Fullscreen HTML5Med Res (??MB)HTML4

MK-ARO '111 RO'I*I'LE FOR MODEL ROADS Want to build a walk-around throttle for your model railroad layout? This one offers a host of features including pulse power, inertia (momentum), braking and full overload protection. By LEO SIMPSON & JOHN CLARKE 32 SILICON Cll/1' Over the years we have seen a number of solid state throttles for model railroad layouts but none matches the circuit presented here for features and versatility. For example, consider the walk-around throttle feature. These days, few model railroad enthusiasts want to be tied to a fixed console in order to operate their trains. They want a walk-around throttle so that they can observe the train closely while they are controlling it. The walk-around throttle concept is simple - just a small box on the end of a lead which has a knob to vary the speed and perhaps a couple of switches to provide direction (forward/reverse) and braking. As such, it is a pretty simple concept but what if you have a large layout? You don't want to have a very long lead otherwise it will get tangled and you will trip over it. No, you want to be able to plug the handheld controller into various sockets around the layout as the train moves over the tracks. And when you disconnect the controller from one socket in order to move to the next, you don't want the train to suddenly speed up or stop; the train should continue at its pre-determined speed; ie, the controller should have memory. Having made such a point about the walk-around concept, as you might expect, our circuit has this desirable feature along with those listed in the accompanying panel. Let's talk about some of these features. Main Features • Pulse power for smooth and reliable low speed operation. • Monitoring of motor back-EMF for excellent speed regulation . • Adequate power for double and triple heading of locos. • Inertia (momentum) so that the model acts as though it had the sizable inertia of a real train . • Full overload protection in- eluding visible and audible overload indicators (short circuit duration: one minute) . • Power and track/direction LED indicators. • Provision for maximum output voltage adjustment (to suit Z scale). • Fixed 1 2V DC output for accessories. ,.., ,.., Vo I \ \ I \ I \ I \ I \ I ' \ \ lb) Pulse power Pulse power in model train controllers is not new although to most most model train enthusiasts pulse power means something different to what is used in our circuit. We'll set the record straight on this point before going any further. To do so, we need to briefly review the current state of the art. Most basic model train supplies consist of a low voltage transformer feeding a bridge rectifier to produce unfiltered DC as shown in Fig. l(a). This unfiltered DC voltage is then varied by a simple transistor or resistor controller to set the train speed. Fig.l(b) depicts the waveform when the controller is set for a low train speed. Now the problem with this basic approach is that when the controller is set for low speed, the output voltage is low, as you'd expect. This means that when the loco wheels and track are not scupulously clean (they never are), the train may have trouble starting or may run jerkily. Designers of commercial model train controllers have taken a number of approaches to improve the situation and they all involve in- I I I I I ,.., ,.., ,,, \ I I I I I \ I I I I I I \ ,.. ' \ / I I I I I I ,.. .... \ I (d) Fig.1: most controllers operate by varying the level of an unfiltered rectified DC waveform as shown in (a), (b) & (c). An SCR controller (d) chops the fullwave rectified DC but best results come from a pulse power controller (e). creasing the peak voltage applied to the track while the average voltage for low speed settings remains low. The simplest and crudest of these approaches is to use half-wave rectified DC, as shown in Fig. l( c ). This gives a higher peak voltage for a given low speed setting but has the disadvantage that it makes the loco motors growl, particularly at low speed settings. Now this crude approach is often referred to as "pulse power" and, in the truest sense of the word, so it is but it is crude nonetheless. Some controllers