GS-4 re-motor project update

[peteski]

The Kato GS-4 drivers (especially the new ones with splines in the axle) should be quartered perfectly out of the factory.  I will be really surprised if they aren't. Unless of course you pulled the drivers off the axle and then reinstalled them slightly out of quarter.

Yes, this loco has a fairly loose mechanism but it is "precision loose" to coin a phrase. It is designed to be that way to traverse sharp curves and uneven track, yet remains very smooth in operation.  This loco runs really well even with the stock motor.  It can even smoothly pull the full Morning Daylight consist up grades (as shown in the video on Kato USA website).  Whatever is happening on your loco is highly likely to be introduced by your conversion.  The binding seems to be multiple times per driver revolution which should be something other than the drivers or valve gear.  If you can, remove the worm and then push the engine down a piece of straight track (using your hand).  It should roll very smoothly.  If that is the case then the driver quartering and valve gear is all working correctly.  Then concentrate on troubleshooting your conversion.  If the loco still binds without the worm installed then the original mechanism has a problem.

[mmagliaro]

I don't think it's the quartering even though the counterweights may not appear to always
be in the right alignment.  The Kato driver sets are new in that engine, and they do tend
to have a lot of play in them, making them look like they might be out of quarter because of the appearance
of the counterweights, but they aren't.   

I like Peteski's idea,
You can just pop the worm out and then try rolling the engine on some track and even rolling the wheels in your fingers
to see if you feel a trouble spot.

But if you don't,  I was going to look at this problem from the other end - the motor.
Gearheads make noise, but I don't like the noise I hear in there.
That noise sounds like the pinion making noise in the gearhead.

Take out the worm and the rubber tube, and just run the motor with a little black
Sharpie mark on the output shaft from the gearhead.  You should be able to run that thing at well under a volt,
in fact more like 0.5 volt, and you should be able to get that gearhead output shaft turning so slowly that you
can easily count the revs/minute by eye.  If you can't, something is amiss in that motor.

Sometimes, those gearheads aren't mated onto the motor quite perfectly. 
The motor pinion can bind in the planetaries.   THey seem to run "pretty good', but noisy and
they don't have the good low-end running when they are under load.

I would take the motor out and measure the current with the gearhead on it.
It should be less than 20 mA at 6 volts.
 It it's more like 30-40 or more, the gearhead
is not fitted to the motor correctly. 

Then unscrew the gearhead from it, and then measure the current of JUST the motor.
At about 6 volts, it should be no more than 10 mA.  In fact, it will probably be less.
Then put the gearhead back on .  It should be no more than 20 mA.

Many times, you will find that just unscrewing the gearhead 1/4 or 1/2 turn
on the motor will make it quiet down and draw a lot less current

I fine-tune mine by running the gearmotor with an ammeter in-line on the bench, and I
screw the gearhead in and out a little, watching for minimum current draw.  When I find that spot,
I put a few drops of cement (a little epoxy, loctite, etc) at the joint between the motor and gearhead
so it won't unscrew from that spot.

[victor miranda]

keep in mind that the counter weights are NOT always square to the crank pin.

this is most often true for the driver with the connecting rod.

the engine does not act consistent with a quartering problem.

[carlso]

Peteski
Max
Victor

Thanks guys for the suggestions. First Pete, I think it is something in the mechanism and yes I have pushed the mechanism on the table top without the worm and it hags up somewhere. It does not require a great force to move through the problem, that is why the gearhead will force it along.

Max, I am going to do as you suggest and test the motor with/without gearhead. I don't know much about this stuff, obviously, but I did not think the gear box should sound off like thais one regardless of the connections used.

Victor, I agree with your comment.

Thanks for replying.

Carl

[Loren Perry]

At about the 13:36 moment in your video, I observed the front end of the chassis lift (apparently caused by the roller block coming dislodged from the rails) and the lead drive wheel (No. 1 axle) "derailed" with the flange riding on the roller instead of the wheel tread. It is at this moment that the severe vertical movement of the No. 2 driver set begins.

I examined my own Kato GS-4 - a later model - and realized that only the third driver set with the traction tires is gear driven. All the others are rod driven. This may or may not affect the absolute necessity for quartering precision. My engine runs flawlessly with a 12-car train in tow as well as running light. I've never done a thing to it since purchasing it. And there's plenty of freeplay in the drivers making it a fairly "loose" engine compared to some of my brass models.

I also have observed a noticeable fore-and-aft "surge" in time with the driver rotation, much like the surge I mentioned here some weeks ago when I was dealing with a brass ATSF 4-8-4. I still haven't fully solved that problem.

But at this point (13:36+) in the video, I now believe your engine's surge may actually be a yaw (hunting) because I can see the fore and aft movement at the front end but none at all at the rear. The camera angle to the chassis may be helping me to pick this up. If it were an actual surge, uniform movement would be visible at both ends. The hunting would almost certainly be a result of the lead driver derailing.

