Lifelike/Walthers (LL) Berkshire tinkering

[mmagliaro]

So, on the heels of drifting the thread of the Bachmann Berkshire engine review and discussion off course, I am
taking my remarks here to keep them out of the way of that discussion.

I discovered something quite shocking tonight as I continued tinkering with my Lifelike Berk.  So in addition to
all the other well-publicized "fixes" for that engine (adding weight, tape on the driver, cutting a traction
tire groove, and so on), here's what I found.

This got very long, so you can skip to the end if you already know all this stuff and just want to see the
"new" discovery.  Sorry there aren't any photos this time.   I may yet take some as I continue tinkering.

First, some history:

My Berk currently weighs about 106 grams.  The stock engine weighs 86.  Early in its life, I opened the smokebox
front, which just pops off, and I stuffed as many little tungsten disks in there as I could fit and still allow me to get the front back on (and not have a disk short across the frame gap at the top).

I also glued as many little disks up above the pilot truck as I could so that when they are all painted black, they don't show.  That got it up to about 96 grams.

Finally, I "winged it" and cut a groove in the #4 driver by holding a spinning Dremel cutoff disk on the wheel as it spun upside down in a foam cradle.  This last move was insane.  I expected I might just destroy the wheel, but I got away with it, and it worked.

At the end of all that, it could pull about 20+ cars on flat track and pull, I think, 16 or maybe more up a 2% grade.
It was way better than the stock engine.  I also tried Bullfrog Snot first, but the tire produced much better results.

I also added 15g of weight to the tender, and with two live trucks in mine, the pickup has worked well that way.

-------------------------

Fast forward to now, 2015.

Everyone is talking about the Bachmann Berk, so I dragged the LL engine out to see how it runs these days.
Since I now have a milling machine, and I'm a little more brave about pressing drivers off axles and spinning them in a machine, I pressed the #4 drivers off, chucked them in the mill, and recut the traction tire grooves with a proper tool.  All I did was dress off a little more in the groove so that now it is all even and square.   The Dremel-chopped grooves weren't too bad, but they are better now.  I was able to grip the wheel in the chuck by the little metal nub on the back of the wheel.

The results?
It could pull about 40 cars on flat track and 32 up my 1.7% curved grade.   I was happy with that.
It was, however, on the hairy edge of lifting the nose with all that load on it, so I weaseled some more tungsten
into the front.  I also took the shell off and stuffed lead or tungsten up into the big rectangular dome,
and over the top of the worm, and I put tungsten disks in all the larger open voids between the frames
that you can see when you open the gear cover plate on the bottom.  Just be sure not to let a piece of
tungsten touch both frame halves.

So now I was up to 106 grams.  But it didn't really pull any more cars, and the lifting problem wasn't really any better.

Then I decided to make a whole new drawbar and attach it to the post right behind the rear driver, instead of
on that post that is way at the back of the cab (a terrible place).   I cut a drawbar from a piece of
printed circuit board, slit the copper plating down the middle, and soldered .010" phosphor bronze pressure
wires at each end.  I then drilled and tapped 00-90 screws into the underside of the frame for the wires to
press on at the engine end.  On the tender end, my wires press on the tabs just like the original drawbar.
This allows me to take out 1 screw and separate the engine and tender.

This also did not help the lifting problem or make the engine pull any more cars.  I was a bit surprised.

Tonight's Discovery

I noticed that when people put tape on a driver on this engine, they always seem to talk about doing it
to driver #3.  My traction tire is on #4.  They are both geared.  I figured that having added all that weight and moved the drawbar, the only reason I wasn't getting any more pull must be that #4 isn't down on the rails enough.

I thought, "what about the old "shim-the-driver" trick?  So I slipped a .005" shim under the bearing blocks on #4.

All I can say is.... WOW.
It just dragged 46 cars over the 1.7% grade at really low speed without the slightest hint of slip, and all hint of lifting on the curves, even after increasing the car load from 32 to 46, is gone.

I don't know the limit yet.  I'll keep adding cars.

Perhaps we have been thinking about the geometry all wrong.  #4 wasn't getting enough pressure, even with a traction tire on it, which is already making it stick out a little more.  So, I suspect that #3 is sticking down,
like the fulcrum of a see-saw.  Even though it doesn't seem to rock on the track.  And perhaps, jacking up
the back by .005" not only increased the weight on #4, but keeps the nose from lifting.

And if you put your strip of double-sided tape on #3, you might get more grip, but you also might
make the nose lifting problem worse.

At least that certainly seems to be what happened.  I don't know if that shim would have the same
effect on a stock engine.   But I would like to know.   Double-sided tape and a shim on #4.  Maybe
that is the way to go.

[mmagliaro]

Okay... the limit is 52 up the hill.
On level track, it can pull more.

