Kato ES44AC - Faulty Motor?

[peteski]

Dwight,  to be honest I'm not very surprised with the outcome. The whole scenario made no sense from the start.

I have been a troubleshooter for my entire professional life.  First, I troubleshoot electronic hardware (to the component level), then I troubleshoot software issues. At this point I troubleshooting problems with computer storage arrays and their connectivity to servers through various types of connections.

I have had my share of NPFs (No Problem Found) resolutions where something was not functioning correctly but then it just started working.  Well, we both know that there really is no such thing as spontaneous fix.  What it basically mean that no enough time, resources, and historical data were available to conduct thorough enough investigation to explain the instantaneous "fix".

I'm curious how that motor would behave if connected again to its original decoder.  But please do not use the MTL coupler spring as the brush spring.  When you compare the springs you will see that the MTL spring is shorter, has smaller diameter, and is a lot softer than the Kato brush spring.  It does not apply sufficient pressure to the brush, and as the brushes wear down, the pressure will be even lower.  As I see it, the best option is to relegate this motor to the spare parts bin and use the new motor you bought from Kato.

[Dwight in Toronto]

Thanks Peter - I fully concur with everything you said.  I have indeed relegated the motor to the parts bin.  And since I lost one of the Kato brushes, and the hobby shop had no spares, I’m just leaving the MT spring in there … no biggie.  Unfortunately, still no new motor delivery.

Now, regarding the decoder.  Again, just for something to do, I alligator-clipped the standalone decoder between a spare standalone motor and the LokProgrammer.  Lights, sound & motor were working. 

I downloaded a different ES44 sound project from ESU, hit the “sound” load button, and received a “Error writing decoder data’.  I tried a different sound file and got the same result.

LokProgrammer reads the decoder fine.  I’ve updated the firmware several times to no avail.   In ‘Drivers Cab’ the front & rear led’s work and the motor ramps up & down, but the sound file won’t write.  When trying to write, the yellow LokProgrammer led turns on & off at regular intervals for about a minute, and then gives the error message. 

I described this to ESU in an email - they said to send it back.  There’s another thread above discussing how to return decoders to ESU, and I’ve summarized my experience there, fwiw. 

So, now I’m awaiting a Kato motor AND an ESU decoder.  In the end, it will be nice to start over with all-new components.

[peteski]

Now, regarding the decoder.  Again, just for something to do, I alligator-clipped the standalone decoder between a spare standalone motor and the LokProgrammer.  Lights, sound & motor were working. 

The reason I asked to hook up the originally misbehaving, but now apparently working correctly motor (not some spare motor) to the original decoder was because I was curious if that motor would now also work correctly with that decoder.  It is now a moot point - no big deal.

What we also see is that the more complex the decoders become, the more prone they are to suffering failures.  This just wasn't a lucky install for you.

[Dwight in Toronto]

I have another question -  if you alligator-clip jumpers from your multimeter to the motor outputs on a decoder, will you see varying voltages on the DC scale as you increase & decrease speed steps on the dcc throttle?  In other words, can you substitute your multimeter in place of a test motor?

[peteski]

I have another question -  if you alligator-clip jumpers from your multimeter to the motor outputs on a decoder, will you see varying voltages on the DC scale as you increase & decrease speed steps on the dcc throttle?  In other words, can you substitute your multimeter in place of a test motor?

That is a tricky question. It depends on the meter itself.  Most will give you some reading, but it might not be close to being accurate.  While the polarity of the motor output does not change (like the DCC track signal), it is a PWM signal.  The actual voltage is a series of short pulses between 0V and 12V, where the percentage of when each pulse is at 12V (the Pulse Width) changes with the speed step.  Some multimeters might have some sort of averaging or smoothing circuit built in. Those will show more accurate voltage than meters which sample the voltage from the meter's leads without any averaging.

You could add a low value capacitor in parallel with the meter's leads and that would do the averaging for you. If you have a 0.1µF capacitor handy, that would likely work.  Do not use typical tantalum caps with much larger capacitance used in keep-alive circuits.  A tiny capacitor often present on model locomotive light boards would be perfect for the task, or a capacitor wired across a motor as a part of Radio Frequency Interference (RFI) suppression circuit (often seen in more recent model locos made in China) would also work.

