Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Unable To Speed Match--Do I Have An Impossible Situation?

3040 views
16 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    January 2005
  • From: Summit NJ
  • 308 posts
Unable To Speed Match--Do I Have An Impossible Situation?
Posted by fkrall on Sunday, January 15, 2012 11:34 AM

Speed matching for the first time; have spent the weekend w/o success.  I have a P2 Mikado, a P2 SW7 switcher, and a Model Power 0-6-0T I've converted to DCC with a TCS decoder.  All run well individually (well, the Model Power not quite as well).  I'm matching against the SW7 (slowest loco) and have tried all loco combinations. I'm doing this to be able to use my CMX track cleaning car--none of the locos can pull it individually.

I can match time-to-traverse a measured course by optimizing CVs 2, 6, and 5.  I've not yet coupled a consist, because when I run the locos consisted but uncoupled one performs significantly differently than the other through the speed steps. What starts as 4" separation ends up as a foot or more.  I've tweaked momentum, but it hasn't helped. The P2s don't have some of the CVs (e.g., torque compensation) that I've read in other threads may help solve the problem.

I'm thinking I have (2) problems that might prevent me from consisting these locos:  (1) they're dramatically different; even the two switchers differ in age and, I assume, type of motor.  (2) My layout is mainly broad curves in 2% grades--there's relatively little straight and level track.  This may be too much challenge complicated by too much loco diversity to speed match effectively.

I've not tried Decoder Pro--is that an option?  Or am I right that I'm out of luck until I acquire another loco closer in type to one of my current ones?

Thanks for your input.

Rick Krall

 

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Sunday, January 15, 2012 4:39 PM

 Your start speed can only match the one that runs the fastest at step 1. You can't make one that starts like a jackrabbit run any slower, other than some minor motor CV tweaking, although if you are using newer TCS decoders with automatic BEMF, they should all start as slow as they can. You'll have to bump CV2 up on the slower ones so they all run the same speed at step 1.

On the top end, you can only slow the faster ones down, so adjust CV5 so they all run approximately the same at full throttle.

Mid, you cna move either way. At the same throttle setting, adjust CV6 until they are all similar at the same midpoint on the throttle.

This should eb plenty close enough. It does not have to be exact at every speed step, just close. The load of a train will even things out, as well.

                      --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    April 2007
  • From: Clearlake, California. USA
  • 869 posts
Posted by Lake on Sunday, January 15, 2012 7:16 PM

I have had an easy time (well close to easy) on speed matching most of the engines I run as pairs together. Two Atlas GP38-2 I have worked on, off and on over the summer and never quite got them close enough. Finely I spent about two hours on Wednesday and manged to get where one would not be pushing or pulling the other to much. Now they are close enough to work together.

I still have 2 other SD50's that can not be made even close as each is completely different in every CV setting, but these do not really need to run as a pair.

These are all rather new production models so I figured it is the difference in the decoder works.

Ken G Price   My N-Scale Layout

Digitrax Super Empire Builder Radio System. South Valley Texas Railroad. SVTRR

N-Scale out west. 1996-1998 or so! UP, SP, Missouri Pacific, C&NW.

  • Member since
    February 2008
  • 8,878 posts
Posted by maxman on Sunday, January 15, 2012 7:35 PM

These are my personal opinions, but I think when you start fooling with things like momentum and torque compensation while trying to speed match you are wasting your time.  Or at least throwing mud on the situation.

The fact that you have little level track also complicates the issue, because the loco on the downgrade will naturally want to run faster the the one going uphill.  And the one going uphill will run slower than the following loco on the flat.  I think that ideally for speed matching you need a good length of reasonably level trackage.

Also, ideally, you need some sort of speedometer to see the actual speeds of each engine.  What you are trying to do is what some call the "gold engine" method of speed matching.  That is, you use one engine as the standard and attempt to match all the other locos to it.  I tried this, and rapidly came to the conclusion that I would very quickly drive myself batty trying to adjust CVs 2, 5, and 6 of locos trying to match them to the gold engine.

Yes, JMRI (decoderpro) will be helpful.  If you have locos with diverse operating characteristics you can adjust 28 individual CVs to get the locos close.  But you'd still need some way of checking the speed of one loco against the other.

  • Member since
    October 2004
  • From: Colorful Colorado
  • 8,639 posts
Posted by Texas Zepher on Monday, January 16, 2012 12:02 AM

fkrall
I'm matching against the SW7 (slowest loco)

I agree with rringer.  Basically one has to match to the worst slow speed performer, not the best.  Likewise on the top end one has to use the slowest locomotive as the highest possible mark, and then curve the lot between those two points.

