The Perplexing Case Of The Melty Lounge Cars

[Davo Dentetsu]

Yes, all very strange.

 

Running a full set of Suisei/Akatsuki on Friday (as I usually have for quite a while and no issues) when I found the second coach in formation started to tilt over.  Very odd I thought and proceeded to be confused by numerous teardowns without spotting anything causing it.  All of a sudden, we had noticed the lugs on the bogie that keep the coach at the correct height had melted on one side of each bogie.  I was stunned to find that as the only coaches that have light circuits operating are just the standard end coaches, of which this isn't one, so for now I sanded the unaffected side and switched the ends, no tilting issue so all good I thought.
 

 

At this point I separated the sets, thinking it best I don't overload it for the moment.  Next thing I know, the first coach (the one in front of the previously leaning lounge) has now displayed the same symptoms.  Uhoh, now what?  Well... I removed it and... yeah.



Oh bloody hell.

That is the corridor end of the Legato, another bogie dead.  So now I'm kind of keen to get to the bottom of this.

The layout uses wireless controllers which outputs to a DC transformer, the voltage of which has mysteriously not been tested by anyone with a more knowledgable head than mine.  Someone mentioned a pulse something or other, which some things dont like, I dunno.  My suspicion is really on that just chucking out far too high a voltage really.

But club members did test these two coaches thoroughly and found no shorts at all.  They noted that the light in the Legato draws a fair bit, so for some reason that seems to attract the attention rather than the layout hardware itself.

Anyone ever have such issues with the lighting circuits on the sleeper coaches (end light only) before?  Or even some happy electrical basics I should look out for when I test the layout with a meter?

Edited May 15, 2016 by Azumanga Davo

[kvp]

There are two types of pulsed power. On-off pulsing and chopped (on-disconnected). On-off pulsing will generate a braking pulse while in the off cycle heating up the train motors while draining any directly connected storage caps. Chopped pwm is better as the motors are just freewheeling while in the disconnected state, unless there are any directly connected capacitors, in which case they will drain them first every cycle. This is the reason why constant lighting (CL) compatible lighting units must use an isolating diode before any buffer capacitors.

 

I would also test the impedance of the cars under load as conventional bulbs require quite a lot of starting current before they light up, after which their resistance increases and current drops. If you turn off (pulse) the power multiple times a second and the time between two cycles is long enough for the bulbs to turn off, you'll get a high current consumption.

 

What kind of pickup and light system do those cars use?

[mrp]

I'm wondering where the electrical load would be on a coach that doesn't even have lights?

  

Just a random thought, but you haven't started using a new lubricant on the bogies that might be conductive?  Like graphite?  Or somehow accidentally managed to get some type of conductive substance bridging the gap between the wheel and the axle on the side that's supposed to be insulated?

  

Seems to me that in order to melt like that, the wheel itself must be getting hot - and if it's only on one side then it might be the side with the insulating gap between wheel and axle.

[Densha]

I'm far from an expert at electrical components, but if I were you I would keep my trains off the track before you confirmed the voltage. Japanese N gauge trains are rated for maximum 12V and shouldn't be ran at higher voltages. I'm really curious as to how high the voltage was to lead to actual plastic melting. I'm suspecting it to be scarily high.

[kvp]

It's not the voltage, but the current that could cause problems (power, thus heat is actually the product of the voltage and the current). You can have a hundred volts on the tracks and until a spark bridges the isolation through the air at a much larger voltage, nothing would happen if there are no active loads in the section.

 

The cars on the picture seem to have pickups installed. Having a one side isolated wheelset turned around could cause a near short within the bogie, manifesting itself as low resistence path. It's also possible to have a cross car low resistence path through the two bogies. This doesn't show as a short or similar when measuring each bogie individually, but you could detect it on the fully assembled car standing on a piece of unconnected track by measuring the resistence between the rails.

