Tomix CL controller - does not work with some new Kato models!

[clem24]

Forgive me if this has already been discussed, but it doesn't seem like it has. I've brought it up in the past but likened it to possibly an issue with my model. But... I can now positively confirm that at least 2 Kato models DO NOT WORK with my Tomix CL-1000 controller, even if I have everything set to nil (i.e. no lights at a standstill). The 2 models that don't work are my Kato TGV Duplex (but oddly, the Thalys and AVE works, meaning the Duplex must have newer electronics) and my just arrived Toyama Tram.

 

TGV Duplex: the lights come on, the train moves about 3 inches, than stops. I can hear a slight buzzing/whining.  I cannot restart the engine even with full throttle, but the lights stay on.

 

Toyama Tram: the lights come on VERY briefly and then the power LED on the Tomix controller goes red indicating a short somewhere! I freaked out when I saw this, thinking I killed my brand spanking new tram, but luckily, no damage was done.

 

Any thoughts on this? I've now had to revert back to my MRC controller for these guys...

[CaptOblivious]

Without an oscilloscope trace I'm not sure that I could explain why with the lighting control set to zero, the throttle still wouldn't work…

[cteno4]

now this is a nasty way to be competitive!

 

with all the little circuits going into trains these days it was bound to run afoul of the more sophisticated throttles like the tomix cl series...

 

bummer.

 

jeff

[Martijn Meerts]

I've noticed an old, 15 year old Minitrix loco doesn't much like the Tomix CL-1001. Lights come on instantly, and there's a weird noise. That's with both the throttle and adjustment ring set to 0. It definitely outputs something, even at 0 ...

[The_Ghan]

Yes, both of my CL-1001's put current onto the track with both dials set to zero.  I'm waiting for my new multimeter to arrive to investigate further.  At first I thought this might be to power the headlights / taillights ... but it doesn't work with my Tomix Shinkansen 0 Green set.  Any thoughts?

 

Cheers

 

The_Ghan

[brill27mcb]

Forgive me if this has already been discussed, but it doesn't seem like it has. I've brought it up in the past but likened it to possibly an issue with my model. But... I can now positively confirm that at least 2 Kato models DO NOT WORK with my Tomix CL-1000 controller, even if I have everything set to nil (i.e. no lights at a standstill). The 2 models that don't work are my Kato TGV Duplex (but oddly, the Thalys and AVE works, meaning the Duplex must have newer electronics) and my just arrived Toyama Tram.

 

TGV Duplex: the lights come on, the train moves about 3 inches, than stops. I can hear a slight buzzing/whining.  I cannot restart the engine even with full throttle, but the lights stay on.

 

Toyama Tram: the lights come on VERY briefly and then the power LED on the Tomix controller goes red indicating a short somewhere! I freaked out when I saw this, thinking I killed my brand spanking new tram, but luckily, no damage was done.

 

Any thoughts on this? I've now had to revert back to my MRC controller for these guys...

I reactivated this topic so that is doesn't misinform people that find it through searches. You cannot "positively confirm" that modern Kato equipment doesn't run with the Tomix 1000-CL controller, only that two of yours don't seem to. I say this because our club has run many Kato Unitrams with no problem on a 1000-CL. (Unfortunately, no one has brought a Kato TGV Duplex yet for us to play with...) Your post is carefully worded to say just this, but the topic title is not.

 

I have read a number of past posts on this forum about the mystery of the Tomix "constant lighting" CL feature. It actually is not very mysterious, and is much like the old Linn Westcott/Model Railroader TAT-IV electronic controller design, as well as many commercial controllers that use pulse width modulation (PWM). It applies high-frequency pulses, which get wider as the CL adjustment ring is turned. Turning the speed throttle dial increases the steady DC voltage underlying these pulses. Here is a link to a Japanese description (author unknown) with oscilloscope images (a multimeter won't work to show this):

 

http://www.geocities.jp/motorcityrally_2061/N/cl_power.html

 

The pulses are limited to 12 VDC, unlike the high-voltage pulses in old U.S. "Troller" power supplies that zapped right through the insulation in some can motors. As the author shows, the pulses range from being on about 5% of the time (note that is does not turn fully off) to about 30% on. His later scope images show the CL left on full, as the baseline DC rises from the zero-volt line as he turns the throttle dial up.

 

The pulses are short enough that an iron-core motor doesn't really react to them, and a small capacitor across the motor for electrical noise mitigation (as required in Europe) would actually serve to act as a bypass to this "AC" component. I do not have any Tomix rolling stock with CL lighting installed, and it would be interesting to reverse-engineer the CL  lighting circuit board diagram for them. Anyone care to do this? Chances are it has a capacitor and resistor to "spread out" the pulses to keep the lights on steadily, plus maybe a bridge rectifier or diodes to isolate this lighting circuit and keep the voltage polarity on the capacitor unchanged when the track polarity is reversed.