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Monday, January 16, 2012 6:34 AM

 One of the hardest things to match is when one loco has a BEMF decoder and the other does not. This is why I use TCS decoders in all my locos, they have enough different form factors to fit different loco models, but they otherwise at least work the same and have the same CVs, The differences then become just in the motor and gearing in a given loco. In the case of different gear ratios, you probably have to use 28 step speed tables to get consistent runnign across the speed range. I haven't run into that so far, and the only ones that might have an issue would possibly be my Baldwin switchers and in my timeframe, the Baldwins still had their original electro-pneumatic control system and could not MU with EMD and Alco anyway. My other locos seem to all run fine together with a slight tweaking, often just to CV2 to get them to start and run smoothly at the slowest possible speed on step 1.

 A speedometer is useful, especially if your layout is either large or has no continuous run, but you can accomplish good speed matching just by running a continuous loop, with the locose not coupled but set a distance apart, if the close up, the trailing one is too fast or the lead one is too slow, and if they seperate, the lead one is too fast or the trailing one too slow. Close is good enough. It's particularly easy with Digitrax since the throttle runs 2 locos at the same time, and you can op mode program either one to adjust CVs 2,6, and 5. If you have to run one then control the other, it's more of a pain to switch back if it's the first one that needs the adjustment. JMRI can help here because you cna run each on a different throttle, and JMRI can op mode program either one. Maybe the MRC software can do this for MRC users, otherwise you're on your own.

                     --Randy


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    January 2005
  • From: Summit NJ
  • 308 posts
Posted by fkrall on Monday, January 16, 2012 8:31 AM

As always, great input guys.  I'll keep at it.  I'm all too happy to avoid fiddling with torque, etc., and will check out investing in a speedo and maybe even sectional track to form a circle.  Although frustrating so far, I must admit I'm learning a lot about my NCE system, programming CVs, etc.

I'll print out these posts and get back at it.

Rick

  • Member since
    December 2004
  • From: Bedford, MA, USA
  • 21,483 posts
Posted by MisterBeasley on Monday, January 16, 2012 8:36 AM

My first thought while reading this was "Why?" until I got to the part about the CMX machine.  Then it made sense.  I pull my CMX with two P2K subway motors, because none of my surface engines will fit in the tunnels.

Since you are using this bizarre consist for only this reason, I wouldn't worry much about speed matching.  Put the fastest engine in front, followed by the next and on back to the slowest engine, and finally the CMX car.  That will maintain tension on the couplers.  If you're reasonably close, this will be fine.

Another suggestion is to use lacquer thinner as the solvent in the CMX car.  It produces a lot less drag than isopropyl alcohol, so you may get away with just one good-sized locomotive.

It takes an iron man to play with a toy iron horse. 

  • Member since
    January 2005
  • From: Summit NJ
  • 308 posts
Posted by fkrall on Monday, January 16, 2012 4:29 PM

MisterBeasley

[snip]

Since you are using this bizarre consist for only this reason, I wouldn't worry much about speed matching.  Put the fastest engine in front, followed by the next and on back to the slowest engine, and finally the CMX car.  That will maintain tension on the couplers.  If you're reasonably close, this will be fine.

Another suggestion is to use lacquer thinner as the solvent in the CMX car.  It produces a lot less drag than isopropyl alcohol, so you may get away with just one good-sized locomotive.

Intriguing thought, MisterBeasley.  OTOH, I might have been lost in the forest obsessing about perfect matching.  OTOH, my engines' performance is all over the ice, and I'm concerned (too concerned?) about broken couplers, burned out motors, etc.  At the very least, I think I need to get closer to matched performance, but perhaps I don't need to strive for absolute perfection. It's a good learning experience regardless.

Rick

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Monday, January 16, 2012 4:52 PM

 Absolute perfection is not required. If each of the locos will spin its wheels if you hold it back, rather than stall, you also don;t really have to worry about the motor or decoder. Couplers, well, unless you have some plastic clone couplers, you'll never break a Kadee all-metal coupler.

 Trying to get absolute perfection is where the DC guys start to poitn and laugh and say how complicated DCC is. There's no more need for an absolute speed match in DCC than there is in DC. Close is good enough. If one circles the 4x8 in 10 seconds and the other takes 10 minutes, that's no good. But if one takes 10 minutes and the other takes 9 and 45 seconds, it's fine.

                    --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    December 2007
  • From: Gateway City
  • 1,593 posts
Posted by yankee flyer on Monday, January 16, 2012 6:35 PM

Hi Rick  Big Smile

If I had my druthers,   I would not try to speed match those two types of locos. I have an SW8 And it's speed characteristics are too different to impose on another type loco. I also have a 2-6-6-2 freight engine that will only run at half the speed of my other steamers. If I matched it with a 4-8-2 mountain I wouldn't be happy with either one.
Just my thoughts.
Have a good day.  Smile, Wink & Grin

Lee 

  • Member since
    February 2002
  • From: Mpls/St.Paul
  • 13,892 posts
Posted by wjstix on Monday, January 16, 2012 7:04 PM

Are you trying to speed match them over the entire speed range?? I suspect if you're using them together just to pull a track cleaning car. I would try to get them to run together at a slow speed, using CV2 and 6 and not worry about the top speed. (BTW that's a way to speed match a 4-8-2 and 2-6-6-2...make it so at half power they run the same, but make is to the top speed of the Mountain is much higher than the Mallet.)