with this design have a refinement(?) whereby the output voltage waveform makes a transition from halfwave rectification to full wave rectification as the speed setting is increased. It's still crude though. Another approach is to use a silicon controlled rectifier which chops the full wave rectified DC waveform to provide speed control. This approach is better but still has the disadvantage that, at low speed settings, the track voltage is still relatively low - see Fig.l(d). Then there's the way our circuit does it: the proper way, as shown in /\i'HII, Hl[lfl 33 +12V +12V 100k 100k VT 100k VP SPEED A.Iv\ OSCILLATOR Fig;2: basic pulse power control circuit. IC1d is wired as a Schmitt trigger oscillator while IC2a is wired as a comparator. The output (Vp) is a 200Hz pulse waveform with pulse width determined by the setting of the speed control pot. VT /'( /\ I\ "/1/K} I . nn XT/\/\ VP-VP~...____.~....______.~.___r (a) HIGH VOLTAGE (b) LOW VOLTAGE Fig.3: how the output of IC2a varies with the setting of the speed control pot. At higher speed settings, the output pulses are longer. Fig.1( e ). This is essentially the same method used in switchmode power supplies whereby a relatively high DC voltage is varied by rapidly switching it on and off. This means that the peak voltage across the track is always the same, regardless of the speed setting. Varying the width of the pulses applied to the loco varies the speed see Fig.l(e). In our circuit, the track voltage is about 17 or 18 volts peak. This relatively high voltage is better able to overcome poor contact resistance between the loco wheels and track and so gives much better low speed running and starting. the back-EMF (EMF stands for electromotive force, another term for voltage) is proportional to the motor speed. So the circuit monitors the backEMF of the motor and if this voltage drops, as it tends to when the loco starts lugging up a slope or whatever, the circuit actually increases its output voltage to help maintain the selected speed. We haven't overdone this feature though, so that a loco will still tend to slow down as it is loaded, but the speed regulation is certainly better than if this feedback was not included. Speed regulation Real trains have inertia, hundreds or thousands of tonnes of it. When the driver opens the throttles on his loco(s) very little happens at first. It may take many kilometres for the train to get up to operating speed and similarly, when he applies the brakes, the speed does not slacken very rapidly. By contrast, model trains have no inertia at all and when full power is Another worthwhile feature of our circuit is the speed regulation. This helps the loco to maintain its speed even though the gradient may change or the load may change, as in shunting. What happens is that the circuit monitors the back-EMF of the motor. This is the voltage the motor generates to oppose the current through it and, as it happens, 34 SILICON CHIP Inertia or momentum applied to the track, they accelerate like startled rabbits. Similarly, if power is abruptly removed from the track, they skid to a stop, which is hardly what you'd call "realistic operation". For this reason, the Railpower controller incorporates inertia circuitry so that the track voltage builds up slowly when the speed control is wound full on and drops slowly when the brake is applied. It makes the trains look a whole lot more realistic. Overload protection All model train controllers need some sort of short circuit protection because short circuits can occur quite frequently. Whether it's because a loco is derailed, or because points are faulty or because someone deliberately shorts out the rails with a screwdriver, overloads do occur. The Railpower controller has "foldback" short circuit protection (we'll explain that later) plus a LED indicator and a buzzer to indicate that an overload has occurred. Thus, it will indicate even when momentary shorts occur, as can happen when a loco is crossing points. Power output While model loco motors rarely pull much more than one amp, some model locos can pull considerably more than this, depending on whether they have smoke generators, sound systems and lighting. So if you