Hi,

My future father in law likes electric trains. He has quite an extensive one, actually. 16VAC power supply, fed thru the tracks as you might expect. I believe (but I'm not sure at this juncture) that the AC is converted to DC inside the train itself.

A problem he sometimes encounters is that a dirty or otherwise unfit section of track can make the thing 'stall' for a second. You have to tap it w/your hand to make it pick up correctly again. He does keep the track clean, it's just a 'transient'. Nothing you can get rid of permanently by conventional means.

So my idea is to charge a Big Cap, and when the circuit quits like that, it would provide a burst to move the engine by the bad spot. Total time it would need to provide power is like 5ms or something. The rating on the power supply is 2A, so I'm figuring that would be the uber-max the engine could draw.

Question: how to spec out a cap that can provide something like 1A for a very brief time like that to kick the engine a bit?? The cap would mount inside a following car and be wired across the motor terminals just like a filter cap. It would have to have some limiting resistor or something to get it to charge without seeming like a short, and also to bleed when not in use.
Weird problem, but a neat concept, I think. He is very curious if this is possible.
Any thoughts?

Thanks!

 

If you are sure the motor is DC, it sounds like it is worth a try. You need to know the rectified DC voltage. Read here on what a Farad is: http://en.wikipedia.org/wiki/Farad

Connect the capacitor to the terminals of the motor, so you get DC to charge it. You can calculate the capacitance needed for a particular voltage change using equations in that link. But, I am an empirical type and would just try whatever you have available that is about 5000 uF or more and 25 V or higher rating capacitor. Even a 2400 or 3200 uF might help.

John

 

What energy pulse? I thought it was to be connected across the motor terminals permanently, sort of like a smoothing capacitor. I am skeptical that 5 mS will be enough. Assuming the train is going 1 mph, 5 mS is about 0.09 inches. I was thinking a small resistor in series with the capacitor, say 10 ohm, 5W might help. But before getting too complex, it would just try a capacitor.

Two things to be sure of:
1) The motor and power to it must be DC; and
2) The voltage rating on the capacitor must be greater than the highest DC voltage available. 16VAC after rectification is about 23 volts.

John

 

Perhaps you could put the capacitor in the coal tender. After all, it is a bit of coal as far as the engine is concerned.

Yeah, I know, you are probably using pseuo-diesel-electrics.

 

2A for 10ms with a voltage drop of 5 V (about half of a 9V supply) gives a capacitor of 4000 uF.
1A for 2ms with a voltage drop of 10V gives a capacitor of 200 uF

 

Not entirely sure this is going to work....
Where exactly are you going to put the cap? Most electric tran sets I have seen, allow the train to go forward and back.

There are two types that I have seen.

1) Definitely DC
The voltage on the track is +ve for left wheel and -ve for right wheel to go forwards, and reversed for reverse for example. This obviously would be disrupted by a cap across the supplies, the cap would be fighting any braking or reverse signal

2) Digital
+ve and -ve are in fixed positions, and each train is addressed via a digital wave through the +ve wheel. Much like the dallas 1-wire protocol I would think.
This would be rectified out by the cap!


I haven't investigated large circuits, but these two seem to be what is available to hobbyists in the UK.

To my understanding, unless the trains are controlled via a 3-wire system, I don't see how AC to the rails would give any control of the train (unless there is a control signal mixed in with the AC for a digital system)?

 

I think you need to use AC electrolytics. I assume the train can be reversed by changing polarity, and checking what a train transformer and its output looks like seems to confirm that reversing DC needs to be planned for. Beware of overdoing it - with too much Capacitance unless you want to build in the realistically slow stopping time for a train.

 

Electrolytic capacitors are polarity sensitive. They have only one correct polarity. Inverting that could cause total failure. If two-way motion is a must, electrolytic capacitors are not an option, in my opinion. All this, of course, for a pure DC motor.

 

Yes, there's a control signal sent along with the AC....it is probably a DC control signal (again, "?"). And the AC is rectified prior to being applied to the DC motors, so we're still in business. He is trying to send me the exact specs today, so I'll know more this evening...

He has an old engine and coal tender to experiment with, and has been told the risks, ha ha. VERY good suggestions, guys, that complement the idea I have had of using a cap across the motor itself. Possibly with a resistor in series...
How about 2 caps, one in "each direction" so that when the polarity is reversed (direction change) the other one is "in action"? Could use a diode across them, something I've heard of?

Thanks again!

 

Since your father-in-law's (future one) train is a little old, as you said, it is possible that is uses a transformer in the control box to variate the voltage fed to the rectifier of the train engine. If it is so, maybe you could try this trick too:
http://www.allaboutcircuits.com/vol_2/chpt_9/6.html
You can see what I mean in the last picture of the page. The reasonance freaquency you want to get from that tank circuit, is that of your wall-plug, 60Hz in the States, 50Hz in Europe, but I guess you arleady knew that.
It's incredible what you can find in that e-book, right?

