Slow Down on Curves

Which locos do you have?

Do you have a meter?  If not, get one from Harbor Freight.  They are less than $10.

Measure the voltage drop at the problem location in the following way.  With the train running around the loop, measure the highest and lowest voltage.  The lowest voltage should be when the train is at your problem area.  The highest voltage should be with the train near the transformer.  If the difference is more than 1.0 volt, you have a power distribution problem.  

I have posted several times my method for eliminating voltage drops around the layout.  Multiple power drops is too much work.  I have an easier way.  Make the voltage readings and we will go from there.

Have you checked the track's gauge (width between the inner rails)?  If it is off - tight - by even a quarter a millimeter it could really inhibit free running and stall a loc.  I would start with that.  Then check the electric stuff. 

But I see a lot of flange binding on O-27, and sometimes it just happens in one spot more than another.  

EDIT: frankly, if it were me and I had exhausted all over fixes, I would use leverage to "widen" that piece by about 1/2 mm and see if it fixed the problem.  All you have to loose at that point is one section of track.  Its a problem and its a new piece of track you said:don't assume it is not without is problems.

Servo and Lee - thanks for the replies. 

Servo -My conventional engines are a 4-6-2 steamer and a 4-4-2 steamer. I don't have any strictly conventional diesels, etc. Both respond the same so I've pretty much eliminated engine issues. I've got a meter and will try your suggestion tomorrow morning. I seem to recall trying that before and not seeing much, if any fluctuation but then again, I may be imagining things. Will try in the AM and report back. Thanks. 

Lee - I'll double check the gauge but I don't think that is it, especially since I've had this on both loops, in the same spot, with different track, etc but you never know and I'll double check and try to spread the rails a hair just to eliminate any binding. Thanks again, will report back in the AM. 

You can check for binding by placing a loco on the track and rotating it about a vertical axis.  Move the front to the left while you move the rear to the right.  There should be a little slop between the flanges and the rails.  If there isn't any slop, then either the rails are too narrow or the wheels are too wide.  The rails should be 1.25" measured on the inside.

could be dirty track joints despite the nearby power drop. Solder jumper wires across pins,especially center rails. You really can't find this problem with a volt meter easily,unless you rig up a load  using a wire wound resistor. 

Dale H

Check for the presence of a strong magnetic field in the problem area (use a toy compass or some other method) as there may be a large - yet unseen and undetected-- magnet  lurking below the carpet. You never know....;-)

Dale - thanks. This is certainly a reasonable thought if the track was the old track. This stuff is brand new, fresh out of the factory wrapping as of this afternoon. First run around the track saw the same slowing that was common on the previous track. 

Does it do this with the loco only? No tender or cars.

If you power up to just before the loco moves close to the curve, that should be a good load to then take voltage readings. Loco only. No tender, no cars. 

How is your platform built?  Perhaps parts of it have sagged due to humidity and/or too much of a gap between braces.  So at the point in question, the locos slow down because they are climbing out of a sag.  Is the whole platform of uniform thickness?  A height delta of 3/16" is enough to cause noticeable slowing.

Also the 4-4-2s (both MPC and modern-era can motored variety) are about the worst when it comes to changing speed in response to a changing load, especially if it's not pulling many cars.  One of the new LionChief Plus or smaller RailKing PS2 steamers would excel in this scenario.  My $.02.  -Ted

Is there an Engineer and Fireman in the cab? Perhaps they are just throttling back.

Gilly

Are your other O27 non problem curves exactly the same configuration as the problematic curves?

I'm back - Thanks to all who have offered opinions and suggestions. 

Carl/Moonman - yes, engine/tender only still has the same issue. 

Ted - that was something I also thought of but don't think it is much of an issue but I don't have much of a way to really check due to the way everything is in place. The table top is uniform thickness and no major sags - there are lots of joists under there for support. It is a 4x8 table that is WAY overbuilt! PS2/3 engines with cruise turned on are fine, no cruise is not as bad as the pure conventional engines but still noticeable.

Dennis - Yep, all are the same configuration. 

As for voltage readings suggested earlier - I ran the engine only on the inner loop.

Voltage coming out of the transformer:

Voltage at problem spot

Another thing I noticed was the voltage dropping as the loco approached. Say the voltage is set at 4.2, as the loco approached and was on this curve, it would start dropping to 4.0, sometimes 3.9 but no further. Did this each time. 

