Fullscreen HTML5Med Res (??MB)HTML4

By Tim Blythman Decoder Base Station Using DCC Remote Controller DCC Booster DCC Base Station Image source: https://unsplash.com/photos/ a-toy-train-traveling-through-a-lush-green-forest-rxBE5UF-Dsk Following on from the DCC Decoder last month, the other main component needed to add DCC to a model railway is a DCC Base Station. It provides power and data to the tracks. It’s based on a Pico 2 microcontroller module connected to an LCD touchscreen, so it’s easy to customise. D igital Command Control (DCC) is a system to operate model railways in a more realistic fashion than previous techniques such as DC/analog voltage control. The latter involves using a controller to apply a voltage to the tracks, connected straight to the locomotive motor via wheel pickups, to allow control of speed and direction. To turn a conventional analog model railway into one using DCC requires specific equipment in the locomotives and for the controller. The DCC Decoder from last month can be installed in a model locomotive to enable DCC operation. That article also covered some of the background of DCC. The DCC Decoder receives power and commands from a DCC base station. It then controls the motor and lights in the locomotive according to those commands. The DCC base station thus takes the place of a controller in a DCC system. The DCC standards are maintained by the NMRA (US National Model Railroad Association) and our designs have been tested to work with commercial gear from brands such as DigiTrax, NCE, TCS and DCC Concepts. siliconchip.com.au On page 49, we have a detailed guide on working with DCC. It will focus on using our Decoder and Base Station, but much of it will also be applicable to commercially available devices. The Base Station There is quite a bit of variety in what might be expected from a DCC base station. Some, like the Complete Arduino DCC Controller (January 2020; siliconchip.au/Article/12220) rely on a computer running the JMRI (Java Model Railway Interface) software. JMRI can show layout maps, mimic panels, rolling stock rosters and can even automate operations. At the other end of the spectrum are simple base stations that are designed to allow simultaneous operation of a few locomotives and allow some amount of decoder programming; enough for someone converting to DCC for the first time. Features & Specifications 🛤 Designed for small HO/OO and N scale operations 🛤 Pluggable screw terminals for easy connection to tracks 🛤 PCB front panel to suit UB3 Jiffy box or on a custom control panel 🛤 DC jack for plugpack, or screw terminal power inputs 🛤 Controls for five locomotives including speed, direction and four functions each 🛤 Automatic current sensing with adjustable trip limit 🛤 128 speed steps 🛤 Uses a Raspberry Pi Pico 2 and 3.5in 480×320-pixel LCD touchscreen 🛤 Main track output: up to 10A 🛤 Supply/track voltage: 8V-22V (don’t exceed 17V with our DCC Decoder) 🛤 Programming track output: limited to 250mA Australia's electronics magazine January 2026  35 At a minimum, the base station needs to have a processor of some sort to process user input and encode the DCC data for output. The user input might include selecting a decoder address or a request to send programming data to the track, as well as the operation of locomotive controls. Some sort of driver is needed to generate the DCC signal for power and control. There are systems that even provide a means to connect extra user panels. Since DCC is intended to allow more than one train to operate, it makes sense to allow multiple users to have control over the data that is transmitted, more on this shortly. On a hardware level, a base station must be able to drive the DCC track voltage, which is an AC square wave about 12V-15V in amplitude (24V-30V peak-to-peak) with a frequency varying around 6kHz. There should also be some current sensing and circuit protection, since the output could easily be a few amps or more, and it isn’t too hard to accidentally short the tracks. Our Base Station is intended as a simple and economical way to try out the world of DCC, but it still offers many features. All the controls are based on a 3.5in LCD touchscreen showing one of five different ‘pages’, each of which can be allocated to a decoder address. There are also pages for performing DCC programming and configuring the Base Station itself. We have also designed a DCC Remote Controller unit that can connect to an expansion port on the Base Station. Each Remote Control can