Resistor controlled trains!!

[daan]

The finish of my homemade resistor controller is near. Today I made a temporary setup with all the components installed. I removed the plastic between the contactpoints and made the levermotion freely and smooth. The resistors are mounted on a aluminium cooling plate to keep the heat in control. For the setup I placed the resistors on top of the layout, so that I could contol the amount of heat. After 20 minutes of driving the resistors where stil easy touchable (about 25 degrees celcius) and even a full short cirquit with a screwdriver on the track didn't heat it up. (tried for 10 seconds)

In front is the resistorblock, the wires are still in temporary setup, they will be mounted via screwcontacts on the side of the resistorblock later on.

In front here are the two cirquit brakers. They blow at 4 amps, each one for a trackloop. With the shortage which I made to control heat building and fuses they stayed in for 10 seconds, then they switched off.

This is the resistorblock. I used all kinds of resistors, making a range between 0.33 and 0.47 ohms. They have 18 watts per step minimum, except the ones which are used when the train is starting. they have 10 watts each.

The trains run very good. I didn't make a resistor in front of the 24 volts yet, so the levers are not supposed to travel to full throttle yet. The resistor in front of the rheostats will be made with 8x 2,5 ohms 5 watts, resulting in 1.25 ohm 20 watts when finished. They will provide a 4 to 6 volt drop.
I've tested with 2 trains which both blew the cirquit braker in my heaviest transformer (2amps) at once, they don't lack power anymore and it's a joy to watch my F3's grouwl across the tracks at high speed. The drag of the lighted passenger cars behind it is no more, the lack of power is vanished.
My other train pulls all the cars I've got without stalling and the best is that the e-units stay in position 'till I pull the levers off the power. The steps in the resistors are hardly noticable, they start slowly and run smooth to the next step of the contactplate. They don't jump into a higher gear, but slowly gain speed 'till the leversetting is reached. I feel like a real enginedriver!

Thanks everyone for giving me confidence and help in this project. It started as a try-out and I didn't have the faintest idea that it would work out this well. Resistors which keep in very low temperatures, it can stand longterm shortages and has power to drive everything I like for the cost of $40 dollars in total. It's not completely finished yet, since I want to build a housing around the resistors and make the cablespagghetti much tidier, but the setup works like anything!!

[lionelsoni]

Daan, congratulations! I'm glad that you can experience what a nice feel rheostat operation has. I can't resist saying that I told you so:

"However, the operation is similar to running with a rheostat, or variable resistor, and a fixed transformer voltage. I was pleasantly surprised by the way it operated. It was much more like driving a vehicle, in that increasing and decreasing the current did not cause an immediate change in speed. I have since seriously considered switching over to rheostats for my layout, because of the nice feel it had."

Since you haven't put in the top-end resistors yet, let me make one last plug for the lamp idea, which I still think would be a fun way to operate. Remember that I said:

"I hope that Daan does try putting a lamp in series which would both provide fault protection and drop a few volts in normal operation....I imagine the lamp filaments glowing inside the clear glass bulbs as Daan moves the control up quickly and then fading back to a dull red as the train starts and he backs off on the throttle. I think it could be very dramatic."

[daan]

I'll try to find bulbs that are easy obtainable for a reasonable price. I've looked in the regular autoshops, but they are only 12 volts. The 24 volts 75 watts are quite expensive, around $20 a pair.. The resistors are still very cheap, and easy to get. It's not my purpose to invest another $60 on lamps only, I hope there will be some other way to get bulbs for less money though, it will put the extra touch to it..

[lionelsoni]

That is expensive. I was surprised you have them available at all, since I have never heard of such a thing here. When you first mentioned the 24-volt-75-watt lamps, I googled a bit, but all I was able to find were airplane landing lights.

You would get exactly the same effect as with 24-volt lamps in series with the transformer's secondary by putting the same wattage of 230-volt lamps in series with the primary, but only if there were only one rheostat. With the two rheostats that you have, there would unfortunately be an interaction between them. However, since 230-volt lamps are a lot cheaper and can be reused afterwards in your household, it might be interesting to experiment by doing it that way, with only one rheostat in use, just to find out whether finding 24-volt lamps is worth the trouble and expense.

