Basic electrics question..(photo's!!)

[daan]

Since I'm lacking power I'm thinking of wallet-healthy way's to get more juice to the tracks. I have a giant 24 volts 12 amps transformer. (About the ZW size), but unregulated. The thing in my mind is to make a lever which slides across multiple contacts, wired with 0.5 Ohm resistors. Every contact down will add 0,5 ohms. In theory a 0.5 ohm resistor under a load of 4 amps (max power for a railsection) will drop the voltage by 2 volts. The wattage I'm planning to use for this is 10 watts for every resistor. (2 volts x 4 amps = 8 watts).
If using 10 resistors to step down the voltage it will result in 100 watts of powerconsumption for a short time.
I know that if you put 2 resistors of 0.5Ohms in series it will result in 1 Ohm, and according to that the voltage drop at 4 amps should be 4 volts. BUT can anyone tell me if I'm right with the wattage of the resistors? I mean, am I right assuming that the load on the resistors will be divided equal (2x 8watts) or does it cause that one fries and the other handles the stress with two fingers in it's nose?
If I'm correct, I can make a double regulator with 17watt resistors, which is the heaviest in cheap resistors.

Plan: O-{R}-O-{R}-O-etc. The O's are the contactpoints, the R's the resistors.

That directs me to my second problem.
I want to have a short cirquit safety build in. I was thinking of a lightbulb in series with the regulator. But which voltage and which wattage should I take?
May be someone else has any bright ideas?

[SPFan]

Power will be divided equally between two resistors if they are the same value regardless if they are in series or parallel. However this is a poor way to control engine speed or regulate power even for fixed loads like accessory lighting. It will consume about half the power available at the transformer. Also any resistor used to dissipate power should be a made of nichrome wire. Any other material will change value with temperature. A light bulb is simply another resistor in series with the track and will further lower track voltage. For modern trains with electronics a fast acting fuse is best. For straight AC PostWar engines a fast acting circuit breaker will suffice. Your best bet would be to try and find a variac that would allow you to reduce the transformer output to 18 volts. Placed on the primary of the transformer it would only have to handle about 2 amps assuming your mains voltage is >200 volts? Use multiple fuses or circuit breakers to divide the individual functions of your layout, eg loops, lighting, accessories, etc.

Pete

[jkerklo]

You are right about the power shared by resistors in parallel: the power for the combination is the sum of the power for each (as long as they are the same resistance).

I have to make a few comments, however.

24 volts is a bit much for trains. Since the voltage presented to the train depends on the current draw, when the train is in neutral (small current draw), lights will get nearly
the 24 volts. Not good. Same for e-units and whistle relays. It won't melt them, but its on the edge. Probably a disaster for any electronics in trains; check the spec.

Your switching arrangement will need to "bridge" two contacts (called "make-before-break") when changing resistance, or e-units will drop out. This is why Lionel went to contacts on the transformer secondary in the 1920's.

Resistors aren't a good way to control voltage for trains. As the current goes up, the voltage goes down, which is backwards from what is wanted. It will work, but poorly.

If you really want to do this, I would recommend getting some Nichrome wire and making a variable resistor. Be glad to help if this interests you.

Ditto your "at your own risk" comments.

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

[Anonymous]

A better way to control that power would be to put a medium duty, say 5 amp 0-120v variac in front of the 24v/12a transformer. Use the big knob on the variac to vary the voltage just like a big toy train tranformer. You could actually run the trains from directly from the variac, but it would be too dangerous as you could easily put full line voltage 120v on the track. Variacs can be had for $20-100 and they are very useful items to have around - they are not the same as a wall "dimmer" switch. In front of the 24v transformer, a real variac would give you very fine control of the output at the track with very little loss. Just put a nice block of circuit brakers (one channel for each block) between the 24v transformer and the track. Put a voltage meter on the output to see what is happening. A full constant 24v will kill 12 or 18v lightbulbs very quickly. But I've read that prewar trains may very well need that much power to perform well.

