GS-4 General Problem Observations

[mmagliaro]

Admittedly I am late to the party on the Kato GS-4 in that I never bought one and have only recently
been fixing up a few of them for someone.

I am seeing some really unpleasant problems that seem inherent in the design, and I'm wondering whether if these are known problems. And if so, why aren't more people screaming about them?

These observations are based on 3 engines, so this isn't a 1-shot fluke.   They demonstrate these problems
to greater or lesser extents (meaning one runs better than the others, but I can still see the issues).

First off, I honestly don't think the GS4 runs all that well.  Below about 20 mph, it's "so-so", nowhere in
the same league as the Mikado, or the Spectrum 2-8-0 for uniform, smooth, slow-speed motion.  By the time it's under 10,
forget it.  It's a pretty uneven-running engine, if it will keep running at all.

I'm also seeing problems with the bearings.
These all have the new driver sets in them (in fact, that's what I'm doing --- replacing the drivers
 to avoid future axle-tube breakage problems).   
EDIT:
Measured the old vs new bearings with caliper.  They are the same.  Same thickness
and same width over the ears and over the flats.
END EDIT

1. The bearings have very tiny ears to hold them in the frame slots, and some of them will not stay
in position.  They can be in their slots when the cover plate goes back on, but there is so much play in them
that a little running and one or more will easily work its way out of the slot and leave the axle running in a
cock-eyed bearing.  This is particularly bad on the #4 spring-loaded driver.

2. I have seen an instance where the lateral play on driver #1 is enough to slide the nub
(on the back of the driver) out of the bearing entirely, so that it click-click-clicks trying to work its way
back into the bearing as it comes out of a curve.  This was the same problem that afflicted the Con-Cor 2-10-2.   

Narrowing the driver gauge cannot fix it, because you can't squeeze the drivers close enough together to
avoid the problem without making them so narrow that they will never get through a turnout.
(Pulled a wheel, slipped on some PB washers to limit the lateral play... fixed it, but... that's
a pretty serious flaw).

3. Running anywhere under 7 volts or so (DC), which means 35-40 mph, if it pulls a train up my 1.7% grade,
it slows... way.... down... and will almost stop if I don't juice the throttle.  At 6 volts, it will stop.  This is
with 20 freight cars on it for testing.   That seems like a terribly underpowered motor.   

4. All the wheels are narrow. They are barely in the notches on the NMRA plate, but not "in" well enough to
reliably track through my Atlas code 55 turnouts.

People seem to love this engine so much, I am surprised by what I'm seeing.

Thoughts?

[peteski]

I own couple of GS-4s (SP and the AFT). Still with the original drivers (undamaged).  I bought new drivers for them but haven't installed them yet.

Mine didn't get much running overall but I haven't noticed any of the problems you mention.  I run them on DCC. They seem to have plenty of power to pull the Daylight consist (which is what these passenger engines were designed to pull).    I ran it on friend's layout and on N-Trak setups.  N-Trak has no grades and I'm not sure how steep the grades are on friend's layout.

I did notice that the motor in these looks like a shrunken down version of the Kato's standard N scale motor, but with rare-earth magnets.  I haven't done any performance testing on the engine or the motor (simply ran the Daylight train).

The only problems I found were the original traction tires were relatively soft (came off easily) and a non-conductive wheel blackening substance on the AFT version.  The new version of the traction tire (for GS-4 and FEF-3) seems much better (less stretchy).

Of course, there was also the initial problem with the DCC-converted models blowing up decoders, but that was quickly fixed.

[carlso]

Hi Max,

great write-up. The KATO GS-4 was a "must have" for me and I did buy one from the first production and am still running it today, admittedly it does not run often. Still has the original drivers, I have a set of the new ones but have never tried to remove/replace rods and such. With only one geared axle, quartering should not be an issue. Am I correct or do I not understand. The original shorting issue was found before I went to put a TSU-750 sound decoder in the tender. I was not patient enough to wait for KATO's fix so I dug out my Dremel and removed metal from the sides of the frame so nothing would touch, never had a problem. I basically made "wheel tubs" ala the old hot rod days. Oh yea, the traction tires lasted about two outings, so I very carefully applied Bull Frog Snot into the tire grooves with the unit running upside down, in a cradle. I applied several thin coats and left the loco running until the snot was fully cured. It has lasted 3-4 years now.

