Kato ES44AC - Faulty Motor?

[Dwight in Toronto]

I took the questionable 58741 and a spare motor, soldered the motor leads to the decoder paddles, then connected alligator clip jumpers between the track power pads and my LokProgrammer.

The programmer read the decoder specifications and scanned/loaded the sound project with no problem. 

The ‘Drivers Cab’ feature worked perfectly - lights, sound, and full motor control.

This should be proof positive that the decoder and file are not the problem, correct?  That being said, I have no idea why this same decoder was completely nonfunctional when I put it in that AC4400 loco yesterday. 

[peteski]

I took the questionable 58741 and a spare motor, soldered the motor leads to the decoder paddles, then connected alligator clip jumpers between the track power pads and my LokProgrammer.

The programmer read the decoder specifications and scanned/loaded the sound project with no problem. 

The ‘Drivers Cab’ feature worked perfectly - lights, sound, and full motor control.

Well ok, so if you still have this set up, replace the working spare motor with the misbehaving motor (also soldering it to the decoder) and see how that motor behaves on LokProgrammer's Driver's Cab.

This should be proof positive that the decoder and file are not the problem, correct?  That being said, I have no idea why this same decoder was completely nonfunctional when I put it in that AC4400 loco yesterday.

Knowing how iffy the connections between the chassis frames and decoder are, fully nonfunctional "dead" decoder (which works correctly clipped to LokProgrammer with Driver's Cab) most likely indicates that one or both of decoder's terminals are not contacting the Kato chassis (or if the power is delivered through trucks, maybe there are issues with trucks or truck contacts).

And I also agree that the Kato motor contacts which attach to the decoder are not very solid. You could try to bend the "fiingers" slightly to get a tighter fit. Soldering the contacts to the decoder would be a most reliable connection.

[Dwight in Toronto]

Hey Pete.  I moved that last setup with the spare motor (that worked ok on the LokProgrammer yesterday) back onto my NCE PowerCab test loop, and used alligator clips to connect track power to the decoder.   The spare motor spun up and down just as it did on the LokProgrammer. 

I then removed the Kato motor from the chassis, and (after putting the proper brush spring back in place) connected the motor leads to the decoder motor pads with alligator clips, essentially duplicating the setup that had worked moments earlier with the spare motor.

Even at speed step 28, the motor needed a finger flick to get going, and when it did, it was very slow, weak, and with no torque whatsoever.  I then applied plain ol’ DC again, and the motor works perfectly.

Fwiw, I made a hasty video just so that you could see it firsthand.  Apologies for the crappy quality, but we’re just trying to convey findings here - I’m NOT trying to make a Spielberg production!

/>

[peteski]

Dwight I revisited your initial post.

12 - tried a TCS dcc decoder - NCE ‘program track’ feature worked normally - lights worked ok, but no motor response whatsoever

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.


No, (N scale) MTL coupler spring is *NOT* a good substitute for Kato brush spring.  It is way too soft to produce sufficient force on the brush. It is also smaller diameter.  But since you now found and reinstalled the original spring (yes lost parts suddenly can appear in places we already looked at), that is a moot point. 

So this motor also does not work correctly driven by a TCS decoder?

You also mentioned that another identical ESU decoder worked fine with this motor?

And all along the motor works fine using DC throttle.  Does the motor have good torque running under DC power?

This whole thing makes no sense. It just doesn't add up.  It is almost like the scenario was changing as the testing was progressing.

I agree that at this point the motor seems to be the culprit. But the other DC motor you used for testing the decoder was not a Kato motor. I wonder if that makes a difference.  It will be interesting to see what happens when you connect the brand new Kato motor to that decoder.

I'm wondering if you are just doing all this to see how long it will take for my head to explode?   

[Dwight in Toronto]

Hey Pete - sorry for subjecting you to all of these cranial gymnastics!  But thanks for sticking with me on this.

The ‘spare’ motor that ramped up & down and appeared to work fine was a small Dapol unit that I had removed from my O60 Terrier a few months back (I had replaced it with a Tramfabriek coreless motor). 

Like you, I was left thinking that perhaps this smaller motor appeared to work better because it had less starting inertia or something, so it may not have been a good apples-to-apples comparison.  In other words, maybe the ‘questionable decoder’ IS faulty after all, and the Kato motor represents more of a challenge because of the mass of the dual flywheels or something?

So, I tried that 58741 on two additional, lower-quality, much older motors, thinking that maybe they will put more load on the decoder, or be more demanding somehow.  Tests done using track power from the NCE PowerCab.

First, on a ca: 1972-ish motor salvaged years ago from my very first N scale loco (an Aurora Postage Stamp Trains F unit).  The drive shaft needed a finger-flick to get the motor going, after which it ramped up & down, albeit somewhat crudely proportional to speed steps.  Second, I put it on a late-70’s Graham-Farish motor, and pretty much got the same result.  It’s odd that the Kato motor doesn’t behave anything like any of the others that I’ve tested. 