This would give you a very solid 288 watt (real watts, not toy train tranformer watts) power supply. Any good variac will have a fuse or circuit breaker protection built in. The block of breakers on the track side will protect the 24v/12a transformer. Whistle/bell control is another matter.

Using those resistors will be very wasteful. They will get very hot and self-destruct if they are not rated for the power. Adequite power resistors may cost as much as the variac. I think your math is correct, but many designers use resistors with wattage ratings double what they expect to see in the circuit for a safety margin. You won't find them at radio shack.

Here's food for thought. I just bought a tired KW for $56.00 from a well-known, but little trusted auction site. I put a new $5.00 line cord on it and replaced the rectifier disc in the whistle circuit with a $1.00 diode. The rollers looked pretty good so I left them alone. It puts out 20.9 volts at full throttle. Works like a champ.

Be very careful with electricity. It can kill you or burn your house down if you mess up.

But's still lots of fun to build stuff like this.

Good luck,

[daan]

Guys, thanx a lot. I wasn't sure if it will work anyway, but in theory it should. The problem of tha voltage dropping when power needed gets higher is something I calculated too. With 2 amps, the voltage drop per resistor is only 1 volts per resistor, and if I run a S2 with 2 small DC motors, voltage per resistor will drop even more.
The wire-wound potmeters can handle 40 watts in maximum, due to heat building. There are bigger versions, but they are huge. If using loose resistors they can be mounted on a large cooling plate to get rid of the heat, but it will get hot and is only consuming power.
To get the 24 volts back to a more healthy 20 volts I figured that a lightbulb could be used, but also that has the problem that if the power needed goes up, the voltage will drop more, resulting in less power. The lightbulb also has the advantage that if you have a short cirquit I don't fry anything. But it's not an elegant way.

The whole problem is in the AC. If having DC it wouldn't be a problem to regulate. I've build my own speed controls for h0 with variable load and brake settings and all the fancy stuff, but that's with DC and that's another world.
For AC, as soon as it's out of your transformer, the only way to get it proportional is "eating" the not wanted bit and converting it in something else, heat or light.
The other option is making a transformer myself or use one with about 10 different voltages at it's output. It will need a complicated switch to make and brake all the different combinations in the transformer power output pins, but it won't heat the whole area. Which also requires to disconnect power before switching to another field-combination.

Thanx for all the advise, I'll calculate my stuff over again, may be take another set of resistors for the lower loads or completely bann the idea and get on the search for a decent transformer somewhere else. Hobbyshops nor electroshops have anything matching and the one who sells Lionel in Holland doesn't sell transformers in AC..

[daan]

@ old 2037; guess that the variac in front of the transformer is the only way to go if I can't find a transformer. It's the same basics as I do with the resistrors, but on the "high side" of the transformer, which doesn't require a lot of power. Resulting in less heat and better control.
I've seen those variacs too but don't know the prices asked. I even had one a long time ago, but I trew it away. Couldn't find a purpose for it on the h0 layout[banghead][banghead]...
Thanx everyone for the ideas and help offered.

[lionelsoni]

Daan, don't give up on resistive control yet.

Last Christmas, I very carefully hooked up my Lionel model 107 reducer to a prewar train. This is a resistive voltage divider, intended for running trains in households supplied only with 110 volts DC. (There was another model for 220 volts.) This is a pretty scary item, not only because it is insulated with a slab of asbestos, but because, depending on which way you plugged them in (American plugs were not polarized until much later), you had an even chance that the toy-train track rails were connected to the line voltage.

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.

Lionel made rheostats, like the number 81, 88, or 95, for use with batteries or fixed-voltage transformers. They are not expensive; and I have amassed a pile of them.

Twenty-four volts is not that high. That is the maximum voltage from the type Z transformer, which was sold well into the postwar period.

As for circuit breakers, I recommend 5 or 10-ampere auto-reset ATO-type automotive circuit breakers. (I'm guessing that with the internationalization of the automobile market, these are also sold in Europe.) An incandescent lamp in series is actually not that bad an idea. It can exhibit low resistance under normal current; but the resistance goes up greatly as it heats up in a fault condition. Since there will be some voltage drop under normal conditions, having a few extra volts available from the transformer is convenient.