control three locomotives; multiple Remote Controls can be connected in daisy-­chain fashion to a Base Station. We will present the DCC Remote Controller add-on project next month. Hardware The hardware for the DCC Base Station is fairly simple, so let’s start by looking at the circuit in Fig.6. Modulation of the main DCC signal is handled by two BTN8962 half-bridge drivers, IC2 and IC3. These are the same drivers used in the Arduino DCC Controller, and can handle up to 30A, so should be robust in a circuit limited to 10A. Each chip drives one of the rails to either 15V or ground under the control of signals DCC1EN, DCC1A and DCC1B from the Pico 2 module. If DCC1EN is low, both INH pins are low and the drivers are disabled 36 Silicon Chip (high-impedance). When DCC1EN is high, the outputs of IC2 and IC3 follow the inputs DCC1A and DCC1B, respectively. We’ve provided 1kW series resistors to afford some protection to the Pico 2 in the event of a critical failure of the driver ICs. When their high-side drivers are active, IC2 and IC3 source current from their IS pins in proportion to the current flowing to the output; the ratio is approximately 1:10,000. Since only one driver is active at a time, we combine the currents through dual diode D2. This current develops a proportional voltage across the 1kW resistor at its cathode. The 10kW resistor and 100nF capacitor smooth out peaks due to the varying DCC signal, and the voltage is sampled by one of the ADC (analog-to-digital converter) pins of the Pico 2 via GP27 so it can monitor the track load current. The outputs of IC2 and IC3 are connected to CON1, a pluggable screw terminal. Bicolour LED1 and its dropping resistor provide a visual indication of the voltage output at CON1. There is also a 100kW resistor that pulls the INH pins of IC2 and IC3 low if they are not otherwise driven. The driver for the programming output does not need to be as powerful, so we have used the same DRV8231 motor driver (IC1) as in the Decoder. The standards indicate that a programming output should be limited to supplying 250mA; the 3.7A-rated DRV8231 will handle that with ease. IC1’s control signals connect on lines DCC2A and DCC2B via 1kW series resistors. Its Vref pin is fed from 3.3V and with a 0.1W resistor on the ISEN pin, the current limit is set to 3.3A. The voltage across the 0.1W resistor is monitored by an ADC pin on the Pico 2 to allow the programming current to be measured. We apply the 250mA limit through a pair of 2W 33W resistors on the driver outputs, which can handle short-­ circuit conditions continuously with up to 16V at the input. The programming output is intermittent and only active for seconds at a time under direct supervision of a user, so we think this will be adequate. The resistors provide this soft limiting to prevent the DRV8231 from shutting down its outputs, since that would corrupt the DCC data stream. The outputs from IC1 are available Australia's electronics magazine at CON2 and are shown by bicolour LED2. The presence of a short circuit will be obvious, since LED2 will not light up when expected. Other circuitry The MOD2 LCD panel needs seven data lines to interface an SPI-mode controller to its LCD driver and touch controller; these are taken from appropriate pins on MOD1. MOD2’s LED backlight is driven by P-channel Mosfet Q1 from the 5V Vsys rail. Q1 is in turn switched by Q2, which is controlled by a pin on the Pico 2, GP3. The 10kW/1kW divider combined with a 1μF capacitor across the incoming supply from CON3/CON4 is connected to the last free analog/ADC input on the Pico 2. This allows the supply voltage to be monitored and displayed. Two of the remaining free pins are connected to CON5 (a four-way header) and CON6 (an RJ45 socket) for connection to external control boxes, along with the 3.3V and ground rails. We have chosen the GP0 & GP1 pins since they are capable of both I2C and UART (serial) operation. They have 2.2kW pullups to the 3.3V rail. S1 connects to the 3VEN pin on MOD1; when this line is pulled low, by S1 being pressed, the 3.3V supply on the Pico 2 is shut down. This can be siliconchip.com.au Fig.6: The Base Station circuit is based on the Pico 2 microcontroller module, MOD1, and an LCD touchscreen, MOD2. The incoming supply powers these through buck regulator REG1 and also feeds IC1, IC2 and IC3. Those three chips provide the DCC outputs under the control of MOD1. used to reset both the Pico 2 and any connected control boxes, since they are also powered from 3.3V. An