want to double or triplehead locos or have lots of track lighting, you'll want plenty of amps. The Railpower controller has plenty, around 4 amps or so with the specified 60VA transformer. In testing the power output we ran as many as six HO locomotives simultaneously from the Railpower. Most of these locos also had internal lighting so it really did amount to a considerable load. The Railpower handled it without a murmur and without even getting warm. Current output is not the only important parameter though. While most model locos are specified to operate with a maximum of 12 volts DC, some manufacturers specify less voltage and this should not be PARTS LIST 1 PCB, code SC9-1-488, 117 x 125mm 1 50 x 20mm piece of Veroboard 1 Scotchcal label, 79 x 50mm 1 plastic case, 83 x 54 x 30mm 1 piece of aluminium, 80 x 60mm x 0 .6mm 1 1 2V 60VA transformer 2 8-way PCB terminals 1 6-way PCB terminal 2 SPDT switches 1 knob 1 grommet 1 6-way cable 1 12V buzzer Semiconductors 2 BD650 PNP power Darlington transistors 2 BD649 NPN power Darlington transistors 3 BC54 7 NPN transistors 1 BC558 PNP transistor 1 7812 12V 3-terminal regulator 4 1 N5404 3A diodes 5 1 N4148, 1 N914 diodes 2 red LEDs 1 bi-colour LED 2 LM324 quad op amps 1 4093 quad Schmitt NAND gates 1 4049 hex inverter buffers Capacitors 2 2200µF 25VW PC electrolytics 1 4 7µF 1 6VW PC electrolytic 1 10µF 16VW PC electrolytic 1 4. 7 µF 1 6VW PC electrolytic VP LOGIC IC3, IC4 FORWARD REVERSE o.rn CURRENT SENSE .,. Fig.4: the H-pack output circuit. To make the motor go in one direction, Ql and Q4 are turned on while Q2 and Q3 are kept off. For the reverse direction, Q2 and Q3 are turned on and Ql and Q4 are turned off. exceeded, to safeguard their motors. For example, Marklin Zscale (1:220) locos are specified for a maximum of 8 volts DC. The Railpower controller has provision to adjust for these specified maximum voltages. Operating principles The complete circuit shown in Fig.5 is pretty daunting to try and comprehend at first so let's have a look at the core of the circuit which is shown in Fig.2. This depicts the two key op amps which provide the pulse width modulation. ICld is wired as a Schmitt trigger oscillator while IC2a is wired as a comparator. ICld oscillates by the following action. When power is first applied Cl has no charge and the output of ICld is high. Consequently, Cl is charged via Rl until the voltage at pin 6 exceeds the voltage at pin 5. This causes the output at pin 7 to switch low and so Cl is now discharged via Rl. So Cl is alternately charged and 1 2.2µF 25VW PC electrolytic 1 2. 2µF 1 6VW PC electrolytic 2 0 .1µF metallised polyester (greencap) 1 0 .01 µF metallised polyester (greencap) Resistors (0.25W, 5%) 1 X 560k0, 1 X 220k0, 2 X 120k0, 5 x 1 OOkO, 1 x 27k0, 1 x 22k0, 2 x 1 5k0, 5 x 1 OkO, 1 x 8 .2k0, 2 x 5.6k0, 6 x 2.2k0, 6 x 1 kn, 1 x 1 oon, 1 x o. rn 5W, 1 x 1 MO miniature vertical trimpot, 1 x 220k0 miniature trimpot, 2 x 1OOkO miniature trimpots, 1 x 1 Okn linear potentiometer Miscellaneous Solder, tinned copper wire, screws, nuts, etc. discharged via Rl and the resulting waveform is a triangle (sawtooth) waveform shown as Vt in Fig.3. This waveform has an amplitude of between two and three volts peakto-peak and a frequency of about 200Hz. This triangular waveform is applied to pin 13 of IC2a which compares it with the speed voltage Vs fed to pin 12. Since IC2a is wired as a comparator its output can only be high or low, so when Vt is above Vs, the output will be low and when Vt is below Vs, the output will be high. The interaction of Vt and Vs via IC2a is shown in Fig.3. Fig.3(a) shows that when Vs is set for high speed, the output from IC2a is a series of fairly wide pulses. These give an average DC voltage across the track which is quite high. Similarly, in Fig.3(b), when Vs is set for low speed, the output from IC2a is Vp, a series of narrow pulses which have quite a low average DC voltage. H-pack output So the pulse waveform Vp is eventually transmitted to the track and loco motor via IC3, IC4 and the transistors Ql to Q6, shown on the circuit diagram Fig.5. Again, comprehending how all these devices work together is not easy so we have reproduced the output