Running characteristics are not bad but I have been blaming the Tsunami decoder for having poor motor control, at best, and after your post perhaps it is the loco. It will pull the full Daylight consist up, down, and around 2% grades, with S curves, on our club layout without a lot of adjusting the Digitrax throttle. I can say, 1st hand experience, it is NO WAY close to the FEF in performance.

I have been wanting to replace the Tsunami with an ESU Loksound Select Micro decoder to get their excellent motor control. Originally, I put a Richmond Controls Light module in the boiler in order to replace KATO's blinking mars light. The Richmond Controls is fantastic, looks very real. I have been thinking about tearing it down, to a point, to install the new drivers. Maybe a good coreless from "The Motor Man" could be set in the boiler by making some sort of special cradle ?

Carl


PS: I sure hope we don't find issues with the FEF down the road.

[victor miranda]

Hi Max,

I am puzzled.
how did you stretch out the gauge?

kidding aside,
I am thinking I need to see if I can get a photo or two of the frames.

we can compare tolerences.

on to the power of the motor...

it uses very strong magnets and it does have a mild cog I can feel
as I roll the motor.

I am thinking that the magnets got overheated and did not recover?
leaving the unit somewhat weak perhaps?

from all you have said it sounds like the motor may have been working darned hard in any case.

victor

[victor miranda]

I went hunting

found this...
http://forum.atlasrr.com/forum/topic.asp?ARCHIVE=true&TOPIC_ID=44904

in that thread I mentioned current/voltage/speed
"mine started at 61ma at 3.25v and
this throttle setting gave me timed speed of 20 smph"

so that is a ball park.

pictures when I get home tonight.

[mmagliaro]

On the current one, which is the most troublesome to be sure, it will start at 3.75 volts (level track, no cars),
and draws about 55-60 mA at that voltage.

EDIT:
I spread the drivers on axle #1 using a wheel puller.  The axles do protrude through the plastic wheel centers
in the center of the little stars.  The tips that show there are very small, yes, so you need to use a very
small ram in the wheel puller.  But it doesn't take much to press them.  They are just splines in a plastic
center.  (Not a fan of this, by the way.)
Important: This problem on axle #1 coming out of its bearing was present before I even messed with the gauge.
I pulled the wheel, put the washers on it, and pressed it back on, and set the gauge as narrow as I can to
still function correctly on turnouts (on the narrow side of the NMRA notches)
The splines aren't as troublesome as I originally thought they would be.  Since the axle ends are tapered,
you can ease the wheel on, quartered by eye, and you can feel the spline find its way back into
the grooves on the inside of the axle hole and get the wheel pretty well started by hand.  Then press the rest
of the way.

I'm not having any wierdness with wobbles or wonky driver motion.

This engine is almost brand new.  It only ran a little bit to make sure it worked (the owner tells me, and I trust his
accuracy on statements like this).   

I will try to photograph the issue with the bearings.  It may be hard to illustrate just with photos.
I may have to take some video so I can show how they can easily pop free of their guide slots.

At first (and second and third...)  I thought that maybe I didn't have them correctly in the slots, or "all the way" in the slots, or something simple like that, but no, they are in.  I can even look at an untouched one in the box, and
note that just by bobbing the rear axle up and down on those fingers, you can see the bearings come right out
and kind of sit at a 45 degree angle to the axle as they bob up and down, and then sometimes go back in.

Photos when I can... to explain this better.

I am noticing that the most troublesome one is definitely running better the more I run it.  And perhaps lots of hours
on it will smooth it out.    My experience, however, is that if an engine isn't running good after running for an hour or
so, it isn't really going to get very good, ever, without some problem being addressed.

[mmagliaro]

I have fixed the bearing problem.

Here are some photos that illustrate the problem and the fix.
After fixing this, the engine runs great. (Well,  it runs like a GS4 should, anyway).

