Tomix and DCC

[alpineaustralia]

I should say Capt, my assertion that the Tomix motors need more power to start is based solely on having to turn up the speed to get them to move - it is not based on any scientific measurements as you have carried out.

[CaptOblivious]

I should say Capt, my assertion that the Tomix motors need more power to start is based solely on having to turn up the speed to get them to move - it is not based on any scientific measurements as you have carried out.

What? You don't have a multi-meter? Shame!!  Seriously, though: What are you comparing your Tomix Dr Yellow to? A Kato shinkansen, I presume…I tried a good oranges-to-oranges test, but I don't claim the results to be universal. It is quite likely that the 209-0 uses a different motor entirely from your Dr Yellow, with radically different characteristics. I notice, for example, no capacitor. It just means (most likely) that Tomix models are all over the map (at least in comparison to Kato models) in terms of their operating characteristics.

[Martijn Meerts]

I doubt it's the motor in the Dr. yellow. If it was I would've had problems by now as well, unless I have a different production run of the Dr. Yellow which has a different motor, which is doubtful as well.

Another issue could be if you're using those connection tracks for example rather than soldering wires to a regular piece of track. Connection tracks usually have a capacitor in them.

That said, I have 3 dcc-equipped trains now which don't run as good as I'd want them to, all 3 are Kato..

[alpineaustralia]

I did have a multimeter but it broke and at $60 I couldnt be bothered trying to replace it.

Martijn - my Dr Yellow worked for a few weeks before it started playing up. My 300 worked for the best part of 3 months before it started playing up. My complaint and Bernards complaint is that the failure is intermittent and not predictable.
 
Bernard, please feel free to wade in with any additional informaton that might help the guys.

[CaptOblivious]

I did have a multimeter but it broke and at $60 I couldnt be bothered trying to replace it.

Bah, mine was $3 from Harbor Freight :D

[Martijn Meerts]

I did have a multimeter but it broke and at $60 I couldnt be bothered trying to replace it.

Martijn - my Dr Yellow worked for a few weeks before it started playing up. My 300 worked for the best part of 3 months before it started playing up. My complaint and Bernards complaint is that the failure is intermittent and not predictable.
 
Bernard, please feel free to wade in with any additional informaton that might help the guys.

In those weeks, how many hours did they run, and how often did you service them?

[alpineaustralia]

Not often to both those questions. Why do you ask?

[Martijn Meerts]

It's generally recommended to give locomotives/motorcars a good clean after a certain amount of running hours. Especially when they have decoders installed they tend to be a bit more vulnerable to dirty track/wheels/contacts. Also, dust from the motor brushes will start collecting, causing the motor performance to go down.

Not sure exactly how often they should be cleaned, some people say about 15 hours, some 25 hours, some of the exhibition layouts clean their rolling stock at the end of each day of operation.

Another thing is that running a train for say 10 hours and then storing it for a long time isn't good either. Grease and oil will dry up, and the dust from the motor brushes will settle on the collector.


I'm just trying to come up with possible reasons. I don't believe the cause of the problems is that the trains are Tomix, because the place I buy most my stuff is an official Tomix dealer, and he has a decoder install service. He's had very few complaints. Other then that, I have 3 Kato trains that run bad, so it's not just Tomix. (Those Kato's have actually seen a bit of running before being stored for several months, so those might be fixed with a thorough cleaning and reapplying oil and grease)

[CaptOblivious]

Martijn: You clearly haven't read Alpine's cleaning and lubrication tutorials :D

[Martijn Meerts]

The question isn't how it's done, but how often ;)

[Bernard]

I did have a multimeter but it broke and at $60 I couldnt be bothered trying to replace it.

Martijn - my Dr Yellow worked for a few weeks before it started playing up. My 300 worked for the best part of 3 months before it started playing up. My complaint and Bernards complaint is that the failure is intermittent and not predictable.
 
Bernard, please feel free to wade in with any additional informaton that might help the guys.

Here I am.
Of the 4 Tomix trains I have only 2 have made it to DCC and only the 800 Shinkansen is reliable, it is also a new model by Tomix.
 My Tomix MAX is very unpredictable and have done a re-install 4 times already on that train. It's gotten to the point that you don't even want to touch it once it's running. I have even add extra jumper wires in the MAX for better contact which has improved it's performance.
 My Nagoya 7000 (Sparky in another thread) series was a total failure having fried 2 Digitrax decoders.