Edit: On the other hand, I forgot that your problem is in the power reception at the train, so all the above is just a cool way to protect your system from peaks in the power supply. Too bad...

 

I think if you place 2 identical electrolytics positive to positive you can make the equivalent of an Alternating Current electrolytic with half the total capacitance rating of either.

So two 30 Volts 27uF placed positive to positive across with their negatives connected across the motor would be about 13uF and safe for about 15-20 Volts AC?

I would need somebody else to say if I have that right because I don't want your father in law to balme you and then you blame me if this turns out to be wrong and catches his train on fire.

 

I expect that the AC rides on a +or- DC level that triggers a relay in the train?

Note that the AC power seems to be rectified to DC in the train but it will be changed in polarity when the train reverses direction which is what seems to make the Electrolytics a problem unless you can place them between the rectifiers and the relay that switches the polarity feed to the engine.

Too much guessing is involved without seeing a diagram and details for the engine set up.

 

Here is a version of the circuit I imagine as the one we are describing. I don't know the exact size of the capacitors, it should be found by trial and error, according to the motor's demands. The PIV rating of the diodes should pose no problems (4001:50V, 4007:1000V). The only problem I see is that when you work the train on low speeds, the capacitor won't be charged enough to drive the train for the same distance.
And as always, experiment on futile equipment and toys. If anyone sees a gross mistake in my proposal, please say so.

 

If it's a Lionel, (most O gauge trains) everything is AC. The direction is reversed by a relay inside the loco that flops every time you go to Stop, which reverses the field windings. If it's H.O, it's going to be DC, or pulsed DC. No model trains that I know of use rectifiers in the locos....they're either AC or DC.

Eric

 

He says it's an HO scale train, and uses a decoder inside the engine. The decoder makes nice little sound FX and all that, but in the circuitry, it also rectifies the AC to DC for driving the motor.

I think Georacer may be onto something with the diodes to block the off-polarity voltage. My only concern now would be that I will somehow interfere with the control signals that ALSO ride on the AC. We'll worry about that after some testing )

He has some top-of-the-line engines (it's a "Real" train), but will be using a junker for the test, so things should go reasonably ok. The cap & diodes will ride in a car behind the engine w/connecting wires right to the motor up in the engine....
Thanks for all the replies, good info!!
Keep 'em coming!!

~MJP

 

One problem with the diode / capacitor combination is that the cap can charge but cannot discharge. That rather defeats the purpose, doesn't it?

 

The diode idea looks attractive, but what will happen when the power fails momentarily? The power source to the motor then is symmetrical. How will it know which way to turn? I have not thought about whether the capacitor can discharged, and have no experience with such an arrangement, because there is perhaps another, greater problem (see below).

Assume the engine circuit looks something like this, minus the capacitor.

Then, if you add a capacitor or capacitor-resistor to the supply for the controller, it might keep the train running for a short time after power is interrupted.

But here is the problem: what will happen with the train when the control signal is also missing? That is, assuming the control signal comes through the same track as the power comes, if you interrupt one, you interrupt both. If the train had power to the controller, and the control signal is lost does it stop or continue to run? I suspect it will stop, so none of these solutions will work.

John

 

You may well be right there, JPANHALT. He did mention that if the control signal is lost, the train stops.
I'm just trying to help him figure out if this can be done; I have doubts just like you guys! lol
He can try a cap across the motor - maybe damage it, he would be ok with that.

I wonder why nobody else has posted something like this on the web? You'd think if it was possible, it would be there, as this is a desirable type of circuit....so....hmmmm.

 

Fail safe systems are quite common in model aircraft, maybe boats and cars too. Basically, you program a "safe" setting for the controls. Then, if the signal is lost, those settings are set. There are variations on the system, but that is the basic version.

Unfortunately, for your model train, we do not know anything about the control system. I am sure it is more complex than when I had a Varney "Casey Jones" engine. From what little I have read recently, the controls can be quite sophisticated.

My original response about a capacitor across the motor was thinking on a ultra simple set-up, as there was nothing else to go on. If you were to try to reverse the direction, that system could create a serious problem. At the least, you might pop the capacitor.

At this point, if you know someone with an oscilloscope, it is probably worth using that to figure out the motor control. It is likely an H-bridge with PWM to give both directional and speed control. If that is the case, then you can work back through the circuit to find what is controlling the PWM. It is very like that is a microcontroller. If that is the case, you are probably out of luck. The controller manufacturer might be interested in your idea to add fail-safe programming (i.e, continue in the same direction and speed upon loss of signal). Then all you would need to do is come up with a way to maintain power.

Alternatively, model railroading sites and forums may already have address this problem with DIY controllers.

John