 

Another possibility is a broken joint, partially isolated section or out of synch drivers/boosters (analog/digital) somewhere on the layout, were each pickup bogie could bridge the gap on the same side. This could momentairly send the full layout current through the pickups (either single bogie or through the car). A classic case in real life was the british class 73 which sometimes managed to bridge the 3rd rail gap between two power districts (across its two bogies) and as a result setting itself on fire several times while carrying the load of the two districts across its internal circuits.

[Fenway Park]

I presume your controller is not a feed back unit, as these can severely damage Japanese motors.

[Davo Dentetsu]

Another possibility is a broken joint, partially isolated section or out of synch drivers/boosters (analog/digital) somewhere on the layout, were each pickup bogie could bridge the gap on the same side. This could momentairly send the full layout current through the pickups (either single bogie or through the car). A classic case in real life was the british class 73 which sometimes managed to bridge the 3rd rail gap between two power districts (across its two bogies) and as a result setting itself on fire several times while carrying the load of the two districts across its internal circuits.

Ah, this sounds like a likely culprit.  Recently I have assembled sets in a yard that has a lot of isolators and power feed switches over crossovers to run on the main line.  I promise to take a photo of that particular part of the layout, I suspect it very highly now.  Although there is a lot of good suggestions involving other things like basic voltage and axle insulation (I can at least check that last thing tonight, the rest tomorrow).  I do stress that prior to this, it had not required any servicing and the only real unusual thing was the fact the Legato coach led this time.

[beakaboy]

I had an interesting incident recently on my NCE powercab layout . The layout was using the CP6 bulb system to indicate shorts and everything had been fine except for the odd incident with locos hitting points set wrong due to not paying attention when switching. Suddenly I started to get a glow when 2 sound locos  were running which was strange as I had never had any glow with 3 sound locos running before. Eventually after checking for shorts in points, track,etc, I started to take all rolling stock off track and the glow stopped. The last item I had placed on layout recently had been a second hand Kato coach chassis set up for lighting with wheel wipers and springs as well as brass contact strips under floor, but lights  never fitted. It had a wheelset missing when I got it and I replaced it with a secondhand set. this coach was shorting for some reason and causing the issues. stripped it down and even changed the wheelsets and everything looks okay. Definitely not putting it anywhere near the layout again. Never struck anything like this before!!

I have since changed to a proper circuit breaker.

[kvp]

There are various types of wheelsets:

single side isolated, double side isolated, middle isolated. Using the wrong type for a certain pickup type or installing the single side ones the wrong way could cause a near short (low resistence path). Assembling the cars with the internal strips touching also leads to the same problem. These errors could be detected by placing the car on a piece of track and measuring the resistence between the two rails.

[cteno4]

We had this happen a few times on the jrm layout where a train picked a point and the wheel shorted the point. Wheel got hot enough tom melt the trucks before the transformer breaker tripped. We inserted electronic fuses (thermal ones like used in USB ports) on each throttle so this does not happen anymore! Doremon was melted a long time back and luckily repaired for 500yen by MA! Last just before the breaker was installed was the m250 motor car and 2 months for katousa truck Replacements!

 

I've also wondered how a few cars have gotten their spring pickups fused to the brass pickup plates over they years with no other damage.. They are pretty well welded on and looks just like a spot weld, not corrosion. It's like a large amount of current passed thru the spring. Never figured out quite how that comes about. Nothing else fried. Spring has to be scraped off the contact plate and plate smoothed off and sometimes the spring replaced. I've not kept track well enough though over the years as to motor cars,

 

Cheers

 

Jeff

[beakaboy]

There are various types of wheelsets:

single side isolated, double side isolated, middle isolated. Using the wrong type for a certain pickup type or installing the single side ones the wrong way could cause a near short (low resistence path). Assembling the cars with the internal strips touching also leads to the same problem. These errors could be detected by placing the car on a piece of track and measuring the resistence between the two rails.

Thanks Kvp. I am still a bit of a novice at times with electrics. Will give that a try.

[Ochanomizu]

Hello,

 

I'm struggling to understand the proposed causes of this problem.  Even if you sit the bogie across two power districts there should not be a short circuit.  I have MicroAce Odakyu Romance cars with power couplers and Tomix shinkansen with power couplers that frequently bridge power districts without incident.  They have run on DC and AC.