There have been many times I see a set on eBay or a new release from an online retailer, that I'm really interested and as soon as I see Tomix, I have second thoughts. I just find MicroAce and Kato more dependable when it comes to converting to DCC. 

[Martijn Meerts]

Well, like I said, I have 3 Kato's with below par performance, even though every single one of them used to run well. Granted, the C57-180 runs well, but the squeaking noise can't possibly be good.

That said, trains of any brand can get problems sooner or later. I've seen trains that come with factory installed decoders fail after a few hours worth of running, even Marklin, known for its reliability, has trains that are dead on arrival or die after only a few minutes on the track. Heck, even Lenz 0-gauge has had issues, and these locomotive are designed from scratch with DCC in mind, you can't even get them without decoders. Yet, they can still fail, even with only a couple of hours worth of running time.

I refuse to believe that a certain brand doesn't handle decoders well when the trains are built up in the same way, and the motors are nearly identical. Start and stop voltages may be different, but the motors used in N-scale consist of very few parts, and they are pretty universal. Every N-scale locomotive/motor car I've ever opened has had the same type of motor, the only difference being 3-pole and 5-pole. There are no additional electronics hidden inside the motors.

I can't say what the reasons are, it might be just plain bad luck, it might be a short circuit, overheating, dirty tracks/wheel/contacts, oil that got on the motor brushes, bad batch of decoders, power spikes on the track, etc. etc. etc.

[alpineaustralia]

I refuse to believe that a certain brand doesn't handle decoders well when the trains are built up in the same way, and the motors are nearly identical. Start and stop voltages may be different, but the motors used in N-scale consist of very few parts, and they are pretty universal. Every N-scale locomotive/motor car I've ever opened has had the same type of motor, the only difference being 3-pole and 5-pole. There are no additional electronics hidden inside the motors.

I am not sure this is the case despite the outward appearance of the motors being the same.

I hear what you are saying about all the reasons that might cause it (eg. bad luck, short circuit, overheating, dirty tracks/wheel/contacts, oil that got on the motor brushes, bad batch of decoders, power spikes on the track, etc. etc. etc.) but the fact that we are getting a pattern of behaviour between different users suggests a different reason related to the trains themselves.

[CaptOblivious]

FWIW, I've finally succeeded in getting a decoder into my Tomix EF81. All difficulties to this point were a combination of: Rather technical frame milling requirements because of the chosen decoder; and: me. These difficulties surmounted, we shall see if the Tomix design itself proves a further challenge. For now, it has not.

http://akihabara.artificial-science.org/2009/05/tomix-ef81-dcc-pt-1-disassembly-and-frame-milling/
http://akihabara.artificial-science.org/2010/01/tomix-ef81-dcc-pt-2-small-victories/

[IST]

Although I do not have this type of model loco and I think I will not have also, these are great articles to help others, I appreciate your efforts.

[CaptOblivious]

Although I do not have this type of model loco and I think I will not have also, these are great articles to help others, I appreciate your efforts.

IST, thanks for the kind words!

[CaptOblivious]

Something I've discovered: TCS's BEMF scheme makes my loco run like crap. With BEMG (back-EMF) enabled, running is jerky and the engine shudders like you wouldn't believe. With BEMF off (set CV61 to 0), it runs smooth as silk, as a (late-model) Tomix locomotive should. Go figure. Dither (another TCS option for improving running in some locomotives) seems to make running slightly worse too, but it's not nearly as noticable. To disable dither, set CV56 or CV57 (doesn't matter which; both are fine if you aren't sure) to 0.
Since these features are on by default, I worried that I screwed up the installation, but I figured it was more likely the advanced features rather than poor soldering. Turns out I was right! This might help some here who have installed a TCS decoder into a Tomix loco only to experience poor running. Don't know how this pans out for EMUs.