external momentary pushbutton connected in parallel with S1 could be used to provide an emergency stop feature. Power supply A DC supply of 8-22V is provided to either CON3, a DC jack, or CON4, a pair of screw terminals. These are connected in parallel with the intent that one or the other is used. The DC jack should be good for up to 5A, while the screw terminals can handle up to 10A. We’ve specified an 8V to 22V supply voltage range because those are the siliconchip.com.au limits set by the DCC standards. The components on this rail are all rated up to 30V. You’ll also need a suitable DC power supply wired with a positive tip. A basic 12V supply capable of at least an amp will be sufficient to run some tests and operate the Base Station and a few small locomotives. Fuse F1 provides circuit protection, with reverse-connected diode D1 forcing the fuse to blow in the event of a reverse-polarity voltage being applied. This arrangement is preferred at higher currents, since the polarity protection diode must carry the full current at all times if arranged for reverse blocking. The 1000μF capacitor provides the Australia's electronics magazine bulk bypassing for the supply. REG1 is used to provide the low-voltage rail for the microcontroller and display modules. It is a switch-mode device, since we will be driving a backlight (typically 300mA) at 5V, dropping around 10V from the supply. A linear regulator would dissipate 3W or more if used here. REG1 is an MCP16311, the same device we used in the Homemade 78xx Switchmode Regulator from August 2020 (siliconchip.au/Article/14533). The circuit here is much the same as the 5V version of the Switchmode Regulator, although we have used common E12 resistor values of 56kW/10kW January 2026  37 to give a nominal output of 5.3V with a 0.8V reference voltage. This is a classic buck regulator circuit, with inductor L1 storing energy between switching cycles under the control of REG1. The 5.3V rail has a 100μF electrolytic capacitor for filtering, and the circuit includes the four ceramic capacitors needed by the regulator. Since the 5.3V rail passes through schottky diode D1 to the remainder of the circuit, we have near enough to 5V at the point of use. Connected to D1’s cathode is pin 39 (Vsys) of the MOD1 Pico 2 microcontroller module, along with the MOD2 LCD touch panel module supply. The Pico 2 has its own diode from the USB supply feeding into the Vsys pin. These diodes prevent back-­ feeding from the regulator to USB or vice versa. A connection to MOD1’s micro-B USB socket can also be used to provide power to the low-voltage (5V and 3.3V) circuits for testing. Software From the hardware, we can see that we have a high-power (up to 10A) driver output that will be used for the main DCC track signal. The second driver output will be used for the programming output. We can monitor the drive currents via two of the ADC inputs, with the incoming supply being measured by the third. The Pico 2’s second processor core spends most of its time monitoring the CON1 current so that it can react promptly if there is a fault. If the current limit is reached, the output is switched off for one second, then back on. It might switch off again immediately if the fault has not been cleared. This core also measures the other analog channels when needed. Both DCC signals are provided by a callback function from a timer interrupt; the interrupt triggers every 58μs. DCC uses pulse lengths of 58μs (nominal) to signal a binary ‘1’ and a pulse length of 100μs or longer to signal a ‘0’. Two 58μs periods are used to generate a 116μs pulse length for transmission of a ‘0’. The callback function provides digital signals to control IC1, IC2 and IC3. It processes each packet’s bits in turn, and flags when it is ready to receive the next packet. A packet takes around 6ms to deliver, so our main software loop simply needs to supply fresh packets as needed. If there is no data available, so-called ‘idle’ packets can be sent to keep valid DCC traffic on the rails. This can occur if the processor is otherwise busy doing other processing, such as updating the display. All display pages in the user interface have buttons for switching the DCC output off and on, so power can be shut off to the track immediately if there is a problem. A stop button also sets the speed of all locomotives to zero. The voltage and main track current are also shown at all times. The main control page provides five tabs, each of which can be allocated a DCC locomotive address. There are controls for speed, direction and function (eg, lighting) outputs. Two packets are needed to send all this data for each locomotive, so a queue of 10 packets is kept updated and sent in round-robin fashion. The interior layout is similar to many of our LCD BackPack projects, with the LCD screen connected to a main PCB assembly via a 14-way header & tapped spacers. 