circuit in Fig.4. APRIL 1988 35 Fig.5 (right): the complete circuit diagram. All the IC and transistor numbers correspond to those shown in Figs.2 & 4. IC2c and IC2d provide the foldback current protection while ICs 3 & 4 provide logic switching to the H-pack output stage. a TO-220 plastic encapsulation but have a collector current rating of 16 amps peak (8 amps DC). Main circuit Most of the parts are accommodated on a single PCB . The four output transistors and the 3-terminal regulator are bolted to aluminium heatsinks. This shows the four transistors, Ql to Q4, in an "H" configuration with the motor of the loco connected between the two sides of the "H". IC3 and IC4 are depicted as a logic block with three inputs, one for speed which is Vp, and two for direction (forward and reverse). Fig.4 is really quite a lot more complicated than it needs to be. Instead of using six transistors and two logic ICs, we could have made do with one small signal transistor, a power transistor and a heavy duty relay, which would have reversed the track voltage for the forward/reverse mode. But while the present circuit is complicated, it does have the advantage of being cheaper and more compact than the relay/transistor combination. It also has the advantage of having memory for the direction setting. This is necessary if the walk-around control is to be unplugged at any time. Nor is there anything essentially new in the H-configuration of Fig.4. It is commonly used in industrial circuits used for motor speed and direction control. To make the motor go in one direction, Ql and Q4 are turned on while Q2 and Q3 36 SILICON CIIII' are kept off. To reverse the motor, Q2 and Q3 are turned on while Ql and Q4 are turned off. Putting it another way, for the forward motor pirection, current passes through Ql and Q4; for reverse, current passes through Q2 and Q3. In practice, for the forward direction Q4 is turned on fully and Ql is turned rapidly on and off by the pulse waveform Vp, to give speed control. Similarly, for the reverse function, Q3 is turned on continuously and Q2 is modulated by the pulse waveform Vp to give speed control. Natty, huh? Q5 and Q6 are there solely to provide voltage level translation between the logic block, IC3 and IC4, and the output transistors. This is necessary because the logic circuitry runs from + 12V while the output transistors run from + 17V. Ql to Q4 are Darlington transistors which incorporate flyback diodes connected betwen their collectors and emitters. These diodes are necessary when driving inductive loads such as motors which will tend to generate spikes from their commutators and from the pulse waveform. The Darlingtons come in Now let us relate the circuits of Fig.2 and Fig.4 to the complete circuit of Fig.5 . The circuit of Fig.4 can be seen at the righthand side of the main circuit while ICld and IC2a are roughly in the centre of the circuit. Now have a look at ICla and IClb, at the lefthand side of the circuit. These two op amps are connected as voltage followers. Their function is to buffer and reproduce the voltage from the wipers of VRl and VR2. VRl sets the maximum voltage applied to the track. This is important, particularly for Z-gauge, as mentioned earlier. VR2 sets the minimum track voltage. This is necessary because all locos have some minimum voltage below which their motors will not run. So VRl and VR2 set the overall speed range which is provided by potentiometer VR3, connected between the outputs of ICla and IClb. Inertia The speed setting from the wiper of VR3 is fed via VR4 to the 47µ,F capacitor at the non-inverting input (pin 3) of IClc. VR4 and the 47µ,F capacitor provide the inertia feature, in the following way. Consider that the speed pot VR3 is wound up to maximum. Because of the resistance of VR4, the voltage at pin 3 of IClc does not rise immediately but gradually, as the 47 µ,F capacitor charges. If VR4 is set to its high resistance condition, the circuit gives maximum inertia. The voltage across the 4 7µ,F capacitor is buffered by voltage follower IClc which feeds IC2a, via pull-down diode DL So IClc and Dl provide the voltage Vs fed to IC2a, as shown in Fig.2 and Fig.3. IClc and the 47 µ,F capacitor also provide a "speed memory" in case the "'-l w 0:, 0:, co ..... ~ ] > • "I M2165 60VA OR EQUIVALENT N 240VAC 7 1~2V A .,. vtfJlo8oi :J1<0----."