second loco to the base loco and adjusts the roller speed to match the base loco. When they match, the program makes an entry in the speed table in JMRI then goes on to speed step two, doing the same thing again until both speed steps 2 match. After it goes through the complete table, the locos are perfectly matched. It wouldn't matter if momentum is turned on or not because it's the stabilized speed that is being compared. Momentum could be applied later. I'm not sure how it could be applied in this same fashion.
It doesn't matter what brand or type decoder is in the base loco because only the speed of the rollers is being monitored.
Thoughts, suggestions, cat calls........?
 
Seems to me that this could be done with one set of rollers. The concept would be to profile the golden loco first by recording its performance at a number of speed steps. That loco would then be replaced on the rollers by the unit to be matched. The computer would then set its configuration to match the golden loco.

Sounds like a fine idea to me.

George
Edgewood WA
 
I use 28 speed steps, with a top speed scale of 56 MPH. I set up each locomotive, so speed step one is 1 scale MPH, speed step 2 is 2 scale MPH, I do this up to speed step 10. Speed steps 11 through 19, increase by 2 scale MPH, and speed steps 20 through 28, increase by 3 scale MPH. This way allows for consisting any locomotive to any locomotive, is it perfect? No, but it is so close as to not be noticeable.
 
That's just it, your not applying the same setting to all, you're applying different speed table settings to get the same results, i.e. at speed step 1, 1 scale MPH, at speed step 11, 12 scale MPH. Speed step 24, 44 scale MPH.
 
A friend told me he had used something like this some 20 years ago or so. He's tryin to find it
A friend told me he had used something like this some 20 years ago or so. He's tryin to find it.
Well, you've already found it
Rather than offer an answer, I'm instead going to ask a question. How do the real railroads do it, DJ? How do THEY speed match, when some units might be geared differently than others? Do they just take them "offline?" Or do they sort them out into groups which are similarly geared?

In my area of the country almost all the motive power were geared (going by memory here) 62-15. Supposedly around 70mph (same gearing perhaps, just lower trailing tonnage) with the F9 units on a Budd train (like the Rio Grande Zephyr), but maybe less--65 mph or so--where the rest of the EMD's were concerned.
 
Last edited:
Difference is the drive trains, real locomotives have bull gears between the electric motor, and the axle, allowing a motor to be worked or pushed, depending on its gearing, can't do this with the worm gear drives found in our models.
 
@D&J RailRoad you got me thinking and I've started developed an interesting matching setup. It will be a simple oval using Backman EZ track with a single infrared detector attached to an arduino uno or nano. I've build a single truck car that's about 9' square. Under the truck is attached a piece of styrene. When the car goes over the sensor I know the length of the piece under the car, about 34mm. When the car passes over the sensor the it reads a value < 100. With the speed of the arduino I've determined I can calculate speed up to about 150 scale mph. So I can then determine the time between the first and last reading and determine the speed. DCC++ I'm sure someone could figure out how to make the auto programming work. I'll post my project on a separate thread once it progresses some more. Hopefully doesn't die like so many other projects. :(
 
@D&J RailRoad you got me thinking and I've started developed an interesting matching setup. It will be a simple oval using Backman EZ track with a single infrared detector attached to an arduino uno or nano. I've build a single truck car that's about 9' square. Under the truck is attached a piece of styrene. When the car goes over the sensor I know the length of the piece under the car, about 34mm. When the car passes over the sensor the it reads a value < 100. With the speed of the arduino I've determined I can calculate speed up to about 150 scale mph. So I can then determine the time between the first and last reading and determine the speed. DCC++ I'm sure someone could figure out how to make the auto programming work. I'll post my project on a separate thread once it progresses some more. Hopefully doesn't die like so many other projects. :(
Although I don't use speed matching (my layout isn't big enough to warrant MU's), I intend to follow this when you post your project for future reference.
 
Last edited:
A friend of mine has done some experiments with success. From what I get from him is he takes the speed table of engine 1 and copies and paste it over the speed table of engine 2. Then makes the adjustments on engine 2 to engine 1.
 
I use JMRI as well for speed matching but I also now use the Accutrak II Speedometer to help me get more precise as each locomotive is different regardless of manufacturer. I am experimenting with low speed and high speeds. I am also setting limits to the maximum top speed too. Another words some of the locomotives run way above normal scale speeds mph. So you can set a limit or threshold for it to not exceed that speed etc. I mean really why would ya run a locomotive at 100 plus miles per hour? I like to keep it real. My 2 cents.

Accutrack II has made a lot of improvements from their first run of their device. It is quite an accomplishment with the technology available now. The link below has a real nice product description on what it does with an in depth explanation etc.


 
I currently use the Accutrack when speed matching my locos.
I do the top speed adjustment first and set it at 71mph.
Do the mid-range speed next and set it at 35mph.
I can either do the programming using JMRI or just use the CV 5 & 6 until the loco being adjusted hits those to speeds. If it's within 1mph it's fine. No need to run them a mile or so to see how far they're apart from each other.
Sometimes folks go way out in left field to do things. The Accutrack II gives ya readings in the .1mph. I don't think that's really necessary. As far as enabling taller cars to pass under it, ya just need to glue on a half inch base to both sides. Actually, the device can be cut in half and the panel side mounted next to the track to measure trains speed. It's a simple sensor exactly like that used for a grade crossing. It reflects off the side of the car. You can swipe your hand through the device and measure the speed at 650mph.
 



Back
Top