[Dwight in Toronto]

Pete - it’s admirable that you so readily share your expertise, ideas, learnings and inquisitiveness here on TRW (and the ESU io.group).  Also, taking the time to compose thoughtful, comprehensible responses is equally noteworthy.  Good on ya mate - it’s appreciated!

[peteski]

Thanks Dwight!  I enjoy sharing my knowledge to help out others.

[Dwight in Toronto]

My Kato motor and revised GS4 drivers showed up today (17 day delivery).

I’ll put the motor in place and reassemble the loco, but now I’m awaiting the replacement 58741 decoder from ESU.  I guess I’ll tackle the GS4 driver upgrade in the meantime.

[Dwight in Toronto]

It’s been two months since I posted anything to this topic, so, apologies to those who may have been following along.

Long story short, and to tie up loose ends, I ended up installing both a new Kato motor, and a new 58741 decoder, within that Kato ES44AC chassis.  Successfully loaded an appropriate ESU sound project, and resumed normal functioning (on its own, and in consist mode) without incident.

By my reckoning, the purchase/repair/maintenance cost for this one locomotive breaks down as follows (Cdn $, taxes + shipping etc where appropriate):
- original purchase cost (Kato Kobo factory dcc/sound) - $480
- ESU 58741 decoder (to replace crappy Kobo ‘Select’ decoder) - $150
- replacement Kato motor - $50
- 2nd replacement 58741 decoder - $150
- estimated Total - $830

And now, here’s the icing on this slice of cake … I’ve got an Atlas GP38 with an ESU 58751 drop-in board, one year since the install, exhibiting the very same symptoms:
- motor function dies while in consist, but sound & lights work
- finger-flick the motor and it sporadically, erratically dithers, irrespective of speed step setting
- motor & drivetrain operate fine on DC power
- decoder operates a stand-alone test motor on DCC
- put everything back together, and the Atlas loco again runs & sounds fine on DCC

At least, until the next time.  I have gone through the above procedure four times within the last 10 weeks or so.  I’ve posted the puzzle on the ESU groups.io forum, and Peteski is once again in the loop.  I’ll report back if/when anything verifiable arises.

Oft times I wonder why I put up with the cost and frustrations of this hobby!

[peteski]

It’s been two months since I posted anything to this topic, so, apologies to those who may have been following along.

Long story short, and to tie up loose ends, I ended up installing both a new Kato motor, and a new 58741 decoder, within that Kato ES44AC chassis.  Successfully loaded an appropriate ESU sound project, and resumed normal functioning (on its own, and in consist mode) without incident.

By my reckoning, the purchase/repair/maintenance cost for this one locomotive breaks down as follows (Cdn $, taxes + shipping etc where appropriate):
- original purchase cost (Kato Kobo factory dcc/sound) - $480
- ESU 58741 decoder (to replace crappy Kobo ‘Select’ decoder) - $150
- replacement Kato motor - $50
- 2nd replacement 58741 decoder - $150
- estimated Total - $830

And now, here’s the icing on this slice of cake … I’ve got an Atlas GP38 with an ESU 58751 drop-in board, one year since the install, exhibiting the very same symptoms:
- motor function dies while in consist, but sound & lights work
- finger-flick the motor and it sporadically, erratically dithers, irrespective of speed step setting
- motor & drivetrain operate fine on DC power
- decoder operates a stand-alone test motor on DCC
- put everything back together, and the Atlas loco again runs & sounds fine on DCC

At least, until the next time.  I have gone through the above procedure four times within the last 10 weeks or so.  I’ve posted the puzzle on the ESU groups.io forum, and Peteski is once again in the loop.  I’ll report back if/when anything verifiable arises.

Oft times I wonder why I put up with the cost and frustrations of this hobby!

Of course those are Canadian Dollar prices (for some of US-based modelers thinking "wow, that was expensive!"
I'll be watching the other thread to see what transpires.

[Dwight in Toronto]

Lol - well, I gotta tell ya Pete - it’s bleedin’ expensive in Cdn $ as well!