[daan]

Well Bob, you make me hesitate again on it. Since the resistors are relatively cheap (about $1,50 each) the whole setup costs me about $20 dollars, and is worth the try.
Just for the fun of it... I'll post pictures etc of the building, and I'm very curious how it will work out.
The best thing about it is that I have an ancient BING resistormachine, which works with a 200 watts lightbulb, converting the 220 volt back to 42 volts under load. It also works with resistors for speed control.. It's amazingly dangerous to use so I'll stick with 24 volts and resistors for my try-out.

[lionelsoni]

Daan, your Bing machine sounds like it might be in the same category as my Lionel reducer. I bought mine just as a novelty. It was several years before I used it that one time. It amazes me that these were sold for use with kids' toy trains.

[daan]

The contactplates are finished. I found a heavy cracked safety door from a machine in the garbage at my work, Lexan and big enough for 2 contactplates.
With a bunch of screws and some wiring from my parts box it looks pretty cool.
The two contactplates are bolted together to get a solid unit. The copper bushing on the right down side will be the lever bearing.

The purple wires are for one contactplate, the red ones for the other. And Yes, they are numbered..[:D]
If this thing works, I'll drill a few small holes from the side (plates are 1/3 inch thick) and place LED's in it. It will be a spectaculair power unit...

[SPFan]

Nice Work!!! As for Variac prices, a 300watt version would cost about $100 new but I have found them at the surplus stores for $10-15.

Pete

[lionelsoni]

Wow.

[daan]

@Pete, I've found the prices for variacs too, new they cost about 99 euro's, so about the same as in the US. We don't have a lot of electric dump stores, and on the dutch ebay and marketsites, there where no available hits.

I've located a few old Lionels on ebay too, so if the resistorthing doesn't work, I have options open. (Though the shipping will be frightning high...)

I have 3 separate track's needed to be powered, the shuntyard will be operated with my 2 amps Maerklin transformer, but the 2 mainlines should have the variacs.
Since the contactplates turned out to be nice, I really am curious to fini***his project first and try if it works a bit.
At work I can make levers from stainless steel and poli***hem, the lever bar can be fitted with a black grip. It is not a technical elegant option, so I want to make this thing elegant in style..

[ben10ben]

Looks great! Reminds me a whole lot of an old Strowger telephone exchange switch.

As far as another option, you could always buy a US transformer, and power it through an inexpensive travel transformer(Like http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&product%5Fid=273-1413 from Radio Shack).

Lionel also made the RWS250 transformer, designed for anywhere from 220-250 volts, and 50 cycles or greater. It is rated for 100 watts.

[wrmcclellan]

Being an electrical/electronics engineer, this is getting VERY interesting.

Your contactor assembly is very nice. I had been thinking of how to help you solve your problem. What you are coming up with is much more fun.

Daan, if I understand your configuration correctly, here are some formulas to help you with the resistors:

Voltage (V) = Current (I) x Resistance (R); V=IR

Power (P) = Voltage(V) x Current(I); P=VI

Now substitute V=IR into P=VI and one gets P=IIR (I is squared - but not easy to show here unless one is familier with notations such as P=(I^2)R - but I digress, so back to the point)

For each 0.5 ohm resistor, you will get a varying voltage drop based upon the amount of current your load (the train) is drawing. Postwar Lionel motors draw in the 1.5 amperes (amps) to 2.5 amps per motor. An example is the 671 turbine that I measured a while back; 1.7 amps at 8.8 volts and 2.5 amps at 16.7 volts (includes headlight and e-unit).