That does not change things at all. Could be loose or some contaminants on the track or pins during manufacture.  The volt meter reading is only relevant if significant current is being drawn when measured. You could set up a power resistor so it draws 5 amps or so 8 amps would be better. At 12 volts that is 60 watts at 5 amps,100 watts at 8 amps so you have to allow for the heat if doing this.  I do not rely on track pins as conductors,they are unreliable for any track system. I solder #22 jumpers across all joints. If the layout is temporary the track can still be disassembled in large section. A power feed is placed to each large section. If you do this you will have zero issues with voltage drop.

In your measurements above,the concern is not voltage drop across the transformer output,it is drop on the section where the train gets power. With a load, (resistor at the farthest point) you would measure every section. This is hard to find,even with the test mentioned above. The weight of the train can change things as it can reposition a loose pin.

Dale H

Did you check that the flanges are not binding on the curve?

It appears you do not have a problem with the wiring.

Dale, 0.2-0.3 volts drop should not cause the locos to stop.  That is a small drop.

The last large layout I put on the floor had a maximum voltage drop of about 0.5 volts running post war locos and cars.  The current was about 3 amps.  I used only one feed to the layout which was 11x19 feet and used 072 curves and switches.  I very carefully cleaned the inside of every rail and cleaned every pin.  The track was prewar and was seriously corroded.  I checked every joint with a Kelvin ohmmeter, and the resistance of all the joints was less than 0.025 ohms.  I always splay the pins on the track sections so there is low resistance.  With the pins of the track section pointing away from me, I bend the right rail where the pin is installed to the right.  I bend the center rail to the left.  I grab the rail where the pin is installed so the pliers don't collapse the rail.  I have done this for the last several years with good results.  Lionel's method of squeezing the open end of a rail slightly doesn't work very well.  

Servoguy

That may work,some good tips, but I am old fashioned and solder everything I can. Problem with relying on compression contacts is that the layout bench work can expand and contract with temperature and humidity,the pins and track moves with the bench work and the track itself can expand and corrode in the joints. I use silver grease for such contacts if I don't solder a jumper for something that needs to be removed.. A problem could develop later and hard to find and correct with the track screwed down and ballasted. Soldering jumpers removes all doubt. Before the track is laid,sections can be soldered upside down where the wires don't show.  Just the way I do it. I have a large layout conventional. With an adequate bus, 6 volts off the transformer is still 6 volts 50 feet away or anywhere on the layout.

I still remember at work,tightening hundreds of barrier strip screws and crimp wire connectors loosened from vibration or installed improperly,trying to figure out which one was the cause of an intermittent problem. Personally I solder everything feasible.Takes time but worth it IMO.

Dale H

Dale, I make layouts on the floor.  The floor is covered with carpet.  I cannot solder all the connections because it would most likely damage the carpet.  It also takes a lot of time, and I am changing the layouts frequently.  So soldered connections are not an option for me.  That is why I developed the splayed pin technique which gives enough contact pressure to make a good connection.  The 11x19 floor layout I mentioned had about 100 joints and worked well with only one connection to the transformer.  And, when I disassembled the layout, I do not have to do anything to the track when I make a new layout.  The splayed pins are still splayed.

Based on the voltage measurements SJC made, I don't think he has a problem getting power to the track.

When you say that 6 volts at the transformer is 6 volts anywhere on your layout, how are you loading the track &  wiring?  If there is no load on the track & wiring, you are not measuring anything useful.

Hello Servoguy

I can make tests with the power resistor as mentioned. However the best test is to set the transformer at crawl speed voltage (this depends on the engine) and watch the train creep around the large loop without stalling. Each of my loops has maybe 250 linear feet of track.

If the track is turned upside down a simple jumper wire soldered underneath jumping each joint is not visible and will not hurt a carpet,there are no jagged edges. My current layout I am working on is K-Line Shadow rail.  However the sections can not be dissembled. Instead of 8 , 9 inch straight sections for example you will have 1, 72 inch section,electrically and mechanically speaking. This is OK unless you need to  change the layout configuration. My layout is permanent so it is not an issue.  The unsoldered junctions are marked if I need to pull it up for some reason. In fact one outside rail in each junction has a plastic pin for 3 rail block detection. Each outside insulated rail is wired to a barrier strip and a relay can be easily added to activate block signals,crossing gates,collision avoidance etc.  Each assembled section has its own feeder,1 ground,1 hot and 1 insulated outside rail, so no unsoldered connections need to carry current. The only purpose my track pins serve is to align the track.  Most likely when I croak someone will pull it up and throw the track in the trash anyway. The rolling stock and engines can be sold for enough money perhaps for a down payment on a motorcycle or something.