 

If other consists and users are not having the same problem, I think the problem is within the car.

 

Have you dismantled the car to check if anything is bridging the metal strips that run under the seating parts?  I do not have the model you are referring to but have, in the past, experienced a case where the thin copper strips have moved and shorted.  This does not occur if the car has the heavy coated steel strips that some manufacturers use.

 

I suggest you also strip down an unaffected car to see what might be different.

 

Finally, please give manufacturer and product code for the car.  I would like to consult with my colleagues.  This problem does seem to be elusive.

[Davo Dentetsu]

Alrighty, just had a bit of a meeting with the local electrical gurus and it turns out... you'll like this.  A transformer with no safety break.  Also turns out someone had pointed this out and had remedial action set for it, but was knocked back for being "new" by the powers that be.

 

Isn't that bloody lovely, that could have been massively costly to me!  And not just me, anyone else!

*sits here in an anger filled daze, casually headdesking loudly*

[mrp]

 

I'm struggling to understand the proposed causes of this problem.  Even if you sit the bogie across two power districts there should not be a short circuit.  I have MicroAce Odakyu Romance cars with power couplers and Tomix shinkansen with power couplers that frequently bridge power districts without incident.  They have run on DC and AC.

 

 

  

I'm not sure if it's the same thing that kvp is talking about, but I had a similar problem with version 1 of the PWM system I'm building.

  

If the PWM signal for two adjacent blocks is slightly out of phase - and if the two blocks have a common power ground (possibly via a linked signal ground) - and if the PWM is the braking type - then as the bogie passes over the isolation gap it shorts the power during the time that the PWM signals are out of phase (opposite polarity).

  

In my case it was pretty easy to notice, what with the train motor jerking and interior lights flashing…  And the heat coming off the PWM controller motor driver IC.

  

Edited May 17, 2016 by mrp

[kvp]

Yes, exactly what MRP has drawn above. If one (or both) of the controllers is lacking some form of current limiting then this could seriously damage the bogie. And thanks to the relatively high resistence of the wheelsets, even a transformer with a 3A current limit could remain untripped while still sending a bit less than 3A through the bogie, which is way above its design limits.

[HantuBlauLOL]

Tomix's spring pickup acts like an inductor and it could get hot if too much current gets in.

[cteno4]

Im still puzzled how the springs get fused. Its not like the meltdowns where the wheel shorts out the two rails at the frog and the wheel pickups heats up and melts the plastic around it (so im guessing. here its just the current going thru the wheel and heating it i would imagine.

 

With the fused pickup springs I dont see how the large amount (at most like 1A or so with the plain jane DC transformers we use) of current could come from. in addition these are usually found because the train is running with a truck not rotating freely, not as a derailed car on a point with the plastic meltdown. The only guess i have is that at some point the train motor gets stuck so full current could go thru the circuit thru the car and the spring is the narrowest bit of the current path therefore would heat the most. 

 

any other ideas?

 

should have written all these down over the years to get better ideas! running lots of trains for long periods on the club layout is a good way to run across a lot of oddities. too often i just repair on the fly and go...

 

jeff

[Ochanomizu]

Yes, exactly what MRP has drawn above. If one (or both) of the controllers is lacking some form of current limiting then this could seriously damage the bogie. And thanks to the relatively high resistence of the wheelsets, even a transformer with a 3A current limit could remain untripped while still sending a bit less than 3A through the bogie, which is way above its design limits.

 

Hello,

 

My skill is not such that I can comment on this, so I shall defer to your expert advice.  However, it now raises a concern for me.

 

I am using a single Digitrax PS2012 20A supply to YC52 Cable with 5A breakers built in.  These go to PM42 with each breaker set for 3A.  Including associated yards, any district can have up to 7 consists on it if I fully populate the yards.  At the moment, I rely on the PM42 and the YC52 breakers for safety.  I have purchased several automotive fuse boxes and many 1A fuses.  I was thinking to protect each block with a 1A fuse.  Do you think this is a good idea?