With all the motors, when left to run at a fixed speed step, I occasionally hear quick, short variations in sound pitch.  So, I don’t know … maybe the ESU motor drive circuitry has gone a bit wonky, and just doesn’t have the “ooomph” to spin up that Kato motor? 

I called Kato yesterday saying that I had ordered a few things online but had never received any confirmation.  They said that they’ve had trouble generating auto confirmations, but if my store checkout included an order number (which it did), then the procedure was successful, and I could expect delivery sometime next week.  However, my gut feel is that a new motor will not resolve anything.

In the meantime, I’ll keep plugging away.

[Dwight in Toronto]

Tried using the LokProgrammer (instead of the NCE) with jumpers between the Kato motor tabs and the decoder motor pads.  No change - flicking the flywheels would instigate a weak, half-hearted, short-lived attempt to spin the rotor, even at speed step 28.

It occurred to me that, when I did the speed matching between the three Kato locomotives that I use in this consist, I had set all of the cv5 values to 150.  So, while I had things hooked up to the LokProgrammer I put cv5 up to the max 255.  Unfortunately, Drivers Cab result was same old-same old.

I’m running out of things to try.  I could go back to trying the ‘questionable decoder’ in one of my other Kato locos, but every time I remove a shell or a decoder there’s a risk of some small part flying off, or a speaker wire coming loose etc.  Maybe it’s best to just wait for the new motor (and perhaps a new, spare decoder) to arrive before messing around any further.

[peteski]

But wait.  Didn't you mention that you installed the questionable decoder in another Kato loco (with a Kato motor) in it, and the loco ran as expected? I'm really starting to get confused by all that was tried.

As for needing to flick any motor to start it, you should not have to do that.  The decoder's motor driver should have more than ample power to do just that all by itself. Remember between the full voltage PWM driver and BEMF, you should be able to have the motor running steadily and with decent amount of torque even at very low speeds.  If the motors you tried have not been used for long time then the commutator or brushes in any DC motor seem to get "lazy", and require rather high starting voltage.  It is probably some sort of resistive film (oxidation?) that builds up on the commutator.  But running them at full speed for a minute or so cleans things up, so after that their low speed operation is restored.

I'm also still puzzled that a TCS decoder also had a problem with the motor.  Or was that information incorrect?  Like I said, there have been so many combinations of  things tried that I'm sort of getting lost.  Maybe it would be good to start this diagnostic from the beginning?  Or better yet, just wait for the replacement motor to arrive and then see how things work out.

I just thought of another thing to try:  put the decoder on LokProgrammer and disable BEMF, then try running the motors.  I'm curious as to what difference that will make.

I also recently ordered some Kato parts and never received an email confirmation. After few days I called Kato and was told the same thing you were told. So they haven't fixed the email problem yet.

[Dwight in Toronto]

I know, sorry for all the permutations and combinations … I’ve started to confuse myself as well, and I’ve been having self-doubts and second thoughts on tests done just a few days ago. 

For instance, I’m sure that suspect decoder ran ok in a different loco the first time I tried it, but then the second time it didn’t.  And maybe that was because of the crappy fit between the board and the frame, but then I remember shimming for that and verifying continuity. 

And just this morning I tried both an older TCS and Digitrax dcc decoder on all those motors, and none of them generated a motor response (even though they properly went through the NCE programming routine, and headlights worked). 

So, yeah, I’m with ya regarding confusion.  It IS tempting to start over, but I think I’ll await the new motor.

Apologies for turning this into a nuisance - I’ll report back when the motor arrives. 

[peteski]

No harm done Dwight - I agree  about waiting for the new motor. 

[Dwight in Toronto]

Just checked my mailbox and, unfortunately, the new Kato motor has not yet arrived. 

Looking for something to do, I revisited the old motor, just for the heckuvit.  I never bothered re-installing it in the loco, and it’s not connected to the decoder, so it’s just the bare-bones motor.

I again attached alligator clips between the motor leads and my MRC transformer.  I was surprised when there was no response at all on straight DC … this had never been a problem before now.

I checked motor tab continuity with my multimeter - nothing.  Normally, probes on the motor terminals sound the continuity ‘beep’.

I removed the brass retaining caps, springs, and brushes, and inserted the meter probes into the empty brush cylinders - there was continuity.  I replaced one brush and again confirmed continuity; ditto after having replaced the second brush. 

I replaced one spring and retaining cap - no continuity.  Removed that cap & spring, and installed the opposite ones - no continuity.  Replaced both sets of springs & caps - no continuity.

I repeated the exercise several times, always with the same results. 