I am going to try the rear driver and some other tweaks on an untouched LL Berk I have here.
I won't go all Dr. Frankenstein on it.  I want to try measured things that would be easy to do first.

[Chris333]

I built mine some time ago, but I have another in the works. Erie didn't use Buckeye tender trucks so I never had the problem with their pick-ups. I used Concor Commonwealth trucks. I can't recall where I put weight or what it currently weighs. I do know I did the same trick with the drawbar.

This white line is how my draw bar was bent:



The reason I went down under the axles is because I put a styrene "wall" viewblock where the ashpan would be hanging down. This is so you cannot see light above the trailing truck. It makes the locomotive look much more bulkier and that light through that area is one of my pet peeves of model steam locos.


My current project that resides in a box in pieces. I chopped off a big area of the frame at the front of the boiler for a pretty large tungsten rod:


It also has a TT groove cut in a driver so I have high hopes for it. Perhaps my disappointment in the Bachmann model will get my back working on this one.

[jdcolombo]

Okay... the limit is 52 up the hill.
On level track, it can pull more.

I am going to try the rear driver and some other tweaks on an untouched LL Berk I have here.
I won't go all Dr. Frankenstein on it.  I want to try measured things that would be easy to do first.

Well, this sounds really promising.   Maybe I'll take one of my LL Berks, remove the tape from the #3 driver, put it on the #4, and use an .005 shim on the #4 and see what happens.   The "front driver lift" problem is more noticeable on two of my 10 LL Berks than it is on the other 8, but I never pull more that 25 cars with one (that's the length limit for my passing sidings and staging yard tracks).   If you can pull 40 cars without driver lift, however, then 25 should be no problem at all.

The other issue I've had, which you probably have solved with your drawbar mod, is getting reliable electrical transfer from the tender to the engine and vice-versa.   I've installed Tsunami sound decoders and a speaker in all my tenders, but the speaker enclosures are made from 1mm lead sheet, which makes them pretty heavy.   I think the electrical pickup from the tender is pretty good, but sometimes the engine stops while the sound keeps going - indicating that the tender is getting power, but the power isn't being reliably transferred to the engine.   A new drawbar might help cure this problem.

Good thoughts here - I need to do some experimenting . . .

John C.

[mmagliaro]

Oh!  The drawbar transfer to the engine is your problem.  Okay.  I thought it was in the tender trucks themselves.

Well, personally, when I am finished with this engine, I will more than likely just solder permanent wires at both ends.   With super flex wire and good wire routing, the incidence of the wire breaking off is about zero for me,
and far more reliable than any drawbar wire scheme.

I don't think my new drawbar is any more electrically reliable than the original.  It's just attached in a better place.

Chris, your drawbar is *amazing* !!!  What an elegant piece of engineering.   I'm afraid mine is just a simple straight shot right across the top of the trailing truck.   Blocking that air space above the truck is great.
I was thinking of attaching some wedge-shaped pieces of styrene to the sides of my drawbar to
accomplish the same thing.  It would be phoney in that my ashpan sides would actually move with the drawbar!  But I think watching it on the track, nobody would notice and it would look better.

I will eventually take some pics.


And yes, I would love to hear what happens if you put your tape on driver #4 with a shim instead of putting it on driver #3.

[peteski]

Blocking that air space above the truck is great.
I was thinking of attaching some wedge-shaped pieces of styrene to the sides of my drawbar to
accomplish the same thing.  It would be phoney in that my ashpan sides would actually move with the drawbar!  But I think watching it on the track, nobody would notice and it would look better.

Does it matter that it is attached to the drawbar, as long as the end result is no daylight under the firebox?  Isn't that how Kato designed their FEF-3?  I thought that was a very innovative solution to the problem plaguing N scale steam locos since the beginning of N scale.

[chessie system fan]

That's good to know. Max.  I noticed while working on my 2-10-4 project that LL originally planned on gearing three drivers instead of two.  You can see it planned on the underside plastic piece and also in the parts drawing, and since then I've wondered what would happen if a third geared driver were retrofitted.

 

[Lemosteam]

All good news here.  After all of these traction threads of late, something is becoming clear to me.  All we really have to do is get ONE geared TT (or 2 sided tape, right JD?) driver solidly on the track to provide pulling power.  Based on Max's thoughts, it would seem that on this loco #3 driver is becoming a fulcrum (wheel surface is proud of the rest when compared to a plane tangent to #1, #2 and #4) for the load of the cars, inadvertently lifting the front end.  Maybe that was LL's design flaw on this loco, thinking that the #3 driver should be proud of the rest, after all that's what the Kato Mike does.

I would bet that one could remove all of the added weight and still achieve decent pulling power and lead truck tracking with a traction tire on #4 and that driver being the one that is proud of the rest. 

Max can you try only that before adding any weight whatsoever?

