Pull test project.

SAL9000,

It makes one feel good to actually see the numbers. Especially if all you had was that nagging question and no real answer.

Another use for your test assembly would be to settle one of those age old arguments. "Mine is better than yours", on a level playing field.

Interesting test.  I would expect as you test more, that the percentage increase due to magnetraction would be a function of locomotive weight.  On a very light loco it is probably more than 50% and on a very heavy loco, probably less percentagewise.  

Regardless this is quite interesting.  

Great test, and a nice test rig.   Are you going to publish the results from various locomotives?  That would be some interesting data.

As well as a BOM and build instructions so that others of us can build one and test ours with.  This could give a fun database over time to see how locomotives preform.

Thanks, all. Sure, I'll post more info when I get home tonight (though FYI it's all only postwar stuff).

A sampling of what I've tested. I've tested more but can't find the info:

As to BOM,

3' of Lionel tube track

3' of non-ferrous track (I used MTH RealTrax)

2' x 4' x 1/2" plywood base

4 x stick on rubber feet

2" x 4" x 3" end block

2 screw in hooks

Fish scale found at hardware store (mine reads 0-50 lbs)

Spare 4 wheel truck

3' leads, 18 AWG stranded, fork terminal on each end, 1x black and 1x red

Assemble as shown. I also soldered wire from RealTrax (primary) to Lionel tube track (secondary) so they are powered in parallel.

To use, connect leads between the primary track and transformer, put the scale hook through truck, lash up your locomotive, turn on the scale, and hit the throttle. Move between track types as desired.

That's great info.  I happened to have my Phantom set at the club on the Atlas track, and it has all sorts of issues on the flat profile track.  I suspect, looking at your numbers, that it would probably do better on tubular track.

The RealTrax is actually quite smooth. I bought brand new tubular track. It definitely has a rougher surface finish; perhaps it will buff out in time though...

What are you using for a "pull meter"?  I've  been thinking of getting one lately, but the only one I've ever seen is from Micromark.  Are there other options?

Roger

This:  http://www.amazon.com/Rapala-R...-Scale/dp/B000PB1RFG

Simple and cheap.

Thanks, Sal.......looks good and so are the reviews.

-Roger

I just found this one on an Internet search......

http://www.oldwillknottscales....ultrasportv2-30.html

A bit more than the Rapala, but seems very accurate and has lots of features. Universal reviews.

-Roger

Nicely done research project. I've been thinking about how to quantify the relative pulling capacities of my various locos because I like to push the  limits of train lengths and track gradients.

I reviewed the stats with great interest.  I had often suspected any of the dual motored GG1's were superior pullers but the traction tire gains were well worth the modification.

Even more impressive, and also expected, is the FM.  Whereas it's weight is less than the GG1, it pulls more, discounting the theory that weight is everything.  Well placed weight over a larger footprint does the trick.

Thanks for taking the time to experiment.

Bruce

Thanks. As to the GG-1s, note the ~50% advantage of the (original and lighter) 2340 over the 4866 remake (prior to traction tires) of the early '90s. Both had virtually zero run time. Best as I can tell the 2340 just has a bit rougher surface finish on the drive wheels (with more run time though that advantage will probably decrease).

I don't know if Lionel ever made a Pullmor-equipped GG-1 with traction tires. The recent reissues I've seen did not have traction tires. Very recent GG-1s have traction tires but the ones I've seen used DC motors. Now with traction tires the 4866 can pull ~25 of the heaviest postwar stock I have (like dummy F3 A-units, Burro crane, high drag passenger cars, etc.) and it pulls them all no problem (I'd try more but ~25 is the most I can fit on my layout). If you've ever seen ericstrains' videos he does pull test as part of his review (with an identical test procedure). Only the very biggest heavy hitters - Big Boy, diecast ES44, Centipede,etc. - exceed 3 lbs of pulling force (they max out at ~4 lbs).

As to the TrainMaster, that long wheelbase may have an advantage on weight transfer or the like. I only have the one so can't compare against others (where as I have duplicates not shown of most all the other locomotives listed). I would have to think it would have a practical advantage on actually pulling cars especially on tighter curves.