For power calculations:

Using P=IIR for 1 amp and 0.5 ohms:

P = (1)(1)(0.5) = 0.5 watts

For 3 amps and 0.5 ohms;

P = (3)(3)(0.5) = 4.5 watts

For 5 amps (postwar passenger train with 2 bulbs per car) and 0.5 ohms;

P = (5)(5)(0.5) = 12.5 watts

For voltage drop calculations:

Using V=IR for 1 amp and 0.5 ohms:

V = (1)(0.5) = 0.5 volts drop per resistor

For 3 amps and 0.5 ohms;

V = (3)(0.5) = 1.5 volts drop per resistor

For 5 amps (postwar passenger train with 2 bulbs per car) and 0.5 ohms;

V = (5)(0.5) = 2.5 volts drop per resistor

The current through the resistors (which are in series) will always be the same as the load (the train).

As stated above in an earlier post, most engineers derate a part by at least 50% which means you need a resistor wattage rating at least twice the anticipated maximum wattage that it will carry. For this application you will want to use wire-wound ceramic resistors. Be aware that at 50% load (watts) they will get VERY hot. In some cases hot enough to peal the skin off of you. These resistors can heat up to several hundred degrees (250 C) as designed, so you need to shield the resistors and provide airflow of some sort of case for safety reasons. A fan may be useful. Here in the US, these resistors sell for under $1 each (www.digikey.com; P/N 1.0W-10-ND for a 1 ohm 10 watt)

There are also aluminum housed chassis mount resistors that may be more appropriate (www.digikey.com; P/N TMC25-1.0-ND for a 1 ohm 25 watt). These types sell for about $4.50 each in the US.

I initially used Digikey above because it was handy. Their mail catalog did not show a 0.50 or 0.51 ohm value (which are standard values for this type of part), but they may have it anyway.

I also found www.mouser.com has these. Aluminum Chassis mount 25 watt, 0.5 ohm, is P/N 71-RH25-0.5. The ceramic wirewound 10W, 0.5 ohm, is P/N 280-CR10-0.5.

I eagerly await the next installment of this saga...

Regards

[daan]

@wrmcclellan; I used only the P=VxI (Watt= Volts x Amps) and the R=V/I (Resistance is voltage / amps). With that I made a small list and my guess is that 0,5 Ohms per contact is sufficient to have not a too hogh voltage drop per contact with my 2353 F3 with 2 lit waggons, probably around 4 amps, getting the 2 volts drop and 8 watt needed power consumption. With the 2 amps other trains I think will draw the result will be 1 volt per contact.
I made 16 steps, so 15 resistors will make the speedregulator. Resulting in a max. voltage drop of 15x2 = 30 volts at 4 amps, which is much too high, but for the 2 or less amp trains the 15x1=15 volts, so starting speed will be 5 volts.
The both scales should be wired on 1 set of resistors of 17watts each, because I only have one F3 and all the others are less powerconsuming.
The resistors will be mounted on a 4mm thick iron coolingplate, even with fan if necessary, so I'm not afraid of frying something.
It should be somewhere where I can't incedentely touch it however.[:D]

I also think its a nice project, since real trains also have huge resistors to regulate speed. The only thing I can't reproduce is the series / series-parallel / parallel switch real engines have. they use thet to reduce voltage on the motors when starting. If the train starts, all motors are in series, voltage will be regulated from max resistance trough the resistors to min. Then the motors will be in half series/ half parallel mode, meaning 2 motors are in series, and 2 series of 2 motors are parallel. The same resitorsettings will be operated and then it switches to parallel mode, all 4 motors being parallel and having full voltage. The resistors will be shut off after powering the train to full speed.
This setting is used in the elder 1200 series of the dutch railways, being build in America and being a copy of the EP5, only with 6 axles.

[wrmcclellan]

Daan,

As I said bfore, this is quite interesting and I look forward to some photos of your final assembly and a report on how well it works.

I was also interested in your comment about full speed and the resistors will be cut out at that point. If you are still talking about using the 24 vac transformer, you definitely like full speed! I assume you use something like 0-310 "really wide" radius track or your full speed runs runs are very short!

Regards

[Anonymous]

He he...full speed...that would be an Adam's Family moment on my track!