[randgust]

I don't know about the TT or axle loading, but moving the drawbar to behind the 4th driver was an early modification.    I also added a lot of weight to the nose, in the same way you guys did.   But the reason I moved the drawbar was entirely different.

I've got mostly 13" and 15" curves, what I noticed was the front driver on the outside rail climbing up over top and derailing under load.  The position of that rear post on the stock engine leverages the entire locomotive into that rail under load, literally turning it into a lever against the drawbar post of the tender on a curve.   Something has to give.   So I imitated the Kato 2-8-2 design of taking the drawbar behind the rear axle, and as low as I could get it.   That solved the 'climb over the outside rail' problem.   Bluntly, it didn't do much of anything either for the drawbar pull either.    The original post is low enough, it just is so far back against the wheelbase that it jacks it into curves.

My drawbar design was more like Max's, the only thing I did different was a cutout on one side so it didn't hit the injector.   Made mine of strip aluminum so I could bend it over and around kind of like Chris'.

[mmagliaro]

John... and all..

My Berk is currently a "Frankenstein" monster.  I have done a lot of experimenting, shifting weights, cutting
the tire, adding the drawbar, and so on.   So yes, I am going to do another one, starting with a pristine
example, and see what simple things might help first.  Ultimately, I will cut a traction tire groove, because
I know that is going to make the pulling power go up no matter what.

When I moved the drawbar on mine, it did not help the pulling one bit.  That agrees with Randy's
observation, that moving the drawbar did not add to the pulling.   I wasn't having a problem
with it lifting over the rails anyway, so I can't tell if moving the drawbar helped.  But then, I had
added a LOT of weight to mine (20g).

As for the drawbar attachment height...
Remember, it's not where the drawbar hooks onto the post that matters.  It's where that post attaches to
the engine.  That is where the pulling force is exerted.  That post runs upward above where the drawbar
hooks on,  and is attached at the frame under the  cab.  That's where the pull is from. 

But it still really isn't all that bad.  It's not as high as it looks:



Chris's drawbar, running UNDER the trailing truck, is even better because it attaches even lower.
But the stock drawbar attachment height is just a little above the axle height of the driver.  Right on
the axle height would be ideal, but it's not bad.  It's just the long distance toward the back that
causes problems.

So that explains why moving it forward doesn't matter much to the pulling.  I think Randy is right.
It helps the climbing problem not because it's lower but because it's not as far back, so it isn't
swinging the engine around sideways.

==========================================================================
So... let's begin.
This engine (the one in the photo) can pull 15 cars on level track, and 8 up my curved grade.
It only has one live truck in the tender.  It can run smoothly and uniformly at 3 mph.  I am not seeing
any stalling issues with it whatsoever (which surprises me, having only 1 live tender truck, but that's what it
does).  It doesn't climb on curves, but then, it can't pull enough cars for that to happen yet.

Modifications so far:
1.
The pilot truck did have an annoying habit of derailing, so there is a MT plastic washer above the pivot hole
of the pilot truck now. That causes the pilot truck to shift downward by
.020", and that is enough to make it track very reliably.  It had zero effect on the pulling. 

2. Added about 6g of weight in the nose, and inside the gear cover plate.
Had no effect on pulling whatsoever.  Still pulls 15 and 8 cars, respectively (level and grade).

----------------------------
My next move will be to shim #4.  If we believe that there is a lot of leveraged weight on that driver,
perhaps it will pull a bit more, even without rubber on it (I doubt it, but I want to know).

[Lemosteam]

Great Max.  Hoping you will you be able to pull weights later to see if they have an effect after the TT is added...

[randgust]

Usually I get into this debate only with the diesel guys when the argument becomes that you increase tractive effort by going to a five pole motor, changing gearing, or something like that.

There's only two viable ways to mess with tractive effort.   Pile more weight on the wheels, or modify the coefficient of traction between the wheel and rail surface.   Nothing else really matters significantly unless you're vibrating the wheels to break the coefficient, or lifting the wheels off the rails to reduce the coefficient.  Bad motors and gearing can do that, but the basics are weight and that adhesion calculation.

So it's all about piling weight on the drivers and seeing what you can do about getting more traction.    I use my spring-scale dynamometer car and my electronic scale and do it all in grams, the spring scale isn't dead accurate like the electronic scale but it is comparatively accurate between models.   'cars' is inconsistent here, as there's just no comparison between an out-of-box MT with fresh wheels against one that's got worn, gunked-up treads, let alone the variances on car weights.  Ounces is fine and the math results should be the same as long as units are consistent.