Magnetraction giving way to tires just as Pullmor giving way to can.  Seems the trend.

I believe we both own the same model of Trainmaster....the MPC version?  Again, it pulls well with Magnetraction but I could see it performing way better with tires. 

I should try some refined testing, as you've done, with the 3 FM's on hand, Lionel, Williams and K-Line.  All 3 are like in size but that's about where the similarity ends.

Bruce

Yes, that is a reproduction of one flavor or another (i.e., not a postwar original) owing to the Philips body screws and lack of horn.

When at max pulling the wheels barely turn (unlike most any other locomotives whereby the wheels simply slip like crazy) and current draw is very high (~7 amps). I took it apart and serviced everything but no improvement. I should revisit...

The old Custom Trains FM with the non flywheel Mabuchi can motors was a terrific puller.  These were weighted heavily and featured tires.  The truck blocks were slotted for magnets, which were not installed.  They were tough to run properly with old Lionel transformers because of the high start and stop voltages.   With modern power sources these units would behave a lot better.  The second run featured smaller can motors with brass flywheels, and much more forgiving operation.   Still have some of these around. They were built for Mike Wolf under the table, in the early '90s.  A lot of road names were offered, some of them prototypical.

In the TM book Lionel a Collectors Guide: Advertising and Art, there is an equation from Lionel Engineering Dept to calculate the drawbar pull of a loco with magnetraction.

Looking at the sample equation shown, the magnetraction provides about 1/3 more drawbar pull.

Sal,

I ordered the meter from the link in my above post. It's versatility will allow me to use it for other things.

I've got a question for you. When you test MTH engines, do you test them in conventional or Command?  Thanks

Roger

Sorry, missed this...

I only have postwar Lionel and a conventional (ZW-R) transformer. I modulate throttle to get max pull reading. This generally occurs with moderate wheel spin - either too little or too much and pulling force drops. I can't easily measure scale MPH but my hunch is peak occurs at 30-50.

So I installed the solid state reversing unit (12A model from Dallee Electronics) into that 4866 GG-1. Low speed smoothness improved somewhat and the Pullmor grind and e-unit clang are gone (kinda miss 'em if I'm honest). The motors are definitely more powerful - under pull test I now get moderate wheel spin (where as with AC power the wheels barely turned) and it now pulls at ~4 lbs. The motors also run cooler. The only two steps from here are more voltage or steeper gearing. Neither is easy so I'll probably leave it there.

Hello SAL9000.........

Of course the AC pullmore motors are more powerful is because the armature of the motor is nearly same size as the WHOLE DC can motor is.  Check it out by having a pullmore AC motor armature and sat it side by side to the DC can motor and you will see, no doubt about it smile.  I know because I had a Williams GG1 which had DC can motors and older lionel LTI era GG1 AC motors and saw the differences between them .

Tiffany

Actually, in general terms a DC motor is going to be more powerful per unit of size vs. a universal motor as it uses magnets. Quality of magnets, windings and insulation (higher quality = more flux/current = more torque) is a factor too. The issue with models trains is many use cheap DC motors made for pennies. They seem to get the job done as not many owners to complain about lack of pulling power in modern trains so I don't fault manufacturers. Even the testing I'm doing now with ~30 heavy postwar cars the GG-1 is only ~60% taxed and pulls all those cars fine - the problem now is couplers popping and cars tramlining. 

Sal,

I've got some questions for you about your rig set-up.  I've wanted to make one of these for quite awhile, but never got around to it. Recently, though, I've been working on several Alco sets (early, heavy ones) and decided to add a motor to the dummy unit to improve their pulling power. To put it mildly, it's pretty dramatic. But I wanted to quantify it with a drawbar test rig. So, I ordered the digital scale I mentioned above and this week I constructed a test track. The scale itself is very accurate (I've weighted several reference objects to determine that). But I'm not getting results close to what you've charted and I'm not sure why.