Where momentum is handy is if one engine starts a little faster than the other, you can set CV3 so it delays starting a little, giving other engines a chance to get up to speed. Same with CV4 to get them to coast to a stop.

Stix
  • Member since
    February 2010
  • 191 posts
Posted by bnsf0823 on Monday, January 16, 2012 7:13 PM

i have been reading these threads and am trying to hone my speed matching skills.  I am curious to how jmri works does it just automatically progarm the two locos that are on your throttle?  What does bemf and Tcs stand for? i use mostly digitrax plug n play decoders on athearn and kato sd70,  acc44 engines.  can you explain a little more on how the jmri software works?  i am just looking for something to simplify my cv programming so i can more easily take advantage of all these cool dcc options.  Thanks,

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Monday, January 16, 2012 7:57 PM

 There is a script in JMRI that will do some automatic speed matching but you also need some additional rquipment like the Bachrus speedometer or some block detection on your layout so you cna make a measured distance speed trap. Otherwise, JMRI doesn't do it much different than using your throttle. I mentioned Digitrax because the throttle elts me have 2 lcocos selected at the same time so it's easy to adjust CV 2 for example on either one, right from the throttle. JMRI doesn't do much for me. For users of other systems that JMRI supports, you cna do much the same thing, JMRI has throttles on screen to run trains, so you cna start both locos you are trying to matchm and then use the single CV programmer in ops mode to adjust CV 2 on whichever loco need it. And then on to top speed and CV5, etc. JMRI would help greatly there.

 JMRI also lets you pick the other decoder options from dropdown loists, no need to know values to put in CVs for lights, just pick the option you want from a list. This also all gets stored in a database so you can quickly reprogram a loco.

 TCS is Train COntrol Systems, a maker of decoders. I've found them superior to the Digitrax ones. I won't give up my Digitrax system, but I also don't use any Digitrax decoders.

 BEMF is Back Electromotive Force. An permanent magent DC motor liek in our locos is also a generator. Spin the shaft and you cna measure voltage coming out. Decoders using thise BEMF can control the motor by comparing the voltage back vs the voltage being supplied (it's sort of advanced math - calculus is involved, google it if you really wanna know details) so that the motor runs the speed commanded - if the load on the loco increases, the decoder automatically supplies more power, and vice-versa. Some liken it to cruise control, but that really depends on the settings. Yes, it can often be set so that tyou can start the loco on level, and hit a 4% grade and it never slows down - not realistic. But it can also be set just strong enough to overcome the driveline friction so you cna get nice slow starts and consistent slow speed runnign, but if you hit a hill it will still slow down unless you increase the throttle. If the Digitrax decoders you are using are 16x models, they have BEMF, if you are using 12x, they do not.

                       --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    January 2005
  • From: Summit NJ
  • 308 posts
Posted by fkrall on Tuesday, January 17, 2012 10:52 AM

Randy, Lee, Stix--As I continue learning, I think you've all helped me bracket my problem.  I quickly sensed my locos were too diverse to speed match easily, especially for a beginner, and I believe that more than ever now.

To your points, Randy and Lee, what starts out as matched speeds and a 4" separation on level track evolves to 12" separation after about 25 feet, about half of which involves a 2% decline through a 28" radius curve.   This is at steady speed step 14/28.

As Stix suggested, given my purpose (CMX hauling), I think it would be most productive to concentrate on the low end, so that's what I'll try next.  I'm doing this for both convenience and learning, and I'm learning a lot.  BLI has what I think is a good technical reference manual that I believe will help me understand the capabilities and flexibility of CV programming.

Rick

  • Member since
    February 2010
  • 191 posts
Posted by bnsf0823 on Tuesday, January 17, 2012 7:23 PM

Very good explaination, sounds like my best and least expensive bet would be spend the time speed matching on my dt402 and getting them close as i can.  The jmri does sound like a nice program maybesome day i will have to travel that road.  Thanks for sharing your wisdom..

  • Member since
    January 2005
  • From: Summit NJ
  • 308 posts
Posted by fkrall on Sunday, January 22, 2012 10:38 AM

Success!  Following your advice, I was able to consist successfully.  As Randy said in a separate thread, the process teaches you patience, big time in my case, as I was learning through trial and error.  I celebrated with a Diet Coke and documented the occasion:

Rick Krall

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!