It is important to note that I see this wonky bearing behavior in all the GS4's I've recently been tinkering with.
This one was severe enough to jam the engine.  But even on the others, when I turn the engine over and
poke that rear driver up and down, I can see that the bearing is one eyelash away from popping right out of those slots.

[mmagliaro]

While we're at it...

The #1 driver's bearings do not float.  When you put that driver in, those bearings lock hard into the slots down in the frame.  What's up with that?   That seems like a terrible idea for staying on the track.
Same deal: all 3 GS4's I've seen are that way.

Is that somehow a good design?  To have one axle locked and let the others float?  Victor?  Peteski?
Anybody?  If it is a good idea for some reason, hit me over the head and explain it to me.  (Well, you
can leave out the hit me over the head part).

[victor miranda]

going from memory.

the first driver bearings are not 'tight' they can wiggle just a little.

I think of it as a drop in type fit.  the axle and bearing can plop into place.
and the coverplate will not snug down on them

.... max I have a unresolved problem with your photo....
something does not look quite right and I can't figure it out.

I'll look tonight maybe I can see it then.

I recall that the last driver bearings sit pretty high in the bearing block slot
and that when the cover is screwed down, the bearing block can't slide down (up?) enough
to escape the slot holding the ears.

victor

[mmagliaro]

going from memory.

the first driver bearings are not 'tight' they can wiggle just a little.

I think of it as a drop in type fit.  the axle and bearing can plop into place.
and the coverplate will not snug down on them

.... max I have a unresolved problem with your photo....
something does not look quite right and I can't figure it out.

I'll look tonight maybe I can see it then.

I recall that the last driver bearings sit pretty high in the bearing block slot
and that when the cover is screwed down, the bearing block can't slide down (up?) enough
to escape the slot holding the ears.

victor

Perhaps that is true.  Let's standardize the nomenclature.
When the engine is sitting on the track, the driver cannot push upwards into the frame, pushing against the
spring fingers, so that it goes high enough to let the bearings out of the slots.
Is that it?

If I have interpreted that correctly...
The problem is that they don't pop out when they go UP (again, when engine is sitting on the track).
They pop out when they come down.   I briefly considered cementing some small extensions onto
the little nubs that are molded into the gearplate cover, so it would push the bearings deeper into
the frame.  That actually might work, and would be simple to implement.  But I worry that the rear driver would
now ride a little higher than it is normally supposed to, and that could have unwanted side effects., like not
letting the wheel come down enough when the engine goes over uneven track.  The wheel could find itself in
midair and then come down and derail.

[peteski]

I'm also seeing problems with the bearings.
These all have the new driver sets in them (in fact, that's what I'm doing --- replacing the drivers
 to avoid future axle-tube breakage problems).   
EDIT:
Measured the old vs new bearings with caliper.  They are the same.  Same thickness
and same width over the ears and over the flats.
END EDIT

I spread the drivers on axle #1 using a wheel puller.  The axles do protrude through the plastic wheel centers
in the center of the little stars.  The tips that show there are very small, yes, so you need to use a very
small ram in the wheel puller.  But it doesn't take much to press them.  They are just splines in a plastic
center.  (Not a fan of this, by the way.)

I have noted this when the new drivers first appeared on the market in https://www.therailwire.net/forum/index.php?topic=33803.msg395559#msg395559.   Seems like linear fluting was the best they could do to make a retrofit axle with bearings which will fit the existing model. The FEF-3 was redesigned with much beefirer axles and larger diameter bearings.

The #1 driver's bearings do not float.  When you put that driver in, those bearings lock hard into the slots down in the frame.  What's up with that?   That seems like a terrible idea for staying on the track.
Same deal: all 3 GS4's I've seen are that way.

Is that somehow a good design?  To have one axle locked and let the others float?  Victor?  Peteski?
Anybody?  If it is a good idea for some reason, hit me over the head and explain it to me.  (Well, you
can leave out the hit me over the head part).