[daan]

Those 24 volts 75 watt lamps are used in trucks and (that's why I know) where in my old Nissan Patrol 3.3D (which drives around in Africa now). The idea with normal lightbulbs occured to me too, but their resistance is much higher since they are used for 220 volts. If I would use a 200 watt bulb I have 0.9 amps, resulting in a resistance of 244 ohms. It is too high to use for that.
Since the resistance drops when not fully lit to 10% of it's resistance under power it will stil give 24 ohms. I've measured those bulbs allready on the shelve, and they come up with resistance of around 25 ohms when no power is applied... Shame, it would have been easy.
I used a 200 watt bulb in series with a 40 volts Bing train, with an old resistorbased powersupply (as mentionned on my website) and that went ok. the power was low enough to drive the train. The lightbulb took the 180 volts and left about 40 volts for the train. The only way to go when I want those lights in front is getting 24 volt lamps.
The whole problem is in the glowwire. In a 24 volt version the glowwire is thick and extremely short, resulting in low resistance. But in a 220 volt version the wire is thin and long. The only bulbs I could find cheap are the normal industrial 24 volts 5 watt bulbs, which would be usable as additional christmas lighting seen the amount I would need to get the +/- 250 watts I'm looking for...

[lionelsoni]

Daan, I'm not suggesting that you put a 230-volt lamp in series with the 24-volt secondary winding and the rheostat. I understand that that won't work. Rather, put it in series with the 230-volt primary winding. A transformer transforms not only voltage and current, but impedance as well. It will transform the 230-volt lamp in series with the primary to look like a 24-volt lamp in series with the secondary, from the point of view of the rheostat.

This is probably no good for operating your two rheostats, because it is like putting a single 24-volt lamp in series with both of them, which would lead to interaction as they both drew current from the same lamp. However, as an experiment, you can find out how a 24-volt lamp would work by putting the 230-volt lamp in series with the primary and using only one of the rheostats in the experiment.

(Of course, where I say "lamp", you probably would actually use several lamps in parallel to get the desired power rating, for example 3 75-watt 230-volt lamps in parallel.)

[lionelsoni]

Daan, here is an example that demonstrates how it works:

Suppose that you have 225 watts of 24-volt lamps in series with the secondary and you have a dead short through the rheostat and track. The lamps will draw 225 / 24 = 9.375 amperes.

Now suppose instead that you have 225 watts of 230-volt lamps in series with the primary and, again, a dead short through the rheostat and track. The lamps will draw 225 / 230 = 978 milliamperes. That current will be increased by the turns ratio of the transformer (230:24) to 978 * 230 / 24 = 9375 milliamperes at the secondary, the exact same current.

[daan]

And another idea is born.. Marvelous idea Bob, it will reduce the primary voltage, and thus reduce the secondary voltage. In this case I would want the voltage to drop secondary to 20 volts, so 4 volts have to be dismissed. In the transformer, that would mean (24/20x100=83%=> 100-83=) 17%. The primary voltage will have to be reduced by 17% too, meaning 37.4 volts. With a 6 amp current draw in secondary (24x6=144 watts) the primary would have (144/220=)0.65 amps. With a small amount of waste due to internal heating it would be 0,7 amps primary.
The resistance to get it to drop to 183 volts (220-37) will have to be (37/0,7=)52 Ohms.
A 200 watt bulb has 244 ohms, so I would need 4 in parallel..

Oops... That's a huge amount of power.
I'm sorry, but I think I'll stick with my resistors.. I have a few spare ones here and the additional resistors I need will set me back for about $5.. It's a shame for the lights, but I'm opting for the resistors.. Sorry Bob!

The problem with interfering trains when using the same resistors is very small. It's only noticable when I drive one and start another. When the second train runs, there is no difference in noticable speed reduction of the first train when I gain speed on the second one.. (though I have them both fed through the same resistors) The setup is much better working as I thought it would.

[daan]

To your second post Bob, the resistors can handle full shortage without any problem. I was amazed by that! I don't need any more protection than the two cirquitbrakers I installed. They didn't even heat up a lot. After 10 seconds of full shortage the cirquitbraker kicked in, and the resistors where still easily touchable. It means that the first reason for me to opt for the bulbs is not necessary. They can handle the stress and everything stays under control.
I'm really surprized of the capabilities of this thing..

[lionelsoni]

I would use the circuit breakers anyway, whether using lamps or resistors.