[Martijn Meerts]

The question is, precisely what do you lose with BEMF turned off, and how does it affect any planned layout of course =)

[CaptOblivious]

The question is, precisely what do you lose with BEMF turned off, and how does it affect any planned layout of course =)

Since BEMF is largely meant to improve low-speed running, and since the low-speed characteristics of my Tomix EF81 improve with it off (indeed, are quite good), I estimate that I lose absolutely nothing :D

On my Kato models, on the other hand, very light BEMF settings (using a Digitrax decoder) can improve low speed running just a tad. But they are also pretty smooth.

[bill937ca]

Something I've discovered: TCS's BEMF scheme makes my loco run like crap. With BEMG (back-EMF) enabled, running is jerky and the engine shudders like you wouldn't believe. With BEMF off (set CV61 to 0), it runs smooth as silk, as a (late-model) Tomix locomotive should. Go figure. Dither (another TCS option for improving running in some locomotives) seems to make running slightly worse too, but it's not nearly as noticable. To disable dither, set CV56 or CV57 (doesn't matter which; both are fine if you aren't sure) to 0.
Since these features are on by default, I worried that I screwed up the installation, but I figured it was more likely the advanced features rather than poor soldering. Turns out I was right! This might help some here who have installed a TCS decoder into a Tomix loco only to experience poor running. Don't know how this pans out for EMUs.

Are you using a Tomix controller when the engine is jerky and shudders?  I've heard of Tomix trains running rough on other power packs, but fine with Tomix power packs.

[CaptOblivious]

Are you using a Tomix controller when the engine is jerky and shudders?  I've heard of Tomix trains running rough on other power packs, but fine with Tomix power packs.

No, I'm using a DCC system ;) A Digitrax Zephyr. Also, just to be clear, with those options off, the loco runs smooth as butter, just like it did before conversion.

Also, it also does currently run more or less fine on my low-end Tomix DC throttle (having turned the above-mentioned decoder options off). It just cuts out very suddenly when the voltage to the decoder goes too low, and without flywheels, it comes to a shuddering stop. But that's to be expected.

[Martijn Meerts]

A while ago for some reason a motor in a Tomix locomotive decided to call it a day. It happened after a short circuit on the layout due to an error in the wiring. I ordered a new motor and installed it, but of course the old, broken motor was a great candidate for further inspection.

The motor is from a Tomix EF510, which I've confirmed to work with the constant lighting system without problems. As you'll see, there is nothing special about the motor which should cause Tomix trains to be systematically worse when converting to DCC compared to any other brand.


Image 01:
Overview of the parts. The whole thing is a bit bent out of shape, which happened when I tried to remove the flywheels. They're mounted on the motor rather well and required quite a bit of force. Not something I'd recommend doing with a motor that's still perfectly fine, unless you have specific tools for it :)

Image 02:
Inside the housing, nothing to see here other than the 2 standard magnets that are common to all motors of this type.

Image 03:
The windings/poles and the collector. Also, nothing special here other than the fact that it's a 3 pole motor. I would've expected a 5 pole because the EF510's actually run really well. I'm guessing the flywheels help a lot here.

Image 04:
The plastic cap with the motor brushes. As with the other pictures, no special circuitry here, just your standard motor.

[Martijn Meerts]

Right, I've now installed decoders into around 15 Tomix trains, varying from local EMU/DMU to shinkansen to loco's, from old Tomix to new Tomix, and most things in between. I still haven't seen any problem with any of them on either short or long term. The only problem I've had is a loco (with a selectrix decoder) that started jerking and getting power loss, but that was after well over 200 hours of running time with next to no maintenance done at all.

I know some people will still disagree (and if I had their experiences with Tomix and DCC, be it some systematic problem or just downright bad luck, I would probably disagree as well), but I see absolutely no reason to think Tomix behaves worse than other brands when running digital. Installing decoders can be a bit more difficult, most usually it's pretty much the same as MicroAce and (non-DCC-friendly) Kato.

Oh, and DCC-friendly Kato is definitely NOT DCC-friendly when you install anything other than their own decoder ;)

[alpineaustralia]

Boy you really bent that motor shaft. Could you see what was wrong with th emotor after you pulled it apart?

