Kato ES44AC - Faulty Motor?

[Dwight in Toronto]

ES44AC was leading a consist with a Kato AC4400 and a Kato SD40-2, all three equipped with ESU 58741 “Kato USA” drop-in sound decoders.  The lead ES44AC lost sound and power, and was  being pushed by the other two. 

Exploratory analysis:
1 - rebooted NCE PowerCab - lights & sounds functional; no movement
2 - put loco on LokProgrammer - cancelled the ‘consist’ setting - lights, sound, movement restored
3 - back on NCE/test loop - ran 3 feet and stopped; lights & sounds functional
4 - removed shell - manually moved flywheels, no binding in drivetrain
5 - back on test loop - speed step 4, manually move flywheel - motor exhibits jittery rotation - rpm remains slow, does not change even at speed step 28, and frequently cuts out
6 - remove decoder - motor runs perfectly on DC
7 - decoder checks out on decoder tester - motor speed ramps up & down
8 - put decoder back in loco - see 5 above
9 - swap decoders (Kato AC4400 decoder put on ES44 mechanism, and vice versa) - both locos perform properly
10 - put decoders back in original locos - see 5 above
11 - complete tear down of ES44 mechanism - confirmed drivetrain & trucks are perfect; decoder and motor tested in isolation on workbench (jumpers clipped to rails on NCE test loop at on end, and to track power pads on decoder at other end) - see 5 above
12 - tried a TCS dcc decoder - NCE ‘program track’ feature worked normally - lights worked ok, but no motor response whatsoever
13 - brushes removed from motor - they are clean/fresh/new & over 1/8” long (3/16” maybe?)
14 - reinstall brush assemblies (*a Microtrains coupler spring can substitute for a lost Kato brush spring … who knew!); motor runs perfect on DC; tried ESU decoder one more time … see 5 above.

I am stumped. The ESU decoder works fine in a different locomotive.  The ES44 works fine with an identical decoder borrowed from another loco.   Yet the two will not play well together.  I’m leaning towards something being out of sorts with the motor, but most of the evidence suggests otherwise.

I expect folks to say “just swap the decoders and download the respective sound files and all should be good again”.  But I just had to ask if anyone might have any inkling at all about what might possibly be going on here?

[Scottl]

With my NCE system, I find occasionally that individual consists stop working.  My solution is to clear the offending consist and set it again.  It always seems to work.  I have experienced this with ESU and  TCS decoders in Kato and other 6-axle locos.

[peteski]

To me your analysis seems to indicate a problem with the decoder.  Yes, I know that the problem appears to be a specific combination of the particular motor with particular decoder. but when something goes "squirrely", the decoder is usually the culprit.  As you know the motor is a very simple electro-mechanical device.  Brushes, commutator, armature windings, magnets, and  the housing (with bearings) that keeps everything aligned.  There isn't much there that could cause such oddball behavior.

The only possible motor problem I could think of could be that the gaps in the commutator segments are filled with powered brush material.  But if that was the case, the problem should also be present when it is controlled by another decoder, and  the motor should consume more than expected current running on DC.  Cleaning those gaps is tricky since the commutator is not readily accessible.  I have been known to carve an opening  in the Kato motor's plastic end cap to access the brushes.  Another way (although awkward) is to remove a brush and clean the segment through the brush holder "tube".  I take a flat wooden toothpick and sharpen the pointy end to a thin rectangular cross section (narrow enough to slide in the the gaps in the commutator. Then I line up a gap with the center of the brush "tube"  and use my "cleaning tool" to scrape back and  forth inside each gap.   Since the access opening is so small I do this by feel. Remember never to use any metallic item to clean the gaps as the copper commutator segments are very soft and can easily be scratched or deformed.

I suppose the motor can have one of the winding ends disconnected from the commutator segment, but they the motor would likely have a dead spot and unable to start running randomly from time to time (when the dead segment ends up stopping squarely under a brush).  I had that happen in one of the motors I troubleshoot.

However after rereading your troubleshooting description couple more times, I'm not sure if I fully understand step  4. 
Are you saying that on speed step 4 dialed up on the throttle the loco does not move?  That is how I understand your writeup.
Then if you manually rotate the armature or flywheel slightly, it starts running but jittery?
What if instead of giving the motor a jump-start with your finger, (while it is still not running) you listen if the armature makes a noise (if you have good hearing), or do you feel any vibration wen you touch (but not move) a flywheel or armature?
The other thing is, if the loco does not move on speed step 4, what if you start cranking the speed higher (all the way to max speed).  will the loco eventually take off, or it stays stalled even ad max speed?