[Back2Trains]

If the set-up you are building doesn't do what you want, a KW is an excellent choice, lots of power and versitility for the price. I don't replace the whistle diodes unless they are bad.
Jim

[daan]

@Jim, I've been browsing E-bay already and found some nice transformers for reasonable prices.. A KW costs about 125 euro's for a 110 volt version on a swap meet in Holland. the ZW will do about 250-275 euro's. I found someone having both, so I've to get money apart if this thing I'm building doesn't work.
In the mean time I'm trying to find a good powerpickup. I tried with rollerbearing balls which are sprung, they will make a semifixed position on each resistor position, since the ball will click into the inbus holes in the metalscrews. The problem is that I want to make it continuous suplying power to prevent the e-position problem.
The other possibility is 2 cupper blocks on individual leaf springs, but it is very hard to find a material which will follow the "ups and downs" in the contactplate sufficiently.
I guess the rollergearing ball type has the best options... It's less straight forward then the contactplates, and it should look nice too, so it needs a bit of trying out and tinkering 'till I found a good working setup..

[wrmcclellan]

Daan,

Try just the leaf spring for the sliding contact. Something similar to the sliding pickup shoes on older Lionel locos.

Regards

[lionelsoni]

Dann, how about two balls, spaced apart by half the spacing of the contact screws? It looks like one might always be over a screw head.

One thing you might watch out for: If you use ball bearings, be sure to insulate or shunt the spring that holds them against the contacts, so that it is not destroyed by fault currents. For example, use an insulating tube and put a follower between the spring and the ball with the connecting wire soldered to it, so that the current cannot flow through the spring.

[FJ and G]

Daan,

Good stuff! A very enjoyable read

[jkerklo]

Okay, I will make a rather wild suggestion.

I take a layout to local TCA shows that runs as many as ten trains on a 10 x 18' layout. It was powered by 4 KW's, 4 ZW's, and a Mainline Industries "The MAX." The panel with all the transformers weighed 260 Lbs. (118 kilos for daan).

There were two problems with this: the panel was way too heavy to transport, and the output voltage on the train transformers varied too much as trains passed from block to block. (Train transformers are poor, electrically. Output voltage can vary as much as 4 volts from no-load to full load.)

To save my back, I had a local company make some custom-wound transformers for me with ten taps from 12 to 18 volts @ 28 amps (that's 500 OUTPUT watts). Voltage drop is less than one volt from no-load to 28 amps!

They cost about $100. (for just the transformer, no box, breakers, etc.)

The transformers worked beyond my wildest dreams. The revised powering will be at the RMD/TCA show in Denver the weekend after Thanksgiving. (www.rmdtoytrains.org)

I had planned to use two, but, because of serious overdesign on my part, one will power the whole layout. Weight is about 15 lbs (7 kilos).

Now, instead of the resistors, hook your taps to such a transformer. One of the transformers I described will handle ten train circuits, or more. You should be able to find a custom transformer house in Holland. Inquire, you never know. For the layout you have described, a smaller transformer could be used.

I think the e-unit problem can be solved by connecting power resistors, a few ohms, from the 18 volt tap to each train circuit. If you are interested, I will give this a try.

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

[daan]

@ John, now that's a real bright idea!!!![tup] Powering the track with a switchable resistor, OR BETTER, always leave the trackpower on via an output on the end of the resistors. Even if I use only one contact on my lever on the contactplate the e-unit will hang on to it's direction and a very difficult way to get 2 contacts near to eachother is elliminated within an eyewink. This really is a neat one.
If I want to switch directions I can press a button taking of the power via the resistors off the track, and if the train stops for a long time, power can be shut off that way too.

The power pickups will be made of a rollerbearing ball, wich will be fitted in a fastener used to connect small pressure pipes. If I drill a hole for the spring in it, leave the clampring and flatten the endnut a bit a 7mm big ball will fit in and has a sprung travel of 4mm, which is sufficient. (I allready made a try out) The lever will be stainless steel with a hole in it to hold the pickup unit. On the bottom end of the lever I'll make a powertakeoff with a fixed wire. The levers will be isolated with nylon slide bearings on the axle.
The lever transports the electricity from the powerpickup to the connection to the rails. Since it's only +/- 20 volts max traveling through it, it won't harm anyone when touching it..
Now only building them is the remaining part..