The test shown measuring across the transformer taps shows nothing IMO,unless an under rated transformer itself can not put out the load. The voltage measurement would need to be taken on the curved section with the load being pulled through that section. If there was a bad track joint(s) that is where the voltage drop would be to slow the train because that is the point of power transfer connecting the engine. The bad joint(s) would then act as a series resistor sharing the load with the train and heating up.

As for your accurate meter test with the Kelvin meter. If you check across a center rail joint on an assembled loop with 20 sections for example,the current can flow both ways,,one way across the joint and also the other way around the remaining 19 joints. You also have resistance of the connections of the test leads to the rails. I would get a bit confused with that test,not that it is not valuable. I would think a soldered joint would have negligible resistance.

Dale H

Thanks for the continued replies. Sounds like my wiring isn't so bad. I did try running an MTH SW1 with the cruise turned off - it struggled through this spot as well but not as bad as the steamers. I think flange bind may be a part of it but I'm starting to this it may be just be the nature of the curve although it still makes me scratch my head why two parallel loops will do the same thing in one spot and not anywhere else where conditions are identical. 

Thanks again.

Hi, Dale,

If you look up Kelvin connection on the Internet, you will see that the lead resistance and contact resistance of the leads doesn't affect the measurement.  I can measure resistance as low as o.00002 ohms accurately.  That is why I use a Kelvin connection.  Furthermore, I can measure the joint resistance around a loop of track if I open the loop somewhere.  This allows me to quickly find a joint between sections that has a high resistance.  

SJC, locos with traction tires and/or magnetraction will slow down on curves because the wheels have to slip as they go around the curves.  The outside rail is longer than the inside rail, and the wheels are on the same axle and so must turn the same amount.  If you watch the speed of the loco, you will see it slow down on the curves and speed up on the straight.  Now, if there is a little oil or grease on the rails, the loco gets around the curves easier.  Your new track may be completely free of oil or grease, thus making it more difficult for the loco to go around the curves. To discover if you have oil or grease on the track, take a white  paper towel and rub the top of the rails.  You will see black lines on the towel if there is oil and grease on the track.  

The comment about the voltage readings not having real meaning unless there is a load is correct.  If you have alligator clips or a way to  hold them in place, keep them connected and the multi-meter reading volts as you run the train through the curve.  The voltage will dip as the loco passes, and that is normal: so to see if this curve is worse that others and electrical connectivity is the culprit, also do this same measurement of voltage before and as a train passes on a curve where you are not having problems.  If this curve is two or more times as bad, electrical connectivity is the problem, otherwise now

and at that point, if the electrical is not the problem, then I think it is the gauge, and as I said earlier, the track is tight.  I have seen new track that is out of gauge: one of two pieces in ten years is all, but a small portion of track is manufactured too narrow.  Not often, but it happens.

Check both loops by using a small level or medium-sized marble on the gauge.  I will bet that there is a sag here.  Also, perhaps the track you purchased is not bent to a uniform radius?  Are there any sharp edges at the rail joints??

Regarding locos:  Before PS2 I was never impressed with the performance of MTH's diesels on sharp radius track.  (I had an early GP-9 with the properly spaced Blomberg trucks.)  Instead try a postwar diesel, or MPC (which has both tires on the same side of the power truck.)  I believe it will outperform an MTH diesel in this scenario, unless the MTH has its cruise control engaged.  

A RailKing Docksider or USRA 0-6-0 would hardly slow down, that would be my first choice for this assignment.  A Beep would be fine too, except that they lack flywheels and you will need two MU'd together to get over switches and crossings.  I would rather put that length into the trailing consist.  My $.02.

It has already been suggested that you check the track gauge. It's possible for the track gauge to get squashed when you grip the tracks to join sections, especially if the connection is really tight. It is also possible for some locos or cars to have wheel gauge too wide, which is more of a problem on sharp curves.

I have also encountered some steam locos that had side-play or excessive wear in the drivers that caused binding or drag on curves. You may need to check individual locos and cars on the problem curve to isolate the problems.

An expedient way to check for voltage drops on rail joints is to use an electrical load (like an automotive headlamp) at the furthest point from the track power feed, then check voltage drops at the joints in between the feed and the load.