[Dwight in Toronto]

Hi Pete - I was hoping you’d offer some pointers.

To clarify item 5 - when the frame was on the track, I set the PowerCab to speed step 4 so that the start voltage was sufficiently high, but nowhere near max speed.  There was no discernible sound, such as one might hear when the motor is on the verge of starting to turn. 

Interestingly, when I then lightly gripped the mechanism by the sides to give it a helping push/slide along the rails, that’s when the jittery/dithering motion occurred.  I could then get that to repeat by gently swiping the flywheel.

And yes, I did try advancing the throttle all the way to speed step 28.  There was no motor response until I manually turned a flywheel, at which point the very same slow, jittery dithering ensued.  Also, just to be clear, the motor would not continue running in the slow/rough manner - it would constantly stall out after a few laboured rotations.

I fully agree that the Kato motor is a simple, robust, reliable little mechanism and should be least suspect, and that the decoder is usually the faulty component.  What made me lean towards the motor was when the TCS drop-in dcc decoder programmed ok, worked the lighting functions, but no motor response at all.  But then we have the fact that the substitute 58741 worked fine.

I’ve puzzled and puzzled until my puzzler got sore!  Will maybe have another kick at it today, but in the meantime I just ordered a replacement motor from Kato.  It’s not a big expense, and I’ve been wanting to have the revised GS4 drivers on hand anyway so I ordered those too.

Looking forward to any further comments, ideas, suggestions. 

[Ku7u]

Dwight
What have you done, if anything, to configure the motor control parameters in the decoder? Could be that there is a simple misadjustment that causes this misbehavior with this particular motor.

George
Edgewood, WA

[peteski]

What makes me think it is not the motor is the fact that it runs fine in DC (with the original Kato circuit board).  I assume it runs smoothly in all speeds.

But if it doesn't work properly with both brands of decoders I wonder if there is a problem with the motor contacts which snap onto the decoders circuit board.  You mentioned that the ESU decoder tested ok on the decoder tester. How did you hook the motor contacts to the tester?  Alligator test clips?

There might be another possibility.  Are you absolutely certain that the motor leads and any part of the decoder are not in electrical contact with either chassis half?  If you have enough test leads, take the decoder out and leave it out of the loco and clip the track power pads of the decoder to either chassis halves or even directly to the track, and connect the motor pads to the motor leads, then see if the motor behaves correctly.

[Dwight in Toronto]

Peter, I think you’ve zeroed in on the problem. I was going through some more troubleshooting just now.

I powered the chassis on my workbench by connecting alligator clips to each frame half, with the other ends clipped to my test loop rails.  The loco ran properly, fwd & rev, up & down the speed steps. 

And then it didn’t.  Leaving the PowerCab on 16, I found that gently squeezing the frame sides fired up the motor.  I made a short video:

/>
My suspicion is that the track power pickup pads on the decoder are not making good enough contact where they slide under the frame friction points.  There are 3 power pads on each side of the 58741, but only two on each side slide under this particular frame.  Oddly, the pair of contacts right where I was squeezing are NOT located at one of the friction slots, but I guess there was just enough frame movement to encourage some contact at one of the other spots.

I’m going to try working some thin pvc shims under the decoder to see if that helps.  One of the first things I did at the outset (and neglected to put in my troubleshooting list) was to check continuity by probing pads, frame and copper pickup strips down by the trucks.  Everything checked out, so I began looking elsewhere.  However - eureka! - I just re-checked continuity and sure enough - the left side frame is not touching the left side decoder pads. 

Going to try shimming just now.

[Dwight in Toronto]

My exuberance was short-lived.  A small strip of Kapton tape under the decoder was enough to ensure continuity between frame and decoder pads.  BUT … the loco only ran 12” then stopped. 

It’s reverted to the slow/jittery/dithering that does not change, regardless of speed step setting.  The continuity thing WAS a coincidental problem … it was just not THE problem!  I should have known that it was something else, because even when the motor is acting up, sound and lights remain operational.

George - thanks for chiming in - I have not altered any motor control parameters whatsoever.

Here’s hoping that other possibilities are forthcoming. 

[Bendtracker1]

Dwight, it sounds to me like there is a loss of continuity between the frame and the board.
I had a guy send me basically the same KATO chassis and his was doing pretty much the same thing.

What I found was the chip was a tad bit thinner than the original DC board.
I added some very thin pillows of solder to the board where they make contact with the frame.
First try wasn't quite enough, so I added a bit more solder.  It then press fit snugly into the frames and it ran fine after that.