-tl MINIMUM MAXIMUM ADJUST :JI VR1 100k +12V ~ 1k .... . LM324 .,. .,. 100k 14 .,. REVERSE S2 dORWARO RUN ~A;;-HE~ UNIT - - g_ I 50 I I I I I 3 I 7 2200 '+ 25VW 220k .,. + SC9-1-488 .,. .011 , 100k 0.1 +12V TRIANGLE WAVEFORM VT RAIL POWER +17V .,. 47 16VW1 +12V L ________ J I4 I - 120k +9.BV -: ? ? +12V ..,. 0.lI 27k 10k I +12V +12V • +12V ?? EOc VIEWED FROM BELOW B MOTOR BACK EMF -1 .,.. BCE ~ + .,. GNO -~ITT FOLOBACK CONTROL D3 1N4148 \: ~< 01 B0650 .,. +12V 0.l !l 5W MOTOR OVERLOAD BUZZER +11v--+--------. Model Trains & Pulse Power Myths If you read model railroading magazines or talk to some model railroaders, "pulse power" has a bad reputation. There are claims that pulse power makes motors run hot ·and can lead to motors overheating and burning out. As with most myths, there is some technical basis for this belief but further investigation shows that it is not right. In permanent magnet motors, torque is proportional to the average current while the heat dissipated in the motor is proportional to the RMS value of the current. Based on this, the heat produced for a given speed setting will be higher for a pulse waveform than for pure DC. But, as we have already noted, most commercial train controllers hand-held walkaround throttle is unplugged. Back-EMF monitoring As already noted, the pulse voltage from IC2a is fed via logic circuits IC3 and IC4 to the H-pack output stage but let's ignore them for the moment. Instead, let's flick down to the back-EMF monitoring circuit provided by diodes D4, D5 and transistor QB. There is rather more to this part of the circuit than meets the eye. What it does is to monitor the voltage across the motor when the output circuit itself is providing no power. In other words, the speed monitoring circuit looks at the motor in between each pulse delivered by Darlington transistor Ql or Q2. How does it do it? Well, remember that for the forward motor direction Q4 is. continuously on while Q3 is off. This means that virtually the full voltage appearing across the motor appears at the collector of Q3. So the motor voltage is fed via D4 and a 2.2k0 resistor to the non-inverting input of IC2b (over on the lefthand side of the circuit). But D4 feeds the voltage down the 2.2k0 resistor all the time so it gets the pulse voltage as well as the motor back-EMF which is not what 38 SILI CO N Cllll' use pulse power of some sort. Very few use pure DC . In practice then , the difference in motor dissipation between unfiltered DC controllers and the Railpower design is small. The big danger of motors burning out is if the motor stalls due to a binding gear system. Under these conditions , you run the risk of burning out the motor if you apply full track voltage for more than a few seconds . Note that this applies to any model train controller, not just the Railpower. The risk is higher for motors in the smaller gauges such as N or Z-gauge . Pulse power is also reputed to cause motors to be noisier than with pure DC. This tends to be true partly because a controller such as the Railpower allows the loco to run at much lower speeds than would be possible with filtered or unfiltered DC across the track. At these much lower speeds, motor noise becomes much more significant; at higher speeds motor noise is drowned out by gear noise and wheel/rail noises . Noise is also dependent to some extent on the quality of the gear trains and can be amplified by locos of brass construction. It is · sometimes possible to adjust the , loco gear trains to minimise noise . With the majority of locos we have tested, the pulse frequency of 200Hz has been found to be close to optimum. The pulse frequency can be reduced by increasing the .01 µF capacitor at pin 6 or IC1 d. To halve the frequency , double the capacitor's value. we want. So every time a pulse is delivered by Q 1, the pulse waveform Vp also turns on QB. So the pulse voltage never gets to the input of IC2b. Pretty cunning that! Similarly, for the reverse direction, Q3 is always on and the full motor voltage appears at the collector of Q4 