 

Rich K.

[cteno4]

Rich,

 

thanks for the CL info, pretty much in line with all i have heard. interesting as the circuit must turn the pulse width back into a variable voltage (wider the pulse, the more the voltage). you are right its good to follow up on threads like this as jns posts come up in a lot of searches.

 

i do wonder if some of the circuits now being put in motor chassises may be affected by CL pulses and cause some of these problems.

 

interesting that the voltage regulation circuit in the unitram is not affected.

 

cheers

 

jeff

[inobu]

I have been trying to follow along but something doesn't make sense.

 

Are you running DCC? I ask because you stated you were going back to your MRC controller unless it is the MRC Hybrid.

 

The next question would be are you trying to zero stretch your decoder using the CL-1000.

 

The CL-1000 uses PWM to the track where as a DCC decoder places the PWM onto the motor leads and need the command string from a DCC controller to do it. Does the CL-1000 have DCC command sets.

 

I saw some where that the CL-1000 can damage some DCC decoders.

 

Something does not add up.

 

Did a little search and it seem like you haven't gone DCC or have any decoders.

 

So your trains runs slow or don't run at all.

 

Lets see what going on.  

 

 

5502 Power Unit N-1000-CL (1997 - 2009) - Input: 100 VAC 50-60 Hz. Outputs: 0-12 VDC 0.7 A (dial control, forward/off/reverse toggle) with CL constant lighting outer ring adjustment dial; 12 VDC 0.2 A for TCS; 12 VDC 0.3 A on left side for stackable control boxes. This unit also has an on/off power switch and a power/fault (green/red) bi-color indicator LED. It contains no power transformer, being fully electronic, and has quick-acting fault protection, reset by turning the unit off and back on. Its actual output is reportedly 20 kHz 12-volt square pulses with Pulse Width Modulation (PWM) to control the average voltage level. This power form can light the train while stopped, while still providing speed control. It provides excellent slow speed control, but reportedly can damage some DCC decoders. Replaced by similar-looking 5506.

Note: In July 2007 Tomix issued an electrical product recall for 5502 units built in May and June of that year, although they stated there was no problem in normal use. Affected units have serial numbers JF06501-JF10820 (on back of case). Lot numbers (on lower left of package barcode) include 20JF for 5502 packages, 51YJG for units in 90144 EX 485 Basic Set, and 59YJG for units in 90943 My Plan DX Set.

 

It has to be a polarity issue with the lighting board like Rich said. 

 

Inobu

[CaptOblivious]

Brill, those o'scope readings are nice! I've been asking to see such for a while now, but none of our members with CL throttles have o'scopes, and none of us with the 'scopes have a CL throttle. Those images explain quite a lot.

 

The only difference between the CL- and non-CL-lighting in my Tomix models is this:

CL-light boards have an LED and current-limiting resistor wired in series; non-CL-lighting uses bulbs and have a bypass capacitor wired in parallel with one of the bulbs. We'd often puzzled about that cap, but its purpose is now imminently clear to me.

 

(an image of a pair of CL lightboards can be seen here: http://akihabara.artificial-science.org/files/2009/05/04-trucks-off.jpg )

 

Notice that there is no additional circuitry to the CL lightboards beyond what is needed for regular DC use. The reason is pretty obvious: LEDs are current-driven devices, not voltage driven, and have very fast response times (whereas bulbs are voltage-driven, not current-driven, and have very slow response times). LEDs respond rather nicely to PWM without the need for a low-pass filter to convert the PWM signal to a linear signal. At about 75% duty cycle or so, they appear as bright as when fed a constant voltage, and they fade smoothly as the duty cycle decreases. Bulbs just get hot, and are dim; indeed, all my Tomix lightboards with bulbs have capacitors on them, presumably to shunt away any high-frequency signal to prevent overheating.

[inobu]

The trouble seems sporadic because of the variables. I believe that the lighting board and type of units has its effect.

 

Some trains (older and non bullet trains) have directional lighting boards with independent directional circuitry where as the newer bullet trains have integrated circuitry. The blocking diode and reverse cathode in the IC lighting boards operated one LED's based on track polarity. I think this is the issue.

 

The test would be to operate the motor car only without the lighting board and see what happens. The results will identify the direction of the problem 

 

Inobu

[brill27mcb]

The text that Inobu quoted in blue is from the Tomix Modelers' website, I wrote it and now I see that I need to modify it (once again!) after this further study. I originally thought the PWM provided all power, but now understand that it is only one of two power sources in the 1000-CL. One is 12 VDC pulse and the other is 0-12 VDC smooth DC. The pulse width is controlled by the CL adjustment ring and goes from about 5% minimum (on the scope it looks more like 1 tick out of a 13-tick long cycle or about 8%) to about 30% maximum duty cycle. Anyway, it doesn't go to 0% (off). The 5% or even 10% of 12 volts averages out to 0.6 to 1.2 volts at minimum setting, not enough to turn the motor but it may cause a high-frequency hum to come from it. That is also why there is measurable voltage on the track when everything is in the "off" position. A good reason to use the center-off position on the reversing switch...