38 Silicon Chip Australia's electronics magazine When a control is changed, such as a speed control being adjusted, a priority system allows the relevant packet to be output as soon as it is changed. This makes the system more responsive to user input. There are two other pages. One provides settings pertaining to the Base Station and includes things such as calibration values for the current and voltage readings and user-settable parameters, like a software-controlled current limit. Programming output The remaining page controls the DCC signal on the programming track output at CON2. In general, at most one decoder (and thus locomotive) should be connected to the programming track. This is because ‘service mode’ programming does not distinguish locomotive addresses. CON2 supports direct, paged, physical and address-only programming modes. Of these, direct mode is the newest and fastest, although it has been around for at least 20 years already, so most modern decoders should support it. We recommend using this mode unless it does not work with a specific decoder. Programming involves writing values to certain CVs (configuration variables) to change the behaviour of the decoder. Physical mode only supports a very limited number of CVs, while paged mode supports more through the use of a page register. Service-mode programming relies on specific patterns of packets, including repeated packets and so-called ‘reset’ packets to ensure that programming does not occur unless intended. These patterns are noted in the standard, but we have also validated them against the output of a commercially available DigiTrax base station. DCC also implements an acknowledgement feature, which can be used to read back data programmed into decoders. The acknowledgement involves the decoder loading the output with a 60mA or higher load; typically, by briefly driving its motor outputs. Thus, the Base Station can also perform a read-back of CVs to check their values or confirm them after writing. Our circuit allows the acknowledgement to be seen as LED2 dimming due to the load on the 33W resistors. The Programming page can also siliconchip.com.au send operations-mode programming packets. They are not sent via the CON2 programming track; instead, they go to the main track via CON1 instead. These packets are addressed, so they use the currently selected decoder address from the main pages. There is no read-back, since operations mode does not use the acknowledgement scheme described above. We’ll look more closely at the software operation after the Base Station has been completed. Our separate article will also provide more detail about programming decoder CVs with the Base Station. Control panel If you are planning to fit the Base Station into a larger panel, such as the control panel for an existing layout, then we recommend that you use the panel PCB as a template to trace the outline of the shape. Tracing around the main PCB (before it’s assembled) will give you an idea of the amount of material you need to cut out of your panel to fit the Base Station assembly into it. The LCD panel mounting holes can be used to align the two PCB outlines. You will probably have your own ideas about what connectors you will use, so you may not want to fit the standard connectors until you have worked out how it will connect to your layout. If you don’t think you’ll use the CON6 remote control connector, the tab that protrudes from the PCB can be carefully snapped off. This means that a hole does not need to be cut in the case for CON6. It can still be fitted later, since the traces do not cross onto the tab, but it will lack mechanical support. Construction Start by assembling the main PCB, which is coded 09111244 and measures 130 × 68mm. Most components are on the top, including the majority of surface-mounting parts. The smallest parts are M3216 size (imperial 1206), so construction is not too difficult. Gather your SMD equipment and consumables, including flux paste, solder wicking braid, tweezers and a magnifier. Begin with the SMD parts on the top of the PCB, followed by the two SMD parts on the reverse. After