and is fed via D5 to the 2.2k0 resistor and thence to the input of ICZb. Again, whenever pulse voltage is present across the motor, QB is turned on, to shunt it to ground. So the voltage fed to IC2b truly represents the motor backEMF and therefore is an indication of the motor's speed. It is a train of pulses, because of the switching action of QB. Absolute pulse-power in the palm of your hand. The controls are speed, forward/reverse and run/stop with (adjustable) simulated inertia . Speed regulation IC2b is a non-inverting amplifier with a gain of 3.2, as set by its 220k0 and lOOkO feedback resistors. Its output is a pulse waveform which is filtered by a 22k0 resistor and 2.2µF capacitor. The smoothed DC voltage, representing the motor's actual speed, is fed to the reference input of ICld, the triangle waveform generator. This has the effect of raising the overall voltage level of the triangle waveform Vt, while its amplitude and frequency remain the same. So what happens if the back-EMF generated by the motor for a certain speed suddenly drops? The effect is to lower the overall voltage level of Vt, the triangle waveform. As can be seen from Fig.3, if Vt is lowered in level with respect to Vs, the pulses delivered by IC2a will be longer and so the power delivered to the motor will be increased and the desired speed will be restored. Overload protection Two op amps, IC2c and IC2d, pro- TO HAND HELD UNIT 0 12VAC INPUT BUZZER + LED2 Fig.6: parts placement diagram for the PCB. Be sure to use the correct part at each location and note that IC2 is oriented differently to the other ICs. VR1 and VR2 set the maximum and minimum track voltages. FROM MAIN BOARD Fig.7: this is the wiring diagram for the hand-held controller. The numbers on the leads correspond to the numbers on the terminal block at the top of Fig.6. VR4 and VR5 set the running and braking inertia. vide the short circuit protection and both of these are wired as comparators. The current passing through the motor is monitored by the o. rn 5W resistor connected to the commoned emitters of Q3 and Q4. The voltage developed across the resistor is fed via a 10k0 resistor to the inverting input, pin 2, of IC2c. The voltage at pin 2 is then compared with a reference voltage at pin 3, which is approximately 0.6 volts. Normally, the voltage at pin 2 will be well below 0.6 volts and so the output of IC2c will remain high, as will the output of IC2d. Therefore, operation of the controller continues as normal. When an excessive current flows through the controller output, a large peak voltage will be developed across the 0. rn sensing resistor and the voltage at pin 2 will rise above the threshold of comparator IC2c. This will cause the output to go low which then pulls pin 12 of IC2a low, via diode DZ. This has the effect of reducing the width of the output pulses and so the fault current is reduced. IC2c also turns on the overload LED to indicate the fault condition. IC2c's action in reducing the fault current tends to cause a "hunt" condition whereby as the current is reduced, the voltage at pin 2 reduces and so the controller again delivers the full pulse width. This causes the current to increase again and IC2c again switches on. This " oscillation" is slowed to some extent by the 0.1µ,F filter capacitor at pin 2 of IC2c, so that the action of IC2c is adequate to cope with short-term overloads and short circuits which may occur when a loco is crossing points. For longer term short circuits though, IC2d comes into play. This op amp monitors the output of IC2c via LED 2 (the overload indicator). When a long duration short circuit occurs, the capacitor at pin 5 is discharged so that its voltage is below the reference voltage at pin 6. This causes IC2d's output to go low which then also pulls pin 2 of IC2a low, via diode D3. So IC2c and IC2d together act to reduce the pulse width and thereby control the output current. IC2d thereby provides a "foldback" current limiting action. IC2d also drives Q7 which sounds the buzzer whenever a short circuit, or overload occurs. This very effectively draws your attention to any overloads, whether momentary or otherwise, so that any faults can be corrected. Just