[Martijn Meerts]

The flywheels are on there REALLY tight, so I had to use a bit of force to get them off. That of course caused the shaft to bend like it did :)

I didn't check what was wrong with the motor really. It was shorting out (both on DC and DCC), which I believe started happening after a power surge on the layout (due to bad wiring), so some isolation might have melted somewhere, or maybe the collectors got bridged somehow.

The motors aren't really fixable anyway, or at least not worth fixing considering a new motor isn't that expensive.

[KenS]

That looks like a three-pole straight-wound motor from the photo.  How old was the locomotive?

I've always wondered of Tomix used skew-wound and/or five-pole motors for better low-speed operation. It would appear that in at least this case the answer is "no".

[Martijn Meerts]

It's from an EF510 form about 5 years old I guess. Tomix does seem a bit random with what type of motor they used. On the other hand, an EF510 isn't meant for low-speed operation, so a skewed and/or 5 pole motor wouldn't make much difference, especially not with the 2 decent flywheels.

[KenS]

I think the reason flywheels aren't enough is that at low speeds, the armature isn't turning fast enough (i.e., there isn't enough momentum stored in the flywheel) to fully overcome the "coging torque" that makes the motor want to stop between poles. But I'm no motor expert.

There has to be some reason all of these manufacturers seem to want to use either skewed windings or more poles, both of which reduce the power output of the motor, rather than just using flywheels.

[Martijn Meerts]

Flywheels aren't meant for better low speed running, in fact, I believe to get better low speed running, the flywheels should be removed altogether. The poles, or more specifically the amount of dead space between the poles is the most important there. The motor needs to be able to rotate at constant speed without as little resistance as possible. Skewed poles with very little space in between the poles is great here.

Actually, Marklin has a different type of motor for some of their H0 stuff (C-Sine motor / Softdrive Sinus), which basically swaps the position of magnet and poles. Instead of the poles rotating, the magnet(s) rotate. This gives them enough space to fit in 9 poles. I don't think there's anything that fits N-scale yet though :) Heck, even in H0 they're not using this motor type all that much ...

[rpierce000]

I took apart a Kato EF65-1000 (part#3011) and a round, light brown capacitor fell out. Capacitors MAY not be a Tomix only item. I do know that now the engine runs very badly and I cannot figure out where the capacitor goes back in. It has NO solder on it, it appears to have just been laid in there or press fit into some part of the metal of the body.

I realize this is a bit of a side issue, but has anyone else seen this? If so, where does it go?

Thanks!

[Martijn Meerts]

Is it a Kato or Tomix? You mention Kato, but the thread is mainly about Tomix :)

Anyway, a capacitor isn't something only Tomix uses, but many, if not most of the brands use it one way or another. When converting to DCC, the capacitor isn't needed anymore (the decoder has it built in), so they are removed.

In your case, the capacitor is indeed just stuck in there rather than soldered in. Usually it's held in place by a little plastic wedge like thing that site between the motor (the side with the power pickups) and the frame. The capacitor just bridges the 2 contacts of the motor.

[CaptOblivious]

I took apart a Kato EF65-1000 (part#3011) and a round, light brown capacitor fell out. Capacitors MAY not be a Tomix only item. I do know that now the engine runs very badly and I cannot figure out where the capacitor goes back in. It has NO solder on it, it appears to have just been laid in there or press fit into some part of the metal of the body.

I realize this is a bit of a side issue, but has anyone else seen this? If so, where does it go?

Thanks!

That is very odd. I'm guessing the cap wasn't stock. My kato EF65-1000 had a cap...but it was soldered onto the light board. Those models are such silky smooth runners normally...even without the filtering cap...that I would strongly suspect there is something else wrong with it.

[Martijn Meerts]

Cap, I've come across several trains recently that had the capacitor just clipped in place instead of being soldered. Not sure why they do it, other than make it easier to install a decoder, since you won't have to de-solder the capacitor :)

[CaptOblivious]

Cap, I've come across several trains recently that had the capacitor just clipped in place instead of being soldered. Not sure why they do it, other than make it easier to install a decoder, since you won't have to de-solder the capacitor :)

I've seen that too, but only in EMUs, not in any locos...

[Martijn Meerts]

Good point.. If I remember right, all the ones I've seen it in were EMUs/DMUs/Shinkansen ..