Because the anomaly appears to be timed to driver rotation, whatever is causing your surge and related problems is almost certainly to be found in the drivers, rods, and related hardware and not in the motor/gearbox. If it were the latter, the surges would be spaced far more closely together due to those components' higher rpm's.

One more thing to check: at the earlier part of your video, a definite fore and aft surge (the whole chassis from end to end) is easily seen and it is timed to the driver rotation. It's as though the traction tires are "biting" into the rollers at one point in the rotation and shoving the engine slightly forward. This forward thrust would require some binding in the rollers, particularly under that wheel. Have you confirmed that all your rollers are absolutely free-wheeling under all conditions?

I assume you have at some point disconnected the rods from the drivers and finger-spun each axle to confirm there is absolutely zero binding through several full rotations? If one axle has a bind at some point, this could be a problem. I suspect your No. 1 driver (the forwardmost one) is either slightly cocked on the axle or slightly out-of-true which may be causing the minor vertical movement observed in your video.

BTW, your video is an excellent tool for troubleshooting drive train problems! Nicely done! I hope we help you track the problem down once and for all.

This is for Loren Perry.
You stated that there is a misalignment between drivers 2 & 3. So that I am on the same sheet of music you are saying, as I do, that #1 driver is at the front of the engine. You stated a slight vertical movement in the #1 driver axle. I did not notice much there but the second one bounces way too much. I think.

I really need to talk to someone to understand the quartering on this one geared set of drivers.

carl

[carlso]

Lore,

thanks for the reply. I appreciate your close scrutiny of the thing in operation. The reason the #1 driver lifted was because of me as I was touching my voltmeter to the motor leads and pressed down on the tail end of the chassis too hard.

I am going to run the motor as Max has suggested and check it and gearhead out as I don't think it should make the noise that it is.

After doing that I shall put the motor and worm back in and try again. As I said earlier, I have pushed the chassis across my work table on a cardboard mat and it hesitates every so often and with a little pressure it will move on.

Personally, I think the problem is in the #2 axle. Glad you like the vid idea, nothing shows up flaws like a photo or video.

Carl

[victor miranda]

Ah I do so dislike fixing issues without having the loco in my hands...
and I am not going to have the time to offer to fix it .... anytime soon.

the thing to look for is when the hesitation happens

If it hesitates at about at the halfway fron 12-3-6-9 oclock
-->and<-- it is at a consistant place then you can assume a quartering issue.

the design on the gs-4 is good for avoiding quartering problems
more or less if the quartering is close it will run well.

what I see and hear make me think a gear related problem.
look for flash, a crack, a small rock, the gears rubbing something.
sometimes drivers find ways to rub frames and on the odd occasion
a spot of glue on an axle bearing will give you fits (I will never admit why I know this)

when one gets a loco into a few smph anything and everything causes problems

again, I think the worm/bearings/shaft are the final problem you will have to solve.
cause the noise I hear sounds 'wormy'

victor


 

[carlso]

Thanks Victor,

The drive line is next to study, again, as I tested the motor and gearhead as Max suggested and it performed as he thought it should. It actually pulled less than .5a  @6v, so I am pleased. After removing and reinstalling the gearhead it is much quieter. So, to bed with that piece of the puzzle. I am going to re think the way I installed the motor and see if another way would be better, without using the upper KATO frame halves. Not sure about that, however.

Carl

[mmagliaro]

EDITED FOR TYPOS

Carl,
Your units are probably not right there.  It can't be 0.5A.  I will guess that it is 5 mA (or .005 A)
Assuming that's correct, then yes, it sounds like the motor is perfect. 
And I am glad you got it to run quieter by unscrewing the gearhead and screwing it back on.

Sometimes that does fix it.  A very small amount of extra turning on it one way or the other can make
a big difference in the noise.  Those gears are amazingly precise, and they do not tolerate misalignment well at all.

On to the chassis.  If you can feel a tight spot as you roll the free-running chassis down the track,
that's the place to focus.  Good luck. 

[Doug G.]

As you roll it down the track and feel the binding, stop it there and use a tweezers or other little poky thing and try to move each driving/side rod. They should all be fairly easy to move, even on a loco that uses the rods to drive the drivers. If you feel any stiffness, that's the suspicious place.

It could also be a bearing tight in one spot or two.

Doug

[carlso]

Max,

You are 100% correct, I had my head up my "you know what" and did not have on my transparent belt to see out. The reading I posted was wrong. I just tested it with my handy dandy Craftsman meter and I get .0021 amp or 2ma, so all is cool.

Doug and Max - I have pushed the chassis around many times now and stopped when it hit a bind and have not been able to see any conflicts between rods, cranks, or vale gear. I am thoroughly convinced the problem lies under the driver retaining plate with one or maybe two of the drivers. I will make that my second area to study.

Now, I have removed the original KATO upper halve frames and set the motor on top of a plastic shim  to raise it to the exact level as the worm shaft. I held it together with my hand and found that it runs well. The way I had the motor mounted must have been too much of an angle for the ball/cup connection. I shall pursue that more.