If you want 20 volts at 6 amperes, you need 386 watts of lamp rating, not 800. It's true that the resistance of a 220-volt 200-watt lamp is 242 ohms when it is fully lit, but it's substantially more when it has only one-sixth of its rated voltage across it, as it would have if the secondary drop is only 4 volts. In any case, it would not be consuming anywhere near 200 watts under that condition, but rather 18.66 watts.

If you've got some lamps lying around, why not try it?

[daan]

There is nothing against trying Bob, and I will just to know if it works and how it will affect the setup.

[lionelsoni]

Great! Thanks for satisfying my curiosity.

[jkerklo]

Congratulations daan! Job well done.

You have validated a technique to control toy trains. Now I have to think of how I will make use of your successful research.

John Kerklo
TCA 94-38455
www.Three-Rail.com

[ben10ben]

Excellent work! You have me interested in rheostat controlled trains now, and I may try building a simple salt water rheostat just for the sake of experimenting. Thanks!

[lionelsoni]

Ben, you might look on Ebay. I just searched on "Lionel" and "rheostat" and got about 8 hits, some in the original boxes. You can often find them at train shows, too. They go for only a few bucks. I think the model 81 is the nicest design.

[daan]

Well, the setup is still in this condition.. I've tried with the lightbulbs. The bulbs are not the problem, but the fittings where the bulbs have to screw in are.. Every firring I saw in diy stores can handle 60 watts max. To get to the 380 watts needed I have to use 5 60watt and 2 40 watt lamps if I want to use those fittings..
I'll have to search on fleamarkets to get sockets out of a solarequipment or something..
If I run a K-line 2 motor engine, I have to get some more resistors. It starts to run on 14 volts.. The project is still in experimentation, so I'll use higher resistances on the low end to get the voltage down more..

[ben10ben]

Bob,
Thanks, I will check on Ebay.

I was at a show today, and kept my eyes open for rheostats, but I didn't see any.

As for the saltwater rheostat, I tried it using two alligator clips and a glass full of saltwater. I set my KW for 20 volts, and dipped one clip in and out according to how much power I needed. It applied pretty near full voltage with just a couple of light bulbs, but I had pretty good control when the motor was turning. I was impressed by how finely I could control the speed of the can motored Hudson Jr. I was using, and the excellent slow speed I was able to get. The two problems I had, though, were corrosion of the electrodes(saltwater isn't very friendly to nickel plated steel), and electrolysis of the solution.

Electrolysis of aqueous NaCl(saltwater, in other words) always produces hydrogen gas, and sometimes produces oxygen gas or chlorine gas, or both. Only one electrode was bubbling, which means that it was only giving off hydrogen gas, which isn't poisonous like Chlorine, but is still very reactive. The last thing I want right beside my trains and their sparking is a whole lot of hydrogen gas that could potentially explode or catch on fire. For this reason, I think that I'll go with the Lionel rheostat.

[lionelsoni]

Daan, just to try it out, you could try connecting directly to the lamps, without sockets, and being very careful. I don't know what sort of lamp bases you use in Holland; but at least one connection is probably through a solder connection on the bottom, so you could just solder to that. If you have to make the other connection to the shell of the base, you could solder to that too if it is brass. If it is aluminum you could wrap a bare wire around and twist it tight. This would allow you to see what works without risking buying sockets that you might wind up not using.

[wrmcclellan]

Daan,

Congratulations!

Regards,
Roy

[daan]

Bob, I'll try with the version without sockets. If I use a box or shield the wires off it should work fine for experimentation... For now I'll be off to work. This setup is not difficult to get, so I'll try it out soon..
Thanx for all the help!

[Anonymous]

Beautiful job Daan, Cant wait to see what kind of housing you come up with........Tim

[daan]

Tim, I'm planning a stainless steel housing for the resistors. One made of a plate with holes in it and a fan..
The transformer will be housed in it too, but isolated from the housing, since this is a part where the 220 volts are on.. and I don't want the risk of a loose wire touching the steel casing and me touching it too..
It saves money for the hairdresser though.

[lionelsoni]

Daan, I went to a train show over the weekend and found a guy selling unused old locomotive headlights. I bought one for $5. It is a 200-watt 30-volt PAR-56 (7-inch) sealed-beam lamp with a clear front. In series with 24 volts it would give 4384 milliamperes at 16 volts, 4028 at 18 volts, and 3643 at 20 volts. (Twice that for two lamps, of course.)