Thanks for the replies folks!

I've checked the gauge - lots of "slop" when sliding a rigid wheelbase truck back and forth on the track. I also spread it a bit just to be extra sure. 

All of my locos are PS2/3 except two conventional engines. I've run a variety of steam and diesel engines. Some slow a bit more than others but they all do on this curve which is understandable. I noticed the diesels a bit better. i guess the "flexible" wheelbase and dual motors help a good deal. I guess what really makes me wonder is the other 0-27 curves are fine and trains pretty much maintain tangent track speed on them. I also checked the level and it looks good. 

Again, it was like this before swapping the track out with the new track so it didn't just start. Flange bind doesn't help but I would think it would be like that on all of the curves. Flange bind is some serious stuff - lots of experience there in all gauges, including the real deal stuff. 

I would suggest getting a long jumper wire. Hook one end to the transformer hot. On the section where it slows,connect the wire to the center rail and see if the problem remains. If it does not you have a problem with track pins somewhere on the center rail. Do the same thing with the ground rail. Since the engines have cruise control,it is not likely the wheels are binding,the cruise would compensate for that.

Dale H

Thought I would bring this back up as I took a video and wanted input from others before I spent time testing the DCS in another attempt to bring that back from the dead. 

Take a look at this video. Note the upper loop - quicker at one end, s...l...o....w at the other. That is a small loop (upper). Just two and a half straights on each side and 0-27 curves. 

Inner loop is a basic oval of 35" straights and 0-27 curves. Two feeds per loop. Noted here the train slowed down on the end under the elevated loop and maintained close to speed on the other end. 

The track is close to level but not 100% dead on but with a level, the bubble is about 90% in the level zone. 

Just wanted to get some input of others in regards to whether or this is flange bind. I would say yes IF trains really slowed down on both curves. I thought having a video might be a bit more helpful. 

I watched it a few times,I don't see a problem. The upper loop is going uphill a bit and on a sharp grade when it slows. You have gravity plus friction at work. An O27 curve is not easy for some locos. I could use a couple relays and some diodes and boost and drop the voltage where needed in conventional. But for such a configuration it seems to work well. 

Dale H

Dale - Thanks. Glad you see nothing obvious wrong. Like I said, I'm familiar with flange bind both in toys and real stuff (amazing how flange bind alone took a former Toronto PCC trolley I was operating from 35mph to 5mph real quick!). I guess what I was a little "concerned" about was no major speed change on one end and a lot on the other and I think the carpet tends to "screw" with the level a bit and hide the proper measurements. Thanks again - really appreciate it. 

SJC,

It's hard to use a level in the context of a train layout.  Find a medium-sized or "shooter" marble that will roll freely between the outside rail and center rail.  I will BET that corner is high so the locos strain uphill, and then rush downhill being pushed along by the trailing consist.  Seen it a million times on the floor layout in my old apartment.  Speed control will mask this, as will good gearing.  

Those 4-4-2s will notice ANY grade even those less than 1%.  They're also very sensitive to voltage drops.  And your MTH Rio Grande Pacific with rubber tires on opposite sides of its rear axle can't "skid" freely, so it will tend to strain even though it might have been advertised as O27.  The small motor combined with fairly tall drivers just doesn't have the torque or RPM required to maintain a constant speed in this scenario.  PS2 speed control covered up a multitude of sins, but a lower gear ratio is what's really needed.  Try a K-Line Pacific (the one from the Marx molds) to see what I mean.  Just don't stop suddenly, because that loco's achilles heel is that it lacks a flywheel!

turn the trains and run in opposite direction, does it still have the problem? just a stupid thought

 No its not dumb. If its a loco issue, a reverse run should go across the track area exerting forces to opposite sides of the loco somewhat, and changing the position on the wheels, that the rails touch the wheels. 

Could the center rail be too high? Factory error, or just bent up, this can "beach" a loco on high spots.

The turning ability on wheels, is not the only thing that would make a train able to run well on 0-27.

Some rollers for full O, are longer than others, forcing it to be traveling the center rail, on the extreme edge of the roller on 0-27 curves.

It can drag on the roller shaft holding tabs, or even fall off the rail 100%, but not lose power. While I haven't seen it on steam, or anything but PW Lionel, I have switchers that I must limit the roller drop height on, so they can roll back onto the track as the roller straightens, or shifts.