 

Jeff, I think the creation of the pulses for CL purposes is a totally separate circuit from another that creates smooth DC for propulsion purposes. They are combined within the controller, as Westcott did with the TAT-IV, and separately adjustable by the ring and the main dial.

 

Thanks, Cap'n, for the info and photo of the CL lighting board. It is simpler than I thought, and with no capacitor on it, it needs no polarity protection. The LED, being a diode, simply lights in only one polarity direction.

 

I have seen differing numbers quoted for the pulse frequency, but they are all high enough that they would energize the LED without affecting the motor. I guess the LED, without a capacitor, actually turns on and off but so quickly that our eyes do not see it.

 

The statement that the CL might damage some DCC decoders is hearsay since I have no experience with it. I passed it along because if someone's decoder does get damaged, it will seem like more than hearsay to him!

 

To me, the most interesting part of the Tomix 1000-CL is that it is so light that it cannot contain the usual large transformer. It must drop the commercial household voltage using solid state, which I believe is unique. The replacement 1001-CL uses a converter brick on the power cord, as Kato does.

 

I am curious what sort of circuitry, beyond the small anti-noise capacitor (and possibly a small choke), is being added around the motor in more recent loco designs. Has anyone looked into this?

 

Rich K.

[cteno4]

Capt,

 

thanks mucho this helps my brain wrap around this! now i see the 12v hf pw supplies the led via the drop down resistor and the smooth dc never ends up adding to it as it never really reaches the voltage necessary to drive the leds with the dropdown resistor. the width of the pulses then determines how long the led is on/off for and thus how bright it looks overall. been about 30 years since my ee courses!

 

cheers

 

jeff

[CaptOblivious]

Martijn, in a recent separate thread, has shown that recent Tomix motors have nothing special in them. I myself have yet to see any kind of choke or inductor attached to any motor (besides the windings inside the motor itself, which, of course, is quite a big inductor!) ; the cap is standard on Tomix fare, and appears to be a small-value ceramic (non-polarized) cap of, I'm guessing here since I don't have one at hand, perhaps 100pF.

 

http://www.jnsforum.com/index.php/topic,956.msg37710.html#msg37710

[inobu]

The CL-1000 can trash out some decoders very easily. Riding a chips low or high voltage threshold can cause it to burn out. 

 

I really think there are a few issues at hand.

 

 

• The operating principles of the CL-1000
   
  Our in ability to read the directions.
   
  Using multiple units for testing.
  Some lighting boards and their adverse effects on the controller.
  
• 
  

   

 

 

I suggest testing with a shorty or facsimile of one and starting there.

 

 

Inobu

[Martijn Meerts]

Maybe we should look at it from a different perspective as well. So far, we've all been looking at what Tomix does to make the CL system work in their trains. But it might also be other brands add circuitry that is not compatible with the CL pulses.

 

For example, I have several Kato cars with tail lights which I wouldn't put on a DCC layout for more than a few seconds. They actually start producing a noise as if they're causing an every so slight short circuit (they're usually end cars with bulbs rather than LEDs). Of course, DCC is different than the CL system, but it's just an example that not all trains are the same. (Interestingly enough, the end-car in question is 1 of the 2 in the Orient Express set. The other end-car works just fine :))

[brill27mcb]

I have some further info. I first found the web page with the scope images for CL power a few years back, but did not save the link. It turns out the link I found and posted here earlier this week is a different page by the same person. It actually shows the output of the Tomix 5017 Operation Unit-CL, the first Tomix power pack with the constant lighting feature. The output of the later Tomix 5502 Power Unit N-1000-CL is different, and in fact does seem to match my description on the Tomix Modelers' website. The 1000-CL apparently has only a pulse width modulation output, with no steady DC component added. The pulses go from narrow width (but again apparently never fully "off") to full width, and have about 20KHz frequency. Luckily, I learned this morning how to find an archived copy of the original (now gone) link that shows the 1000-CL output:

 

http://web.archive.org/web/20071201130004/http:/www7a.biglobe.ne.jp/~diagraph/special/n1kcl_hakei/index.html

 

My reading of an Excite-translated version of this page leaves unclear the operation of the two dials -- the CL adjustment ring and the throttle dial. The author only seems to mention turning one dial, and it seems to be the CL adjustment dial.