a small point of explanation here: the reference voltage at pin 3 of IC2c is 0.6V which may lead you to conclude that current limiting will occur for currents in excess of A PHIL 1988 39 IC2a, dpending on the setting of the flipflop. Thus, if Q4 is turned on continuously, pulse signals are fed via IC3a, inverter IC4a and transistor Q5, to turn Ql on and off at 200Hz. Similarly, if Q3 is turned on continuously, for the reverse condition, Vp signals are gated through IC3b, inverter IC4b, and transistor Q6, to turn on Q2 at the 200Hz rate. Power supply This view shows how everything fits together inside the hand-held controller unit. The 6-way cable must be securely anchored to prevent lead breakage. 6 amps peak (ie, 0.6V across the 0.10 sensing resistor). In practice though, the 0.lµF filter capacitor at pin 2 allows higher peak currents to pass before limiting occurs. Output Darlington transistors Ql to Q4 are fitted with small heatsinks which normally stay quite cool. If a short circuit is maintained across the track for any length of time though, the transistors will rapidly become very hot. They can withstand this condition for several minutes although the overload buzzer will be sounding stridently and the short should be corrected as soon as possible. Logic circuitry Now we come to the part of the circuit which looks quite tricky but isn't; if you have stuck with the description as far as this point you will have no trouble with the logic. IC3c and IC3d are the key to it all; they are coupled together as an RS flipflop which is controlled by the forward/reverse switch S2. When S2 is set to the forward condition it pulls pin 5 low (normally held high by a 10k0 resistor). This causes the output at pin 4 to go high while the output at pin 3 goes low. The flipflop will then remain in this condition until S2 is switched 40 S!U CON CIIII' The run and stop inertia adjustment pots (VR4 and VR5) are mounted on a small piece of Veroboard (see Fig.7). over to the reverse condition. When that occurs, pin 1 will be pulled low and the flipflop will change state. Pin 3 will now be high and pin 4 will be low. (If you want to better understand this type of flipflop, have a look at our series on Digital Electronics, in the February 1988 issue). The flipflop determines which output transistor remains on continuously; ie, Q3 or Q4. For the forward setting of S2 , pin 4 of IC3c will be high and pin 3 will be low. As a consequence, the output of inverters IC4c and IC4d will be low and Q3 will be off; the output of inverters IC4e and IC4f will be high and so Q4 will be on. IC3a and IC3b gate through the pulse waveform (Vp) signals from The power transformer is a 60V A multitap unit available from Jaycar (Cat. No. MM-2005) or Altronics (Cat. No. M-2165). It is connected to provide a 12V AC output which feeds a bridge rectifier and two 2200µF 25VW electrolytic capacitors. This produces smoothed but unregulated DC of about 17-18V. This is fed to the output stage (Q1-Q4) and to a 3-terminal regulator to produce a regulated + 12V supply which is fed to all the op amps and logic circuits. Methods of construction The Railpower controller can be built in several ways. Many modelling enthusiasts will prefer to build it into their main control console and thus will bury the printed circuit board under the layout. Others will want a self-contained unit with or without the walkaround throttle feature. Still others will want a bare-bones unit without a case but with the walk-around throttle. We have catered for all these possibilities. Only one printed circuit board is required, measuring 117 x 125mm (code SC9-1-488). This accommodates all components except for those in the handheld walk-around throttle. The board has a six-way insulated terminal block for connections to the handheld throttle and two eight-way connectors for the remainder of the connections. For those who want to get started, Fig.6 shows the parts layout on the PCB while Fig.7 shows the wiring details for the hand-held controller. Next month we will give full details of the construction of the Railpower controller in a number of versions. Kits for the project will be available shortly from Jaycar Electronics.