My test track is a new 36 inch 027 track (since upgraded to a bit longer to accommodate the Alcos). The attachment for the scale is similar to yours, but I'm not sure about the connection to the engines. I started out using a piece of wire (wrapped around the scale hook and then the coupler). I experimented with a spare truck (but unfortunately, it's not as free rolling as I'd like). The directions for the scale point out that anything added to the hook will throw off the accuracy and you have to use the tare function to compensate. So, I have a feeling that the addition of the extra truck and connections to the engines might be accounting for our differences in results. I've been testing mostly PW engines, but I got a red flag when I put my Williams FM Trainmaster on the track. That thing is a real stump puller, but I did not get a high reading for it and I got wheel slip (even though it has traction tires).    How exactly are you connecting your engines to the scale? I'm trying to figure out why your readings are so much higher than I'm getting. One thing is for sure......my PW engines are all great runners. I'm kind of obsessive about that. When I get a "new" engine or set (like the Alcos) I work on them until they run like silk and if I think the magnetraction is weak in a truck, I quickly replace it with a strong one. So, I don't think the differences lie in the mechanical condition of the engines. 

Roger

That has to do with hanging measurements.  Since you are not hanging the locomotives from the scale, the additions are not going to affect the readings.  What will though is if the scale will even work correctly on it's side as not all scales of this type do (Something learned in physics labs of yore.).

I apologize for not seeing this back in March.

Now that I've got a lot more testing in, here's what I've seen:

1.) Condition of both the wheels and track matters hugely. From what I've seen a NIB locomotive on new tube track can have 30-50% more pull than a locomotive and track and with lots of hours. From what I can tell run time polishes surface finish which will decrease traction.

2.) Most of my locomotives despite being prewar were NIB or have very few hours of run time. I'm also using new track (both tube and RealTrax).

3.) As you state Magnetraction strength can very a lot. I don't have a good way to measure it other than with a paper clip or the like.

Here's my complete setup. I simply use the couplers. Seems to work great. Note that I had to modify the scale handle to center when under load. When lateral I'm not sure how addition items on the hook will affect reading. It's not super accurate but my bet is it's within 1-2 ounces at these forces (1-5 lbs).

Since the original posting I've experimented a bit more. I used two early '90s GG-1s (#4866) as they have beefier shells than postwar GG-1s.

First experiment: AC power, traction tires, no Magnetraction.

I removed the traction magnets as in the GG-1s they cause a huge amount of drag. Without, the trucks spin freely as if on bearings. Now, just a miniscule movement of the ZW throttle and the locomotive begins to move. It is slow and smooth and quiet, and quite simply the best postwar puller I have or have seen. End result was ~3.3 lbs of pull.

Second experiment: DC power, traction tires, no Magnetraction, ~3 lbs additional weight (for a total locomotive weight of ~10 lbs).

This was a lot of work. The frame in this model is super soft so it bends and sags but after some time I massaged things to work. End result is ~4.5 lbs of pull. ericstrains uses the same pull test method and the most I've from his vids is 4 lbs, and that is from the Big Dogs - MTH Veranda, Big Boy, diecast ES44, etc. Now, how long this GG-1 can pull a load that big is a question. Pullmor motors won't run as hot on DC but how much less I don't know. In the last month or so I've added a longer run (~45 feet) but the limitation is now couplers popping and tramlining so I can't say how many postwar cars it will pull. My hunch is it would pull ~50 without a problem if everything behaved. This GG-1 is not as good a runner in general as it is a herky jerky at low speed - all that weight is a factor for sure, and perhaps the DC power board has some nonlinear delivery too.

Hello SAL9000

I think its pulling power is impressive but how long can you be willing to pull 50 car

trains? When you add weight you also add more wear and tear on the gears. I think

the main problenm is the worm gears which is made of brass or worse white nylon.