 

Even more interesting is that near the top of the page the author mentions removing a capacitor from his DD51 locomotive to solve an issue. In another current topic here on the JNS forum, removing the small capacitor that is placed across the motor to prevent electrical "noise" is often recommended as part of DCC installation:

 

http://www.jnsforum.com/index.php/topic,3769.0.html

 

So, it makes sense to me that motorized units, especially those intended for the European market (like the Kato TGV Duplex and the 15-year old Minitrix locomotive, the two specific units people have reported with problems) probably contain such a capacitor. That capacitor MAY cause the problem with the Tomix "CL" type of controller, just as it causes problems with high-frequency DCC signals.

 

Rich K.

[inobu]

Most likely the issue will be in the capacitor and its dielectric properties. Kato uses a different capacitor than Tomic and its breakdown voltage is most likely the culprit.

Hence the suggestion of removing the lighting board.

 

 

 

Inobu  

[brill27mcb]

Hi Inobu,

 

If the capacitor across the motor is mounted on the "lighting board," by which you actually mean the only board in a locomotive, then I agree. That would be a simple test. If it works, one could then reinstall the board (to have lights) and carefully unsolder just one lead of the capacitor to disable it but leave it in place, or just remove the capacitor entirely.

 

If you are talking about some other capacitor, on a board involved with lighting only, then I do not understand.

 

Rich K.

[inobu]

There are so many variables evolved that one has to start eliminating the possibilities.

 

Kato has a number of lighting boards , Some have blocking diodes others have capacitors to manage current flow in order to control the directional LED lighting. This configuration has the LED's cathodes in opposite directions so they respond to track polarity accordingly and it is the blocking diode or capacitor that dictates current flow.

 

Based on the symptoms posted the controller sees a short in come cases. The only board that Kato has is a lighting board which consist of LED's, a resistor and diode or capacitor. Kato's design is not based on CL but directional lighting and may use a different capacitor that cannot handle the CL-1000 voltage pulsing.

 

If the breakdown voltage in a capacitor is hit it becomes conductive and may create a short that the controller is detecting. It is hard to say.

 

The best test would be to test with a shorty. It is as plain as it gets. If it runs then we know that the Kato's basic dc motor configuration works and the only other board on the unit is the lighting board.

 

Inobu

 

 

 

 

[clem24]

I am glad this thread got another life. Yes I guess in light new info from others, I guess it is safe to say that the CL controller may not work with "some" Kato models. I did receive another Toyama Tram and I will try this one with the CL controller and report my findings.

 

All the electronics talk above goes over my head hahaha.

[clem24]

OK, so there goes my theory. Yes the CL controller *does* indeed work with the other Toyama Tram. Go fig. CL function sort of works. There's a very fine line between light coming on and train moving forward. It actually adds a bit of momentum when bringing the train to a stop with lights still on, which is actually a neat side effect.

[brill27mcb]

Yes, my hope is to get to the bottom of this issue, sort it out and come up with some sort of solution(s). I don't believe the issues are as complicated as was first imagined. I will be doing some experimenting. It is unfortunate that the Tomix power unit arrangement does not let you just turn the CL feature (pulses) fully off when you want to. That would be an easy solution.

 

Rich K.

[CaptOblivious]

Yes, my hope is to get to the bottom of this issue, sort it out and come up with some sort of solution(s). I don't believe the issues are as complicated as was first imagined. I will be doing some experimenting. It is unfortunate that the Tomix power unit arrangement does not let you just turn the CL feature (pulses) fully off when you want to. That would be an easy solution.

 

Rich K.

 

I find that surprising too. But it wouldn't be hard to build a small lowpass filter that you could attach between the throtttle output and the feeder for when you want to use it with a non-CL friendly train.

[brill27mcb]

Yes, my hope is to get to the bottom of this issue, sort it out and come up with some sort of solution(s). I don't believe the issues are as complicated as was first imagined. I will be doing some experimenting. It is unfortunate that the Tomix power unit arrangement does not let you just turn the CL feature (pulses) fully off when you want to. That would be an easy solution.

 

Rich K.

 

I find that surprising too. But it wouldn't be hard to build a small lowpass filter that you could attach between the throtttle output and the feeder for when you want to use it with a non-CL friendly train.

Please, tell us more! How about a schematic?  -- Rich K.

[CaptOblivious]

Yes, my hope is to get to the bottom of this issue, sort it out and come up with some sort of solution(s). I don't believe the issues are as complicated as was first imagined. I will be doing some experimenting. It is unfortunate that the Tomix power unit arrangement does not let you just turn the CL feature (pulses) fully off when you want to. That would be an easy solution.

 

Rich K.

 

I find that surprising too. But it wouldn't be hard to build a small lowpass filter that you could attach between the throtttle output and the feeder for when you want to use it with a non-CL friendly train.

Please, tell us more! How about a schematic?  -- Rich K.

 

Two yeas late! Lowpass filters are easy: an inductor from input to output, and a capacitor from input to ground. The exact values depend on where you want the cutoff frequency to be.