One thing I discovered, and it sort of supports Victor's worm issue, is that the top gear in the chassis gear tower is extremely loose and wobbly in the gear tower. I think it may be causing some vibration and "could" possibly be the"hang up" that happens if the gear gets out of alignment with the worm. Excuse me for not knowing what that gear would be called, I guess I better learn because I may need to find one or maybe an entire set for the gear tower. I doubt it if KATO has any, perhaps NWSL would have something that would work.

I shall be busy testing this da**ed thing, heck KATO may build an updated version before I ever figure it out. I will get it right or when I croak my wife will place the loco in my casket for me to work on.

Carl

[nscaleSPF2]

I will get it right or when I croak my wife will place the loco in my casket for me to work on.

I like your attitude, Carl.  The world needs more people like us.

I can't offer any specific suggestions, since I don't own one of these locomotives, but I can offer the following approach, so take it for what it's worth.

Now that Peteski has shown you how to isolate the general problem area, you may want to try to remove one part at a time from the mechanism.  Like, LH connecting rod, RH connecting rod, LH side rod, RH side rod, #2 driver, #4 driver, etc.  Each time you remove one part, check for the binding again.  Do this carefully, in a systematic manner.  Take notes, this usually helps to reduce confusion later. 

Eventually, you should be able to isolate the exact cause of the problem, hopefully before you croak, rather than after.

Good Luck
Regards

[carlso]

Thanks Jim, I shall conquer,I think, or else.

Here is my latest 9 minute video with my new efforts and I think some positive results. I may be wishful thinking so any comments are requested......................

/>

Thanks for your time.

Carl

[mmagliaro]

WAY better.
I did not realize that the motor shaft was not in a straight line with the worm shaft.  Even with a NWSL coupling,
it is pretty important to avoid any angle there if you really want smooth low-end running.  It looks like you
have solved that.

I calculated your 1 volt speed, assuming 79" drivers, to be 2.8 mph.

Something to check on those two bearing blocks that the worm is riding in.  You showed a view from the top,
where one can clearly see one of the bearings wobbling around, but the other one does not move at all.
Make sure both of those bearing blocks have a smidge of play in them.  You should be able to wiggle them
with tweezers back and forth, just a little bit.  They should freely drop into the frame slots without have to
force them.  If that other bearing is tight in the frame, you might try lifting the lock out and rotating it 90 degrees
to see if a little tolerance variation there will allow it to wiggle.

The hanging up you show now does not at all look like quartering issues.  It stops and starts at a variety of places
around the rotation.   Now I think it is down to the worm gear meshing with the idler under it.  That other bearing
I mentioned is something to check.

I don't know if you did this, but an important trick to getting that motor positioned perfectly
in there is to hold it in your hands, running at a modest voltage, say, 3v, and with an ammeter in the circuit.
You do this without the motor being glued in place, and you have to kind of hold it squeezed in place with your
thumb and forefinger, or a little clamp if you can manage to get one on there.  The key here is
to run it and gently nudge the motor left/right, up/down and experiment until you get the absolute lowest current
reading that you can, and then tack it in place with a drop or two of ACC.  Once that hardens, test it again to make sure
you didn't lose the spot.

Alternatively, it looks like you have enough room there to actually make a bracket of some sort that you could screw into the
frame.  That takes more work, but it's nice to be able to remove the motor.  I'm one to talk as I often just
glue motors in, but that's usually because I am not willing to give up a bigger motor in exchange for the bracket space.
But you might have room in this case.

Anyway... this fine-tuning of the motor alignment with an ammeter is really critical to getting that final edge
on the low-speed running.  You may well find that .003" or .005" difference in how high that motor is shimmed underneath
could make a big difference in how slow this engine can run.

-
In any event, I admire your perseverance, and you have made a huge improvement.  It would probably make most people very happy with it just the way it is.

[Loren Perry]

Good work, Carlso! A vast improvement. And I'll bet you'll get even better slow-speed performance when it's on the track pulling a load.

While watching the rig operate, I picked up a slight unwanted vertical movement (again!) but this time in the black plastic frame that supports the expansion link (the oddly shaped part that rocks back and forth when the long eccentric rod connected to the driver's return crank moves it) and the movement is timed to the driver's rotation. I don't think it's affecting your drive train in any noticeable way, but it still caught my eye. I'm trying to isolate each and every possible contributor to your engine's roughness - which has all been eliminated as far as I can see.

None of your drivers have any visible vertical movement any more and the surging I noted earlier has disappeared. Your engine is approaching the level of running like a Swiss watch except for the occasional stopping at less than 1 volt.

By the way, the disk-shaped gear that the worm rides on is called the "worm gear". The worm on the motor's output shaft is called just that - the worm. Downstream of the worm gear will be found "idler gears" which engage the gears on the driver axles. Just FYI.