[underworld]

Excellent work Daan!

underworld

[:D][:D][:D][:D][:D]

[daan]

Thanks! The best part is that it still works, though I've now only one selector with the resistors. It was tricky to get 2 trains running at once, since the resistors vary the voltage to the other train too when the first train is slowed down.
Since november I've been trying with lampbulbs, but that is not the good idea. The voltage changes when the bulbs heat up and they where producing a lot of heat.
Instead of the bulbs on the primairy side of the transformer I now made a switchable resistance varying from 1.1 to 8.8 Ohms, made of 8 2.2, 5 watt resistors. Through 3 relais with each 4 double contacts I can switch from 2 sets in series of 4 parallel (giving 2x 2.2/4= 2x 0.55=1.1ohms 20 watts) via 4 sets in series of 2 parallel (giving 4x2.2/2 = 4x1.1= 4.4 Ohms, 20 watts) to 4 series of 1 (giving 4x2.2=8.8 ohm 10 watts)
The 1.1ohm is used for the postwar twin motored F3 and the other older lionels asking huge power, the 4.4 part for the modern area Lionels (sd28) and the 8.8 for the K-line's, though even that is not enough when I run the K-lines without cars behind them.

The feeling of driving trains like this is not comparable to normal transformers. You have to give a bounch of power when the train accelerates, take a bit back if it's on speed and stopping is a gentle push back, or switching over to a higher preresistor rate. It is as smooth as butter, even on the older machines which are stubby and unwilling on a normal transformer.
I removed the cirquit braker, since it never switched off during the past half year..

I tried with a normal transformer, but it really was a huge step back. Only the K-line engines where better drivable alone, but I don't use them very often since they loose their tractiontires all over the place and tend to hop off the track whenever they like it.
(nice to find an old subject back in the board)

[FJ and G]

Daan,

That is really awesome. I'm impressed. And it looks really cool to boot.

Kudos!

[daan]

The trick is in the fact that the voltage drop over the resistors is set by the amps the train uses. If it's increasing speed the motors ask more amps and the voltage step on the left resistors is high. If it gains speed eventually the amps will lower a bit and thus the voltage left behind the resistors will rise. On the end there is a balance in voltage and amps which results in a certain speed.
It has drawbacks too, if an engine is consuming not enough, it won't work unless you add a big resistor in front. It changes power into heat, instead of using only the power needed like with a normal transformer and on a grade or a sharp curve, the train uses more amps and slows down, you need to operate the lever on the selector more then on a normal transformer. If coming down a grade the voltage rises and the train will speed up even more. But since I don't have grade's and I don't mind the speed dropping a bit in a curve, it's not a big deal.

On the older electric locomotives you'll see the same setup. You'll notice cooling tubes on the sides or resistors on the roof of a vehicle. In the cab is a lever or a wheel with a giant selector coupled to it. This also chooses several steps of resistor rating to control the speed. There is another way (sort of gears) to select the speed. On a six axle locomotive you'll find 6 traction motors, they are coupled to switches which can make a setup of all six motors in series, 3 parallel groups in series (2x3) or all parallel.
When moving slowly or starting with a load, 6 in series is used. After gaining speed it's switched to series/parallel (3x2) and when they are up to speed and only need maintaining it, they use 6 in parallel. Like a 3 step gearbox. The resistor selector is used to change speed in every "gear".

[lionelsoni]

It's the other way around, Dann, series at low speed, parallel at high speed.

The same thing is done on DC-motored Diesels. The switching from one connection to the next is called "transition". Originally, it was done manually; now it's automatic. Many locomotives use only two connections. A four-axle switcher is likely to use full-series and series-parallel, while a road engine would use series-parallel and full-parallel. Each type of connection is usually divided into two ranges, with each motor's field winding shunted by a resistor in the higher-speed range.

[daan]

Yep, you're right Lionelsony.. series is for low speed.. The one locomotive I'm referring to is the 1200 series from the dutch railways. It's a six axle locomotive based on the american 4 axle types like lionel made them in the 50's. It was (and still is, after 50+ years, speaking of building quality!) in use for all services. From shunting to high speed passenger transport. Therefor it had series, series parallel and parallel motor setup.