DC Voltages on Bachmann Locos

[Signalman]

G'day from Down Under,
I was wondering if you could help me. I am the proud owner of the following Bachmann loco's, and other things.
Item No. 60301, HO GP40 Diesel Loco - DCC Equipped, "Union Pacific"
Item No. 60348, HO GP40 Diesel Loco - DCC Equipped, "Alaska"
Item No. 50701, HO USRA 0-6-0 W/Smoke, Vanderbilt Tender, "Union Pacific"
Item No. 16947, HO Maintenance Vehicles - Self Propelled Ballast Regulator
Item No. 46202, Gandy Dancer.

At the moment I will be running them on my layout using DC. I will eventually go to DCC, but not just at the moment.

My question is this. What would the maximum DC voltage to apply to these items.

Many thanks.

Col Paton.

[Woody Elmore]

12 volts.

[Hunt]

...
My question is this. What would the maximum DC voltage to apply to these items.
...

Why the question? Are you planning on using a DC power supply not designed to be used with model railroad?

You are OK as long as the true DC voltage is 12 to 16 with 12V at the motor being ideal. 

[Signalman]

I design and build my own power supplies.

[Hunt]

I design and build my own power supplies.

That be the case, maximum power pack output 30VA with the amperes no more than 1.5.

Consider a used MRC Tech 3 Model 9500 power pack and use your time on something else. 

[Jim Banner]

Building the power supply is only half the solution.  The other half is controlling it.  Your trains will run on pure dc or full wave rectified ac and a lot of other wave shapes in between.  Some wave shapes produce smoother operation at low speed and others produce less heating at higher speeds.  Mechanically superior locomotives will run smoothly throughout the speed range on simple wave shapes, including on pure dc.  Whatever wave shape your trains need, it will have to be controlled in some way - voltage control and duty cycle control being the most common.  Variable pulse width control is a very popular form of duty cycle control because of its simplicity but pulse repetition rate with a fixed pulse width matched to each motor gives the absolute best low speed control. 

No matter what how you control the power, your control circuit must be able to limit the power to a safe level.  That is, a level which is safe for your locomotives, safe for your power supply and safe for your control circuit.  Short circuits occur far too often to make fuses practical and circuit breakers are usually too slow with all but the very lowest powered supplies.  That leaves electronic protection, usually some form of current limiting with or without current fold back.  Because brief short circuits can occur at any time, including when the control circuit is operating near maximum capacity, fold back is not the usual choice.  In that case, auxiliary circuitry to monitor circuit temperature and ensure shutdown in case of thermal overload resulting from sustained current overload is a must.  One family of integrated circuits that includes voltage control, current limiting and thermal shutdown is adjustable voltage regulators such as the LM317.  While often considered for voltage control type circuits, they can be used as the output stage in duty cycle type circuits as well.

One quirk with these ICs is their minimum output, typically 1.2 to 1.5 volts.  This is ideal if you wish to use Chubb type block occupancy detectors later on, but must be subtracted for train control.  A pair of diodes does the job, but the voltage must be included as part of the total overhead voltage when designing your power supply.  The IC itself contributes another 2 volts or so giving a minimum power supply voltage of about 16 volts dc to guarantee delivery of at least 12.0 volts to the tracks.  In practice, 18 volts is a better choice as it allows for losing a couple of volts along the rails and through various contacts on the way to the motor.  Going higher just puts more stress on the ICs and requires larger heat sinks to delay thermal shut down.   

[r.cprmier]

WHile on the subject, I am building a coal tower similar to the New Haven tower at Cedar Hill in New Haven, Ct.  I am lighting the two coaling bays under the coal bunker.  What I have found is that using yelloglo type LEDs with about 470Ohm resistors give me a pretty convincing light underneath-not too bright or dim, as I am trying to simulate a standard reflector lampshade that would of course be covered with a liberal dusting of coaldust.  I have used a MRC Tech 11 as a power supply.  the volt reading I got was 4VDC.  It still seemed pretty bright and turning it down produced no significant results, as I suspected it wouldn't.  Can you give me some thoughts about what to do to lower the intensity; like maybe running straight AC into the circuit to get some different response.  I will eventually have about twelve light "fixtures" on this rig; most used for general yard illumination.  I scratchbuilt the lights-again similar to the prototype I used as a model, so obviously, I will be using a better power supply.

[richG]

You might be interested in knowing that DCC equipped locos have 0 to 12 volts pulse power to the motors when controlled by a DCC controller.

Rich

[Hunt]

WHile on the subject, I am building a coal tower similar to the New Haven tower at Cedar Hill in New Haven, Ct.  I am lighting the two coaling bays under the coal bunker.  What I have found is that using yelloglo type LEDs with about 470Ohm resistors give me a pretty convincing light underneath-not too bright or dim, as I am trying to simulate a standard reflector lampshade that would of course be covered with a liberal dusting of coaldust.  I have used a MRC Tech 11 as a power supply.  the volt reading I got was 4VDC.  It still seemed pretty bright and turning it down produced no significant results, as I suspected it wouldn't.  Can you give me some thoughts about what to do to lower the intensity; like maybe running straight AC into the circuit to get some different response.  I will eventually have about twelve light "fixtures" on this rig; most used for general yard illumination.  I scratchbuilt the lights-again similar to the prototype I used as a model, so obviously, I will be using a better power supply.