[daan]

Just a simple obstacle though... If I'm going to use one set of resistors, both tracks will be connected by the powerline on the end of the resistors... It can be solved though by using 2 sets, one for each powertakeoff or hang a capicitor for AC over the resistors which will keep the e-unit stuck when switching voltage..
I also thought about a transformer with multiple steps, like with 2 volts up to 20. They are actually on sale, but they require a difficult connection way for each voltage (pin 23 with 7 for 2 volts, 23 with 5 for 4 volts, 10 with 23 for 6 volts etc.) This requires a very difficult contactplate with 2 contacts on each step of the plate and an even more challenging cablemess behind it.
I never thought of letting a trafo being made to my whishes, I'm going to get some prices in on that too..

[jkerklo]

The e-unit problem arises with a multi-tap transformer because you don't want to bridge the tap-contacts, which would short one winding of the transformer. Not good.
So the contacts have to be spaced so the "arm" leaves one contact before it contacts another. The resistor would be connected to one tap and sized so that it keeps e-units from dropping out, yet does not unduly load the transformer. A little measurement and experimentation is needed.

dann, you don't need the e-unit resistor, since shorting adjacent contacts in your arrangement wouldn't cause a particular transformer problem. If I have your design right, you would just be bridging a resistor, so no harm done.

However, if your mechanical design does not lend itself to easy contact bridging, then the e-unit resistor would still keep the trains running in the same direction.

E-units will hold at about 8 volts, drawing less than 1/2 amp.
However, the train motor is also connected. Say a train motor draws 1.5 amps at 8 volts. So, 24 volts - 8 volts is 16 volts. We want a resistor that will drop 16 volts at 2 amps. So, an 8 ohm resistor, 32 watts. Note that the e-unit resistor is ALWAYS supplying some power. How much will depend on which tap is connected.

I would start with the 8 ohm resistor, then decrease it if the e-units drop out. Actually, I would probably start with something like 12 ohms and measure the drops. Reducing the resistor only if e-units dropped out.

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

[jkerklo]

I don't understand the multi-tap tranformer you mentioned. With mine, only taps are brought out from the winding, not separate windings. I have ten taps, each .4 volts more than the previous. Each tap would go to one of the posts on your switch-plate.

Bringing out the windings separately is possible, but your right, it takes a complicated (binary) switching arrangement. An appropriate switch would cost too much.

Don't try a capacitor. AC capacitors do not do the same thing as with DC. Energy is not stored with AC. AC capacitors filter and provide a phase shift, not energy storage.

Just use a separate e-unit resistor for each track should do the trick.

Note, also, that your resistor technique for voltage control solves the problem of the engine center-rail pickup shorting two transformers when it crosses a track-block boundary. Much less of a short circuit.

Talk to a custom-transformer maker, but if you have difficulty in communicating, I would be happy to send a engineering "template" for giving them the parameters for what you need. Can you talk "electrical engineer?" Your work looks like you know what you are doing.

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

[jkerklo]

I think I just understood part of your eariler post.

I don't think one set of resistors will work for two train circuits. That is, if there are two trains. Each would affect the other.

If there is only one train, then one set of resistors will serve for two (or more) track blocks.

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

[lionelsoni]

Daan, you have twice referred to a "rollerbearing ball". Roller bearings use tapered cylinders and ball bearings use spheres. I'm guessing that you are talking about the latter, but not sure. Please consider my advice above about keeping current from flowing in the spring, which I am imagining is a coil compression spring behind a ball.

I'm worried that the e-unit resistor will provide such a variation in voltage with the variety in motor current and train lighting that some trains will not quite stop and others' e-units will drop out. What is the spacing between the screwheads on your contact plate? Is it possible to relieve the plastic between them so that the ball will roll directly from one screwhead to the next, without touching the plastic? That would solve the e-unit problem. If the gap is too wide for the 7-millimeter ball, you might get this to work with a somewhat larger one.

If all else fails, let me know and I will send you a Lionel rheostat...;-)