Since most model locomotives run at prototypically high speeds, and different models usually run at different speeds, it is worthwhile to match the speeds of your locomotives, especially if you wish to MU them in varying combinations.
My method requires nothing more than a cheap stopwatch and 3 distance markers (I use plastic bulletin board tacks with the pins cut off).
This method applies only to decoders that allow changes to CV2 (start voltage), CV6 (mid voltage) and CV5 (maximum voltage). For extra fine tuning, it is also helpful if your decoder can handle changes to CV116 (kick rate), CV117 (kick strength) and CV95 (reverse trim).
Step 1 - Adjusting the start speed via CV2
In Program on the main mode, I adjust CV2 until the locomotive just creeps at speed step 1 (using 128 speed steps). If you can't get the loco to creep with CV2 less than 30, you might adjust CV116 (kick rate) to somewhere from 2 to 5 and CV117 (kick strength) to somewhere from 4 to 25 to overcome striction in the motor. If the loco creeps faster in one direction than the other, adjust CV95 (reverse trim) by using a value of 1 to 127 to increase reverse speed and 129 to 255 to increase forward speed. In both cases, start at the low number and move up.
Step 2 - Adjusting top speed with CV5 and mid speed with CV6
Use your markers to measure out 2 distances along your track. I place my pins to set a maximum speed of 64 mph and a mid speed of 24 mph. With the mid speed less than half of the max speed, you get better control at low speeds. You can set your speeds wherever you want using the formulas given below.
The formula for HO is simple:
Distance (inches) to mark off for 5 seconds equals speed in scale miles per hour
eg. For 24 mph, measure 24 inches between markers
and for 64 mph, measure 64 inches between markers
The formula for N-scale is a little more complicated, but not very:
Distance (inches) to mark off for 5 seconds equals 0.55 X speed in scale miles per hour
eg. For 24 mph, measure 0.55 X 24 inches between markers (13 inches)
and for 64 mph, measure 0.55 X 64 inches between markers (35 inches)
So my pins are set at 0, 24 and 64 inches along the track:
The photo shows only the first two pins.
The following speed settings given are for my NCE system using 128 speed steps. My maximum speed step is 126 and my middle speed step is 63. You will adjust yours for whatever DCC system you are using.
Running your loco at top speed, measure how long it takes to cover the maximum measured distance (64" on my track). Adjust CV5 to have the loco take 5 seconds to cover it. (This is where the stopwatch comes in.)
Then run your loco at its middle speed step (mine is 63) and adjust CV6 until it takes 5 seconds to cover the middle speed distance (mine is 24").
If you've never tried to speed-match your locos before, try this method. It's fairly simple and the rewards are great. (Of course, you might have to go out and buy a cheap stopwatch
Colin ---------- There's just no end to cabooseless trains.
My PhotoBucket album: http://s31.photobucket.com/albums/c390/CN4008/
My RailImages album: http://www.trainboard.com/railimages/showgallery.php/cat/500/ppuser/4049
My web site: http://www.cmgraphics.ca
cmurray wrote:By my way of thinking, if two locos are running at the same speed at a given throttle position, then they are speed-matched perfectly for MU purposes. Besides, after setting the CV values as described, I have tested the locos in consists and they run fine.
i see what you are saying, but just because you have top, middle and bottom matched quite close, doesn't mean the reat of the steps are necessarilly close unless you use speed tables. you can't factor in mechanism friction with just 3 points. you may well have seen in DC with 2 loco's on the track one catching up at slow speeds and dropping behind when you give more throttle.?? you may well have them working ok in a consist, but with normal decoders without back EMF you can get away with loco's that are quite a bit out of matched on speed aslong as the extra resistance the 'fast' decoder gets is not enough to trip it out.
It depends on how closely they run together 'out of the box' I think. I've used the three CV's mentioned (plus start and stopping momentum CV's) to speedmatch engines so they run together very well. I generally run fairly slow (passenger trains top out around 35 MPH) so maybe it would be an issue if had to run from 0-70 MPH.
BTW I just run the two engines together and adjust the CV's on one until the speed matches the other when starting, stopping and running. Once you have two running together, you can speedmatch further engines to either of the first two, and then all three can run together. Keep doing that and eventually all your engines will be able to run together mix-and-match.
wjstix wrote: It depends on how closely they run together 'out of the box' I think. I've used the three CV's mentioned (plus start and stopping momentum CV's) to speedmatch engines so they run together very well. I generally run fairly slow (passenger trains top out around 35 MPH) so maybe it would be an issue if had to run from 0-70 MPH. BTW I just run the two engines together and adjust the CV's on one until the speed matches the other when starting, stopping and running. Once you have two running together, you can speedmatch further engines to either of the first two, and then all three can run together. Keep doing that and eventually all your engines will be able to run together mix-and-match.
similar here, but i run in all loco's first, and only match loco's to ONE 'control' loco. if you match A-B, then B-C, then C-D, any little differences could mean that A and D are distinctly different. everything gets matched to A in my system. also, i don't try and match my switchers with my SD and GP units. i match the switchers as above for the odd time i would consist them, but obviously there speed range is set slower than the mainline loco's.
This idea seems a lot easier than running the engines on a loop and guessing with one until they run the same.
I'm going to try this with my U-boats to get the DCC only one to match the DCC/sound one.
Vincent
Wants: 1. high-quality, sound equipped, SD40-2s, C636s, C30-7s, and F-units in BN. As for ones that don't cost an arm and a leg, that's out of the question....
2. An end to the limited-production and other crap that makes models harder to get and more expensive.
Hello everybody!
I have tried several ways to speedmatch and I have even done it in a complicated way. First of all it all depends on how accurate the speedmatch needs to be. I can't say I know that myself.
I want to set the maximum speed to the prototypes maximum speed, for the locomotives I have that means 70 or 75 mph.
I Started with timing the locos on a 2 m long part of my track. I became disatisfied with my ability to clock them accurate so I started time them for one complete loop (19 m) which gave me less error starting and stopping the cellphone timer.
The next thing to solve was to calculate how much I should adjust the Cv. I'm using DecoderPro to program my locos. The locos are equipped with Digitrax FX3 decoders and I use the speed table and the forward and revers trim. I try to set the maximum speed forward and reverse equal. I thought simple math should give me fewer trial laps. Maybe ...
In order to help me keep track of all figures and settings i made a simple program to run on the computer. It's not simple anymore, mostly because I enjoy programming (and I'm a better programmer than model railroader). If you are curious on my program you can find it at http://fson.se/mrlocospeed.
Now I had a tool that should make it easy to set the speed I thought. But reality isn't always what you expect. After setting Cv values that should give me the correct speed I wanted to verify it. Then I noticed that the longer the locomotive has been running the faster it ran. I tried to find out for how long time they need to run to have a constant speed (with the same settings on throttle and Cv:s). I have tried to run one locomotive for over four hours and it's not stable, the speeds varies with time. there are periods when it runs with the wanted speed and then it starts increasing speed and then decreasing it again.
Have anyone noticed the unstable speed with the same throttle setting? My locomotives are all Kato SD70, SD80, SD90 or AC4400 N-scale.
Bye the way, the locomotives run nice in consists so the goal is achieved, but this "little job" has given me more questions than answers.