cleaning off any flux residue, the handful of siliconchip.com.au Parts List – DCC Base Station 1 Base Station PCB assembly (see below) 1 black panel PCB coded 09111244, 130 × 68mm 1 3.5in LCD touchscreen module (MOD2) [Silicon Chip SC5062] 4 M3 × 8-10mm black panhead machine screws 4 M3 × 6mm panhead machine screws 4 M3 × 12mm tapped spacers 4 M3 nylon hex nuts 1 UB3 Jiffy box [Altronics, Jaycar, Bud Industries CU-1943] 1 DC power supply to suit layout (see text) Base Station PCB assembly 1 double-sided PCB coded 09111243, 55 × 131mm 1 Raspberry Pi Pico 2 microcontroller module programmed with 0911124B.UF2 (MOD1) 1 14-way 0.1in socket header strip (for MOD2) 2 2-way 5mm/5.08mm pluggable screw terminal blocks (CON1, CON2) [Altronics P2592 + P2512, Jaycar HM3102 + HM3122, or Dinkle 2EHDRC-02P + 2ESDV-02P] 1 PCB-mounting DC barrel jack (CON3) 1 2-way 5mm/5.08mm screw terminal (CON4; optional) 1 4-way 0.1in R/A locking header (CON5; optional, for remote control) 1 RJ45 PCB-mount socket (CON6; optional, for remote control) [Altronics P1448 or P1448A] 1 22μH 1.3A SMD inductor, 6×6mm (L1) 1 6 × 6mm through-hole tactile switch with short (~1mm) actuator (S1) 2 M205 fuse clips (F1) 1 M205 fuse to suit PSU (F1) 1 small tube of neutral cure silicone or similar to secure the capacitors Semiconductors 1 DRV8231DDAR motor driver IC, SOIC-8 (IC1) 2 BTN8962TA half-bridge drivers, TO-263-7 (IC2, IC3) 1 MCP16311(T)-E/MS buck regulator, MSOP-8 (REG1) 1 SSM3J372R or AO3401(A) P-channel Mosfet, SOT-23 (Q1) 1 2N7002 N-channel Mosfet, SOT-23 (Q2) 1 SS14 40V 1A SMD schottky diode, DO-214AC (D1) 1 BAT54C dual common-cathode SMD schottky diode, SOT-23 (D2) 1 1N5404 or 1N5408 3A silicon axial diode, DO-27 (D3) 2 3mm bicolour red/green LEDs (LED1, LED2) Capacitors (all SMD MLCC, M3216/1206 size, except as noted) 1 1000μF 25V radial electrolytic 1 100μF 25V radial electrolytic 5 1μF 50V X7R 4 100nF 50V X7R Resistors (all SMD M3216/1206 size, ±1%, ⅛W except as noted) 1 100kW 9 1kW 1 56kW 2 33W M6332/2512 size, 2W 4 10kW 1 0.1W 4 2.2kW The DCC Base Station is a simple but complete system for starting out with DCC. The Base Station has controls for five locomotives. We have also designed a DCC Remote Control that can provide extra controls. Figs.7 & 8: the board uses a mix of surface-mounting and through-hole components and modules. Most components are on the top side of the PCB, but we have placed F1 on the back to allow easy access if needed. through-hole parts and modules can be fitted. Figs.7 and 8 are the overlay diagrams for the top and bottom, respectively. You can find photos of the PCB assembly on earlier pages. Regulator REG1 comes in an MSOP package with the closest pin pitch on the board, so start with it. Spread flux paste over the pads on the PCB and rest the chip in place. If you can’t see the pin 1 marking on the silkscreen, it is near the 56kW resistor. Tack one lead, check the positioning and then solder the remaining pins when it is correctly placed and flat against the PCB. Follow with the SMD diodes and transistors, being sure not to mix up the three different SOT-23 parts. Single diode D1 must have its cathode stripe facing correctly, towards the ‘K’ on the PCB. Follow by soldering IC1, then IC2 and IC3. Since IC1 will be operating at a small fraction of its limit, we have opted not to solder the exposed thermal pad on its underside. Tack one lead, adjust and solder the remaining leads when you are happy with its position. 40 Silicon Chip For IC2 and IC3, be sure to add a generous amount of flux to the pads before placing the part, and turn up your iron if it is adjustable. Tack one of the smaller leads in place, then add a good amount of solder while applying your iron to the large tab and pad near CON1. If your iron cannot provide enough heat, you can try preheating the board or supplementing the iron with a hotair tool. When the solder flows freely and the flux is smoking, you will know that the joint is solid. Finish by carefully soldering the remaining leads. The remaining SMD parts are all passives. Inductor L1 is larger than the others, so solder that now while the iron is hot, then turn it back down for the remaining passive components. All the values are marked on the silkscreen, so take your time and make sure that they are all placed correctly. Finish with the two 100nF capacitors on the back of the PCB. Clean off any flux residue using your choice of solvent and allow the PCB to dry. Inspect it closely for bridges, dry Australia's electronics