I can’t fathom why everything seems proper with just the brushes in place, but the brass springs and caps are non-conductive.  Makes no sense to me - I am totally stumped.

I then used a gear-pulling tool that I had purchased a few months ago (from Tramfabriek) to easily remove one of the flywheels.  Again, no particular reason, other than trying out the tool for the first time!

Anyway, just wondering if anyone has anything to share about the puzzling spring/brush/commutator situation?

[peteski]

Wow Dwight, you sure come up with some weird things! 

When I have a dead motor, I  connect a DC power supply to it and if it does not spin, I try spinning the shaft or flywheels by hand. That often kick-starts the motor (if it has a dirty/oxidized commutator or a dead winding). I'm also assuming that you have the correct brush springs installed in the motor (not the soft MTL coupler spring).

If you have the brushes fully assembled in the motor, connect DC power (at 12V), and give the armature (flywheel or motor shaft) a spin or two.  Still dead?

Your problem as described doesn't add up.  Using your multimeter you can easily check conductivity of the springs (watch out they don't sproing out when being held by the probes), and you can check conductivity of the brushes (both should be close to zero ohms).  You can do the same to the brass clip that retains the brush (but I just don't see how that would be non conductive).  But don't use continuity test setting - use the low ohm range.

You mentioned that you put the probes in the empty brush cylinders, and there was continuity.  That means the probes must have been touching the commutator segments, and that the windings connected to those segments are ok.  But I suspect that the pointy probes also put a small dimples in the soft copper commutator segments which will affect the longevity of the brushes if you ever get this motor revived.

How about this:  put the brushes back in the cylinders, then put the meter's probes in the cylinder deep enough to put pressure on the brushes.  Is there continuity? 

Report your findings. 

[Dwight in Toronto]

Ahhhh, Peter … I know, man - crazy stuff happn’in here!

At the first sign of a dead motor, with DC power applied, I tried spinning the flywheels to no avail.

The meter probes were indeed lightly touching the commutator when I confirmed continuity. 
And yes, with both brushes in place, the probes verify continuity.

At the 200 ohm setting (which also gives the audible beep), the springs and caps are pretty much zero ohms, and the brushes are about 3 ohms.  And, ummmm, yeah - while testing one of the springs, it did the Houdini thing and vanished.  So, for the here-&-now, an MT coupler spring has again been recruited!  I know you don’t approve, but it DID work in a pinch last time, and, well, here we are again.

Finally, I put JUST the brushes back in, then clipped my DC leads to the multimeter probes, and touched the probes to the brushes, and the motor runs fine.

Reassembled the spring & cap on one side only, and no continuity.  This makes no sense at all - that such highly conductive materials, under spring-induced contact, are behaving as if they are totally insulated.

While reassembling the spring & cap on the other side, the MT spring sprung away, so I’m calling it a night.  This is really gonna bug me though!

[MK]

Bugging you?  Me too!  Have been following since Day 1 with interest. 

Could there be some kind of varnish on the underside of the caps?  Maybe ohm the top and bottom of the cap?  Then afterwards the top of the cap and the spring with the other probe at the end of the spring.

[peteski]

Yes, this problem is so illogical, with only 3 components in play, that the solution should be simple. Actually the brushes were eliminated (because when electrified probes were touched to the installed brushes the motor ran.

So it is just the spring and the caps.  How about eliminating the caps?  Cut thin strips of aluminum tape. Install brushes and springs in the tubes, then put the strip of aluminum foil over the spring as you are snapping the caps on.  The foil should be pinched between the top of the spring and the dimple in the cap.  Then apply power to the foil strips.   

What makes this even stranger is that this exact motor ran on DC (and poorly on DCC) few days ago!

[Dwight in Toronto]

MK - there was no resistance at all between one side of the cover plates and the other.  Regardless, I used a thin, fine-grit sanding stick to lightly buff up the inside surface of the retaining plate that contacts the spring.

Pete - I tried aluminum foil as suggested.  There WAS one super-brief ‘bleep’ of continuity, so an encouraging sign finally.  I could not get it to reoccur, so tried a triple layer of aluminum foil, but no further luck.

Popped over to my LHS (George’s Trains) with the motor in tow, to consult with one of the gents on staff who has decades of sales/service/maintenance experience.  Explained the entire escapade.  He did what I’ve been doing - removed & examined retaining covers, springs and brushes, put everything back together, and jumpered DC power to the motor tabs.  The darn thing ran! 

He had no answers, but his best guess was that maybe the ‘scalloped’ end of one or both brushes was a bit skewed in the port, and therefore not in full-on contact with the curvature of the commutator.  Fwiw, whenever I inserted the brushes, I did my best to try to get that alignment right, but who knows … perhaps being off by even the tiniest amount is enough to make all the difference.

Anyway, got home just now and verified that the motor does indeed run on DC again!