Check to see how much play the board has when installed, if it wiggles even the slightest, add the solder.

[Dwight in Toronto]

Thanks Allen.  As my post of 2:20 pm explained, what you surmise was indeed a problem.  Rather than add solder to thicken the pad, I stuck some Kapton tape on the bottom of the decoder, which was enough to make contact.

However, as I went on to explain, that was not THE problem - the original troubles persist.

[peteski]

I'm unclear again. The video shows a different test lead hookup than what I recommended. Have you done the test I mentioned earlier (using test leads to connect track power and the motor to the decoder taken out of the loco)?

Also when the motor is acting funny, does the decoder continue to produce sound (diesel, horn, bell), and do the headlights work and can be controlled?

[Dwight in Toronto]

Peter - I did something very similar to your prescribed test as part of yesterday’s troubleshooting (step 11, where I had completely torn down the loco). 

While I had the motor out of the chassis, I slipped the decoder motor contacts over the copper motor leads (just as is done when the motor is in the frame), and then used alligator clips on the decoder track power pads (other ends clipped to the rails).  The motor behaved in the erratic fashion.

But I now see what you are suggesting … I will use another set of alligator clips to connect the decoder motor pads to the Kato motor leads, and NOT rely on those copper side paddles.  Those paddles are soldered quite rigorously to the decoder pads, but I will try this later this evening and report back.

[Dwight in Toronto]

Pete - I distanced the decoder from the loco, and jumped the motor tabs to the decoder motor pads with one set of alligator clips.  A second set of clips connected the decoder “track” pads to the rails.

There was no improvement.  I think I might have detected a very faint, intermittent humming sound, but moving the flywheels did not encourage any motor rotation.

I’ll keep plugging away.

[peteski]

Ok, fair enough.  I just wanted to see how the motor and decoder work totally removed from the chassis. So that motor works fine in DC (at every speed), but gets weird with ESU and TCS decoders?  YOu also mentioned that the same decoder works ok in another Kato locomotive (with a similar motor). That really is a stumper.

Do you know if the loco in question (and its motor) are new or have been in use?  Maybe the commutator is the culprit.  Someone asked if you did any tuning of the motor parameters in the decoder. While that would be worth checking, this is a standard permanent-magnet and iron-armature motor.  It should not behave like that when using standard motor parameters. Plus the decoder works ok in another loco with a similar motor.

If the new motor resolves the problem I would be interested in getting my hands on that motor for analysts.  It got me really curious.

[Dwight in Toronto]

Peter - I bought the ES44 new, maybe 18 months ago.  It was a “Kobo Custom”, so, factory dcc/sound, but I never cared for the sound.  When I put it on LokProgrammer to download something better, I found that it was using a LokSound Select.  LS5 projects are not compatible with LS Select, so I removed that decoder and put it in a Kato RDC, then equipped the ES44 with the 58741.  I haven’t used it much - I’m guessing that it has 15 hours on it, tops.

I have not tweaked any motor parameters.

The plot thickens.  I went to the bother of once again swapping decoders - the “known-to-be-good” 58741 from my AC4400 was put in the ES44, and the “questionable” 58741 was popped into the AC4400. 

The “good” decoder delivered lights & sound in the ES44, but no motor response.

To my surprise, the “questionable” decoder was TOTALLY dead in the AC4400 - no lights, no sound, no motion.  I’m SURE that combination had worked FINE when I had tested it several hours earlier. 

I again put everything as they were - the AC4400 worked perfectly; the ES44 had no motor function.

I’ll keep trying and retrying various things.  I’ll put that 58741 on my (homemade) decoder tester again, and then maybe download a whole new project/sound file and see if that helps. 

Maybe I’ll get an ESU micro or nano to try with the ES44 (can always use one down the road anyways).  I’m disappointed at how those drop-in ESU boards are somewhat of a sloppy, hit-or-miss fit where the track power pads are supposed to contact the Kato frame, and I’d have more confidence in wired/soldered connections.

If the new Kato motor works, I would be happy to send you the old one.  I ordered the motor (and GS4 drivers) on Saturday, but have yet to receive any confirmation that the order has been received/processed.

Oh, funny thing - I was cleaning up after doing some airbrushing late yesterday, and that tiny brush/commutator spring that had sprung off into the netherworld had somehow found it’s way back onto my workbench.  It was in an uncluttered spot, literally right in front of my nose, in a place that I had scoured a dozen times before I decided that any hope of finding it was fruitless.

I’ll keep updating this thread if/as/when new findings arise.