Those gears will wear out fast even it is well lubed. Something to think about. The

worse ones are the steamers such as #681,736 has the single worm gear so they may

be worse for wear and tear. I know what i am talking about because i Have worked on those postwar style lionel trains and the worse ones I deal with were the LTI dual motor F-3 as this engine had all nylon gears and nylon worm gear but just the gears on back of the wheels were metal. This engine stock number is 18117. I paid pretty penny for that new in the box. I spend quite bit of money to change it to all metal gears like the old postwar 2383's, took me 15 hours of labor to do the whole job(I was taking my time). I will not try to get it to pull more than 10 cars so it will last. The Williams f-3 has all metal bronze gears and the worm gear is bronze as well as i just puchased that last April and is a much much better made postwar style enigne for less.  This Williams f-3 will run smoother and pull more than those postwar style MPC, LTI f-3. I was very surprise how well its made compared to the LTI and MPC f-3's. Just my nickel's worth

Tiffany

Oh, with a non collectible as this I'd be happy to run it till the motor insulation cooked or the gears stripped if I could load up that many cars . Sometimes I find it fun to push the limits.

I doubt this locomotive will get much use though. The other GG-1 I experimented on is a much better puller - far smoother and easier to control - and though it doesn't have as much raw power I'm sure it will still pull 20-30 cars with ease. So, I guess, my lesson learned is: no Magnetraction + traction tires >>> more weight.

This statement might be true for small motors, however for 500 hp motors like those used for 1:1 scale locomotives, the universal motor was the much better choice. DC can motors were never used drive SD40's, C30-7's and such. Why, because DC can motors are not as efficient or as high in torque per size.  Until the perfection electronic commutation and inverters to drive squire cage (asynchronous induction) motors, universal motors turned the wheels on electric and diesel electric locomotives.

http://www.railway-technical.c...ml#DCMotorProtection

http://www.railway-technical.com/tract-01.shtml

Some pictures diagrams of universal DC traction motors:

These diagrams should look familiar as they are identical to Lionel's universal motor. The Pullmor motor could also be operated like the separately excited DC motor where the stator is on one circuit and the rotor on another circuit. 

Another item to consider is costs. While the Pullmor motor is more expensive than the small DC motors used in our O Gauge trains, the same statement on costs is not true for larger motors. High cost neodymium or samarium rare earth magnets must be used for a DC can motor to approach universal motors in terms of efficiency. When a motor may weigh several tons, the amount of rare earth material makes the motor cost prohibitive. Universal motors use either solid iron or iron plates which is much less in cost than rare earth materials making universal motors the much more economical choice. 

Thanks for the info. I would challenge however .

Field wound DC/universal motors were used for locomotives because as you state they were relatively cheap and easy to both build and control, and just in general in the early 20th century there were few if any alternatives for safe, reliable variable speed control of large electric motors.

The physics however for small motors are the same for large motors - for a given constant power rating a field wound DC/universal motor is generally larger (magnets generally have higher magnetic flux density than electromagnets save for something exotic like cryo cooling) and less efficient (current flowing through the field windings is a source of loss) than a high quality DC motor. The practical hitch is after ~20-50 hp magnets become prohibitively expensive (esp. back then) and variable speed control of a DC motor would have been more difficult back then as well.

Going to disagree again.  The field magnets do not impact efficiency. Here is a diagram for a synchronous motor such as those used on the Queen Mary 2.

This type of synchronous motor has a wound rotor (armature). Slip rings serve as the electrical conduction rather than a commutator. DC is supplied to the rotor. The stator (field) is wound and supplied with an AC signal to produce the rotating magnetic field. Both the stator and rotor are wound. However, these synchronous motors regularly achieve 95% efficiency. Another reason that this type of synchronous motor design is liked is that iron plates are used for the construction which limits costs. Less expensive materials + more efficient is a win win combo.

Compare the synchronous motor efficiency to that of the poultry 70% maximum for a DC can motor with rare earth magnets or 75% for a universal motor. So really the wound field does not impact efficiency at all.

Synchronous motors can also be produced with rare earth magnets where the rotor is on the inside (in runner) or outside (out runner). There is still a wound stator with an AC signal. However, these are less efficient than the dual wound field/armature design. 

Asynchronous motors like those used for locomotives or ships use are also in the 90-95% efficient range. The design of asynchronous motors excludes the need for any permanent magnet materials.