This is by design Max.  That was part of the revolutionary design of this model (why it pulls so well).  I have also described this in few older posts  (one in your thread) https://www.therailwire.net/forum/index.php?topic=30346.msg333242#msg333242



Mechanically this loco behaves like a 4-4-4 locomotive.  The first and 3rd driver axle is fixed (vertically and laterally). That is the middle "4" in 4-4-4.  Most of the loco's weight is over the 3rd (traction tire) driver but the 1st driver also supports appreciable weight. The 2nd and 4th drivers are strictly cosmetic. The 2nd driver has plenty of free vertical and lateral play and it does not support any of the loco's weight (on level track).  4th driver has has lateral play and vertical play. The vertical play is controlled by a very soft sprung suspension. That spring is there to make sure that it follows the track behind the 3rd driver without jumping track.  If you don't want to call this a revolutionary solution, it is at least a very elegant solution to maximize the pulling power of a steam locomotive model.

Also playing devil's advocate, when a real steam loco pulling a heavy train starts climbing a grade, it will also slow down and need a wider open throttle to keep a constant speed.    Also, if DCC is used, the BEMF circuit will somewhat compensate for increased load while climbing a grade.

[mmagliaro]

Thanks Peteski.  Yes, I remember that drawing and explanation.
(Very nice, by the way!)

But I would still not expect the #1's bearings to be really locked tight.  I mean, sure, no vertical slop
to speak of, and it should be firmly on the track (unlike the floating #2 or #4).  But I would still expect
some nominal wiggle in the bearings.

When I say "locked" I mean... really locked.  You push the bearings in and put the cover plate on, and the
cover has little nubs that protrude down into the bearing slots.  Once you've done that, if you take the cover plate
off, those front axle bearing are in there so tight that you almost need to pry them out with a screwdriver
(don't do this of course... bad idea   )

In fact, if the #3 with the traction tire and gear on it were locked this tight, the engine would be running
on 4 rigid, spring-less "legs" and it would probably wobble like a table when it went over anything but perfectly
flat track.

The extreme slow-down-to-a-stop on the hill was related to the bearing problems.
Now that I have that fixed, the whole engine can start, run, and pull the train up the hill a full volt lower than it
could before (5 vs 6).   It slows, but keeps beating along smoothly and pulls the hill.
I'm with you on that.  I don't mind some slowing on hills because that is natural prototypical behavior.
I balked because this one had an excessive slow-down in my opinion.

In any event, with the "U" bracket in there, it's running quite nicely now.

[SkipGear]

I believe #1 is locked down because it is side rod driven. The majority of the weight is on the geared driver so that is what keeps it in check. The #1 driver runs light so slop free bearings lessens the chance of the driver wobbling under the drive of the side rods.

[peteski]

I believe #1 is locked down because it is side rod driven. The majority of the weight is on the geared driver so that is what keeps it in check. The #1 driver runs light so slop free bearings lessens the chance of the driver wobbling under the drive of the side rods.

Good theory but on this loco only #3 is gear driven - all the other drivers are driven by side rods - which just like on the 1:1 loco are not a single solid rod spanning all 4 drivers but segmented into separate side rod between each pair of drivers. So the very loose #2 (which is free-floating) and #4 drivers are driven by side rods and they don't wobble because of that.  I have to take my (smooth-running) GS-4 out of its jewel case and take a look at it.  It's been a while...

[SkipGear]

Good theory but on this loco only #3 is gear driven - all the other drivers are driven by side rods - which just like on the 1:1 loco are not a single solid rod spanning all 4 drivers but segmented into separate side rod between each pair of drivers. So the very loose #2 (which is free-floating) and #4 drivers are driven by side rods and they don't wobble because of that.  I have to take my (smooth-running) GS-4 out of its jewel case and take a look at it.  It's been a while...

I never said #1 was gear driven. #1 supports some of the nose weight of the loco and thus does add slightly to the tractive effort. The #2 driver doesn't do anything but set there and look pretty.  the #4 driver is effectively the wheelie bar to keep the loco from nosing under extreme load and really doesn't contribute much to the tractive effort of the loco either. #2 and #4 are basically freewheeling. #2 only serves to hold up the side rods and when the loco is under load it will wobble a bit because of it's slop. The load on the #1 driver is what causes the #2 to wobble because of the stress on the side rods.