 

Also, am I correct that I should not have a problem so long as I am only using the single power supply to power the whole layout?

 

Thanks in advance.

[kvp]

The spring fusing can happen when the spring tip hits the same spot repeatedly. This creates a very small, but very clean spot where less current is needed to weld. It would be a small and weak weld though. Normally the bogie moving around would make the spring wipe over the pickup plates and that should prohibit this.

 

About the breakers: digital breakers should be fast but the full load (3A) will still go through the shorting equipment and if internal resistence is high enough it could stay at say 2.9A, leaving the breakers untripped.

 

On the other hand automotive fuses are slow, doesn't have auto reset and 1A could be a legal current for a bulb lighted train with sound decoder, even in N scale.

 

The number of transformers doesn't really matter, but all boosters/drivers (digital/analog) should be in synch. With DCC, the booster bus (CDE or Loconet-railsync) should keep everything in synch. The only problems come from using two centrals or feeding a booster from another one's output as the latter causes jitter due to processing delays.

Edited May 18, 2016 by kvp

[inobu]

Either the bulb wattage is too high or the springs contact with the pickup is poor.

 

Inobu   

[inobu]

Hello,

 

My skill is not such that I can comment on this, so I shall defer to your expert advice.  However, it now raises a concern for me.

 

I am using a single Digitrax PS2012 20A supply to YC52 Cable with 5A breakers built in.  These go to PM42 with each breaker set for 3A.  Including associated yards, any district can have up to 7 consists on it if I fully populate the yards.  At the moment, I rely on the PM42 and the YC52 breakers for safety.  I have purchased several automotive fuse boxes and many 1A fuses.  I was thinking to protect each block with a 1A fuse.  Do you think this is a good idea?

 

Also, am I correct that I should not have a problem so long as I am only using the single power supply to power the whole layout?

 

Thanks in advance.

Your YC52 cable protects the PM42 from the power supply.

Your PM42 protects the command station or booster.

 

Adding the physical 1A fuse is redundant as the 4 track sections are already protected electronically by the PM42.

 

The PM42 is programmable and can be configured to set the trip current from 1.5 to 12 amps.

 

Inobu

Edited May 18, 2016 by inobu

[Ochanomizu]

Hello Mr Inobu,

 

Yes, I understand how the Digitrax parts work.  

 

My PM42 is set to 3A because I can run up to 7 x 8-car consists in one power district, if I have consists in the sidings also.

 

I make the following allowance per consist: 150mA for power car and 5mA for each direction lighting and 5mA per in-car lighting.  For an 8 car consist = 200mA.  To be safe I round up to 250mA ... x 7 = 1.75A.

 

From time to time I run EF65 + Tomix Cleaning Car + Tomix Cleaning Car.  I allow 0.5A.

 

PM42 is only programmable in 1.5A increments.

 

Of course, only one consist runs in each block at a time, so each block will have max 0.5A.  This is why I considered 1A breakers for each block.  However, I shall omit for now.

 

BTW, did you mean YC52 protects booster from power supply? I think so.

 

Thank you for your advice.  I shall continue without 1A breakers for now.

[kvp]

The idea is to have only one active train in each power district or at least make sure that the highest current set for each district is tolerated by all trains. For N scale, i would set the pm42 to 1.5A. If the stopped trains in the siding trip it, then subdivide it and add more pm42-s, also set at 1.5A. This limits the peak current sent through a single bogie or car in case of a problem.

[Davo Dentetsu]

Either the bulb wattage is too high or the springs contact with the pickup is poor.

 

Inobu   

I'm more inclined to go with the more basic "the layout chucks out a current that is rather unchecked when an issue does arise" issue first.  No bulbs have been added/ changed and all locos/coaches have discoloured springs thanks to the overloading.

[inobu]

I'm more inclined to go with the more basic "the layout chucks out a current that is rather unchecked when an issue does arise" issue first.  No bulbs have been added/ changed and all locos/coaches have discoloured springs thanks to the overloading.

ok, how are you going to regulate current? 

 

Inobu

 

I am asking questions to invoke thought

Edited May 19, 2016 by inobu