 If you have moved different sections of track to that position, and the problem stayed there at the old position.... 

It could be the levelness, but I can't see it. And that slows more than anything I've ever owned without a track short.

Are there any nail/screw tips poking up out of the wood?

 Are the flanges "sharp" or rounded? Compare to others. Sharp edges usually climb, but that is a loco & good sized so a sharp flange may bind. Still worth a check.

  Despite tests, it may still be a small short, caused by only by loco weight. Heat doesn't always build, esp. at 4v. and small amp draws.

 Like Dale mentions, the power will flow mostly one direction, & the easiest path is where it flows, not the shortest, or longest route. I suspect, if it is a power issue, it is between the girder bridge, and the slow curve, not in the curve. Your exact power drop positions(unknown 2 me) might reveal more on this

.

  I've had the spacing of feeds, sort of cancel each other out, when running certain trains. I learned as a boy to move the lock-on as little as one inch, not even taking it off that section of track. It wasn't a lock-on to track issue. Put new track there, and it reacts the same.  E.g. Today, to run one Hudson on my shelf layout, without a slowdown like you have, I must disconnect one certain center rail feed. If I don't reconnect it later, all other trains will slow in that section, including another same year Hudson. That's more theory than I wish to chase, the "slow" loco runs great. I just disconnect reconnect as needed. But I suspected waves cancelling each other at "balance points" where the easiest direction of the path of power to loco may change.

  A temp. zip tie to raise one roller off the track may reveal this better, by losing more power, or "cleaning up" the paths shift. But the test may do nothing also(?)...Dale? Servo?

    If its not any of that, my next guess is a black hole, forming below the spot, causing an increase in effective gravity 

Not much time today but had some...

Pretty sure that corner/region is high both on the inner tabletop loop and the elevated loop. 

Running in reverse, something I really don't do, shows the opposite corner shifting the "slow down" on the inner tabletop loop. Something I also did - the MTH C&O passenger train seen in the video is still on the outer loop. I turned the speed control off and ran it the same direction as the inner loop (counter clockwise) and same thing...fast then slow. Ran forward and the speed/slow changes sides. 

Now the head scratcher - how to fix permanently. The table frame is reused from a previous layout (and I think THAT layout reused it from the layout before) with the table top changed a few times over the years. The Mianne stuff is beautiful but the cost for that just isn't in the budget. Unfortunately, I've tossed all the leftover small tiles from the master bath remodel that would have been perfect "levelers". I'm tempted to replace the tabletop but not sure I want to risk that when that may not be the culprit. 

Really appreciate everyone taking the time. Sometimes the most obvious thing isn't that obvious. 

 How many lock ons per loop  ?Can you slip an extra lock on for testing? Heavier wire?

Gregg, 

You might have missed it somewhere in the thread but I've been there, done that. Nothing. Appears the table/table top is not level and that curve is the high point. 

Okay, this may be a really dumb question, but here it is anyway.  Are these locos actually rated for O-27 track?  The one pictured on the upper level sure looks big for O-27.  

Yes they both are and I regularly run 0-31+ engines on 0-27 with no problem but they've got speed control so the flange bind isn't as obvious. 

SJC, so as not to confuse our readers-- it's not Flange Bind.  In the case of your MTH Pacific on the lower loop, it's an issue of rubber tires on both sides of the axle preventing the wheels from skidding (which is necessary since the inside and outside rails are different lengths.)

Rubber tires are a bad idea to begin with.  But when Lionel began using them in the 1960s, right up until their Korean-made products in the early 1990s, they NEVER put tires on both sides of the same axle.  That design might work with locos meant for wide-radius track, when the difference in inside & outside rails isn't pronounced.  But on sharp curves one wheel must be allowed to skid.

If you can't level the platform surface there's one more approach you can try...  Separate the loops into insulated segments, and use variable resistors like Lionel's no. 95 to set an appropriate voltage for each section.  Full power for climbing, and a bare minimum "holding" voltage for the descent.  This will work better with a pullmor motored loco than with the can motored loco on your upper loop, because by comparison, the can-motored loco has very little hysteresis.  Even one volt makes a lot of difference.  

I can tell you from experience that MPC-era diesels, or some of the other locos I mentioned earlier on this thread will perform better than the ones you're running in the video.  I ran O27 for years on layouts that had issues like yours.  Only a select handful of locos will maintain consistent speed under those circumstances.