The 'best of breed' even with traction tires is rarely more than 20%, I'll get as high as 21% if you divide pull/weight for a calculation of the same units.   What's bad is that with the wrong (slick) wheel material you'll see that pancake down to 8%.   So as you're tinkering, that's the number to watch here an set up some kind of consistent test to get it.  It's been a while since I tested my Berk but I think I'll give it a fresh test so some comparative numbers can be run here.   My champ is still the GHQ L1, equipped with the traction tire, which will even put any diesel I own to shame - it's the single-unit pulling champ I own and almost off the scale (literally) that I can measure.   My opinion, which should be data-proven, is that the LL isn't as light as it is slippery, and it all goes back to the wheel material.   But we'll see.

About the time you watch a lighter Kato outpull a heavier Atlas you'll realize that wheel material (even without traction tires) has a lot to do with it, and wheels with the plating worn off (raw brass) far outpull new ones.   I'm kind of a fanatic on this stuff, admittedly, and really, really like things like this where I know Max takes as much of a scientific approach to this as I do.  I'd 'like' to think that a more modelers catch on to this approach, the automatic willingness to accept new models that can't pull worth anything will be suitably challenged and addressed.   If you start testing most legacy units out there, you'll find they WAY outperform today's production, one of the few areas that N has significantly backslid and given manufacturers a bye about it.    I'm from an era where it was expected that a model was capable of pulling a train beyond a handful of cars.    The other dirty little secret is that the rolling qualities of cars and wheels is FAR better than it was in the old days, so once again the old MRC/Rowa 2-8-4 that could pull '8-10 cars on a level grade' in 1969 could probably rate as 15 today just due to that factor.

I'd appreciate some comment though on what is 'fair' as a method of testing, do you want to do total locomotive weight including tender, or measured weight on the drivers only?    Pull is easy, but if you include tender weight you'll penalize some calculations.    I'll include the tender for the drawbar pull (which also penalizes for leading and trailing trucks) but I think weight on the drivers is the denominator for steam and the pull is the numerator, comments welcome.

[mmagliaro]

Randy,

I'm not really doing this scientifically, but I am doing it so it is repeatable, comparable to previous results,
and testable.

One thing I do, which I haven't mentioned, is use the same string of cars as I test the engine.

In other words, I have a string of cars sitting on the layout,  8 in the case of this Berkshire.
And I will make mods and then pull those same 8 cars  plus some more that I will couple behind them.

And then I will make another mod and use that same string, and add some more, and so on.
So while this doesn't mean that anybody else would get the exact same car count when they make
these same mods to their engine, it does tell me whether or not I am increasing pull at each step.  I also
use only 40 foot cars that are mostly all weighted the same (24g).   I say "mostly" because they
aren't all MT and aren't all equally weighted, but most of them are close.

I don't necessarily use the same string of cars from engine to engine, so results across engines aren't
comparable... but they are close because, heck, I've only got so many cars and most of what I test with is
the same string sitting here on my layout.

Within the realm of our hobby, which is not a physics lab, it's good enough. 

-----------

I TOTALLY agree with you on the slipperiness of the metal on the drivers.  I can see it with my eyes and feel it with my hands on the LL Berkshire.  Those darn drivers are smooth and glassy, moreso than most engines.  That's a big reason it doesn't pull well.  And I will wager that's why adding 6g of weight on the nose didn't get me even one more car of pull.  If you have a really low coefficient of friction at the drivers, you need to add a LOT of weight to make any difference.

It is a lightweight, though.  The engine alone only weighs 86g.  By comparison, an old-generation Kato F unit weighs about 125g.

-----------

[mmagliaro]

I put shims on the rear drivers with no other mod (no traction tire and no added weight beyond
the nose weight I have in there.  As I expected, it cannot pull any more cars.   I just wanted to make 100% sure
of this.

Cutting the TT comes next.

[mmagliaro]

Here's the traction tire groove cutting.   This is the same way I've done groove and flange cutting before.
The little hub on the back of the driver clamps quite nicely in the chuck.
Allowing that the Kato Mikado traction tires I am using for this are about .010" thick, I want about a .007" deep groove.  So I cut and periodically stop and caliper, until I see a diameter across the cut that is about .014" smaller than the tread diameter.

With the Kato tires on this, the engine's pulling soared.  It could only pull 8 cars up the hill before.  Now it is pulling 40 and that's not the limit yet.  HOWEVER, the front driver is starting to lift.  It's not enough to derail on curves yet, but all that pull on the tail is starting to lift it up.  I do not have any extra weight inside the body on this one.  My other one had all the weight I could pack up in the big rectangular dome and over the worm cavity inside.   Based on the owner's wishes with this one, I may not crack open the shell and go that far.  It can pull between 30-40 cars even on grades with no trouble, and that's more than adequate.

However, adding the traction tire gave up some pickup on the drivers, and that was causing it to have some stalling problems, so I had to do some work on the tender trucks.  See the other thread on ConCor Hudson and LL Berk tender trucks for that.