Hi Rich,
(I did not see your off topic question buried in this thread. Someone called it to my attention.)

LED brightness control is by type of LED selection and control of voltage to the LED.

LED --  use DC not AC power makes things easier.

The more commonly found LED to simulate incandescent lamp is the Golden-White, Sunny-White or Yeloglo.  An issue is different type or make LED require slightly different voltages. So it is hard to control brightness of a mix of LEDs in a circuit.

Hopefully you have a LED knowledgeable electronic supplier nearby with a good selection of different types of LEDs. Ask about a  LED brightness control module if using the DC output of a model train power pack with speed control does not meet your requirement.

[Jim Banner]

  Can you give me some thoughts about what to do to lower the intensity

First off, stick with dc.  Excessive reverse voltage from the alternate half cycles in ac can reduce the life of an LED.  Secondly, reducing the voltage is only one way to dim LEDs.  The other way is to increase the resistance.  If you wire a 10,000 ohm potentiometer in series with your 470 ohm resistor, you will be able to adjust the resistance until you get the brightness you want.  You can then use an ohmmeter to measure the total series resistance and replace the pot + 470 ohms with the nearest standard resistor.  When wiring the potentiometer, use the center terminal and one end terminal.  Technically, this use makes it a rheostat but if you go to your electronics supplier and ask for a rheostat, all you may get is a blank stare.

LEDs are funny things to deal with if you are used to normal incandescent bulbs.  With LEDs, the light output is directly proportional to the current through them.  So if you reduce the current to one tenth, the light output reduces to one tenth.  With the way the human eye works, this means it appears one half as bright.  This makes LEDs very useful as headlight for locomotives because they have relatively little change in brightness over the operating voltage.  Locomotives may start as low as 1.5 volts.  And we may run them up to about 15 volts.  So over the whole operating range of voltages, the headlight brightness only doubles.

With small incandescent lamps, the light output varies as the 9th power of the voltage ratio.  This means if you reduce the voltage to one tenth, the light output is reduced to one billionth.  The apparent brightness is reduced to about one five hundredth, or at least it would be if your eye could respond to light that dim.  For most practical purposes, the light has gone out.  This makes voltage control of incandescent lamps quite practical but does not make them great headlights for locomotives.

All of this is further complicated by the fact that our eyes respond to the peak brightness of varying light sources.  So if you drive an LED with 4 volts dc that is derived from full wave rectified ac, our eye will see the LED as if it was powered by 5.7 volts of pure dc.  The incandescent lamp, on the other hand, cannot respond fast enough to follow the variations in the rectified ac so we see it as if it were driven by the average voltage which is 3.6 volts.

[Jim Banner]

You might be interested in knowing that DCC equipped locos have 0 to 12 volts pulse power to the motors when controlled by a DCC controller.

Rich

Right on.  Normally, this is fixed pulse width with variable pulse repetition rate.  The pulses have fast rises and falls between 0 and about 12 volts so that the output transistors in the decoders can be very small.  The motor receives an average voltage between 0 and about 12 volts, depending on the duty cycle of the pulses (on time divided by on time plus off time.)  With many decoders, you also have the option of controlling the width of those pulses using CV9.  You can use CV9 to tune the pulse width to fit the motor, making it just wide enough that each pulse will just kick the motor past any cogging it may have.  This approach is not particularly popular because it tends to be noisy and because motor cogging is not the problem it used to be in the bad old days before skewed armatures.  But it can still be useful if you are installing a decoder in an old locomotive.  Adjusting CV9 can also be useful if you do not want to remove the capacitors in Bachmann and other locomotives that include RFI suppression.  Otherwise, you will probably leave CV9 set for ultrasonic ("silent") operation. 

[r.cprmier]

Hunt, Jim;
Thank you for the feedback.  The thought I had behind using AC was that the half-cycle that would be produced by using the LED would drop intensity; but I forgot about the effect the reverse cycle would have on everything.  I am using 470 Ohm resistors, and I think that  perhaps adding a second resistor-say in the 200 ohm or so range in series to the existing stuff, (only because I am too lazy to take the assemblage apart) which would give me about 700 or so, and going with the 470 Ohm on the yard lights, might achieve what I am trying to accomplish.
THanks again, guys.

[ta152h0]

darn, that question brings back memories of my younger days, about a thousand years ago. I built my own power supplies way back then and controlled the 120 VAC side of things and used  a 5U4 rectifier tube to get 12 vDC. It was easy to vary the 120 V input. Kept my room warn also.