magazine joints and pads that do not have solder adhering correctly. Fix any problems before proceeding. At this stage, the circuitry around REG1 is complete, so you can test it by applying a current-­ limited power supply (such as a 9V battery) to the pads of the (not-yet fitted) 1000μF capacitor. You should see 5.2V-5.4V on D1’s anode relative to ground (the negative lead of either electrolytic capacitor). Programming the Pico 2 We suggest programming the Pico 2 now, since it will be more difficult to access its BOOTSEL button when the LCD is affixed above it. We also recommend using the flash_nuke.UF2 firmware image to ensure that the Pico 2’s flash memory is blank first, although this should not be strictly necessary if the Pico 2 is brand new. Hold the BOOTSEL button on the Pico 2 and connect it to the computer, then copy the flash_nuke.UF2 file to the RP2350 virtual drive that will appear. Wait for the drive to disappear and then reappear, then copy the 0911124B.UF2 firmware. The LED on MOD1 should light up, indicating that the firmware has been loaded correctly and is running. The remaining through-hole components For simplicity, we recommend soldering the Pico 2 directly onto the siliconchip.com.au PCB, surface-mount style. You can use headers, but since full-height headers would be too tall, you will need to use low-profile headers and no sockets. That will work, but the clearance is very tight. The 14-way header socket for the LCD can also be fitted now. Make sure all headers are mounted squarely; you can use MOD1 and MOD2 to align them. Solder CON1 and CON2, the pluggable screw terminals, and follow with your choice of CON3 or CON4, since only one of these is needed. You should also fit either CON5 or CON6 if you plan to use the Remote Controller. We preferred to use the RJ45 socket (CON6), since we can use standard Cat 5/6 cables to connect Remote Control units. Next, mount the fuse clips for F1 (with the retention tabs on the outside), large diode D3 and the 100μF capacitor on the rear of the PCB. You can slot a fuse into the clips to keep them aligned. Switch S1 and the 1000μF capacitor are the last parts on the top of the PCB. Make sure to bend the capacitor leads the right way before soldering and add some glue or silicone to secure the capacitor bodies to the PCB. Leave off LED1 and LED2 for now. To fit the LCD panel and align the LEDs, start by attaching the M3 × 10mm machine screws to the front panel PCB using the nylon nuts. The nuts will act as spacers for the LCD panel below. Slot the LCD panel over the screws, making sure that the 14-way header is at the end opposite the LED holes in the panel. Secure the LCD panel with the tapped spacers. Now guide the LEDs into their holes on the main PCB, but do not solder them. Their polarity does not matter, since a DCC signal is effectively alternating current. Attach the LCD panel assembly to the main PCB, making sure that their 14-way headers connect. Secure the main PCB using the M3 × 6mm machine screws. Bring the LEDs up so that they are just poking out through the front of the panel and solder them in place, then trim the LED leads and check that you have a fuse fitted. The rating of the fuse should match that of your chosen power supply. Initial checks The 5V-powered parts of the Base Station can be supplied from the Pico 2’s USB socket, so USB power is sufficient to check that the processor and LCD touchscreen are working. Connect USB power to the Pico 2 and see that the LCD backlight switches on and Screen 1 is visible on the panel. Verify the touch panel calibration by trying some of the buttons. The default touch calibration should work for all 3.5in panels, but there are parameters that can be edited if it does not. Check the “Arduino library and software” panel overleaf, as this has more details on adjusting the calibration and customising the software. If all is well, disconnect the USB cable. Testing and setup A 9V battery is a good choice for a current-limited power supply, but just about any plugpack that can deliver a few hundred milliamperes at 8V-22V Screen 1: the SET and PR buttons can be used to access Screens 2 and 3, respectively. One of the five tabs can be selected using the L1-L5 buttons, while the address can be changed by using the button at upper left. siliconchip.com.au should be sufficient. Check that the Base Station powers on and that the voltage display at lower right matches the power supply voltage. The displayed current should be 0.0A. The SET page (Screen 2) holds the calibration parameters (I1x, I2x, Vx and IO/S). The software current limit for the main track is ILIM. Most of the calibration parameter defaults should be usable, but the current offset (IO/S) can vary wildly. This parameter is due to IC2 and IC3 producing a non-zero current signal at zero current, so the offset setting is necessary to cancel this out. If you use a different supply voltage, this may change. If you wish to redo this calibration, reset the IO/S parameter to zero before doing so. Make sure nothing is connected to the MAIN (CON1) output and change the ILIM parameter to 9A by pressing the ILIM button and typing 9 ENTER on the on-screen keypad. Press the yellow ON button, which should cause the MAIN LED to light. You should be able to see that both the red and green elements are on in the LED; if not, then there is likely a fault in one of the drivers. The current display should also show a non-zero value around 4A to 5A, although anywhere between 1A and 9A can be expected according to the BTN8962 data sheet. Take that reading and enter it in the IO/S field. The current reading should now drop to 0A with the offset applied. Now adjust the ILIM value to suit your power supply. All values are immediately saved to flash memory, Screen 2: apart from the calibration parameters, the button at lower right saves the currently selected locomotive selections (L1-L5). Pressing this button should show SAVED, after which L1-L5 will be automatically loaded when the Base Station is next powered on. Australia's electronics magazine January 2026  41 so you don’t need to perform an extra step to save them. The other parameters should be within a few percent without adjustment, so should not be changed unless you have an accurate way of measuring the voltage and currents. Enclosure preparation The panel PCB has been designed to fit a UB3 Jiffy box; Figs.9 & 10 show the cut-outs needed to fit the assembly into this box. We have not included holes for CON4 or CON5, since we have not used them in our prototype. The 6mm hole for the CON3 DC jack suits our power supply, but you may need to enlarge it if your plug has a short shaft. If you’re planning to use CON4 instead, you can make a hole in front of that for wires to pass through. Similarly, if you plan to use CON5 instead of CON6, you could omit the rectangular cut-out for the RJ45 socket and drill a hole for wires to pass through instead. The three rectangular holes can be made with vertical cuts from the top of the case. Score the horizontal cut with a sharp knife and snap off the tab with pliers. The round hole is simply drilled with a twist or step drill. The PCB assembly takes the place of the Jiffy box lid, and can be secured using the screws that are provided with the box. We prefer to surface-mount the Pico 2 module. If you find that the holes are slightly misaligned, you can trim the sides of the holes using a sharp hobby knife. Using it We’ll now take a look at the basic operation of the Base Station. Those who have experience with a DCC system should be able to take what they need from these brief instructions. Note that only 128-step speed instructions are issued. For more details about getting started with DCC for the first time, refer to our separate article in this issue, starting on page 49. When the Base Station is powered on, it starts on the main page, seen in Screen 1. Buttons L1-L5 select the active locomotive, which is highlighted. The controls above this operate on the active locomotive. The top left button can be used to change the selected address controlled by L1-L5. DCC uses two types of addresses; a short address is seven bits, and is valid between 1 and 127, although values above 99 are generally avoided since they conflict with some programming packets. Long addresses are 14 bits and are valid from 1 to 10239, enough to hold all four-digit numbers. In both cases, zero is not valid, so it is used to indicate that the tab is inactive. This is shown as three dashes in the address box. Any address entered with three or more digits is treated as a long address. To use a long address in the range of DCC Base Station Short-form Kit (SC7539, $90): includes everything in the parts list, except for the case, power supply, glue, CON4 & CON5 headers 42 Silicon Chip Australia's electronics magazine siliconchip.com.au Screen 3: the default of DIR (direct) mode programming is the newest and should work with all modern decoders (including our own design from last month). You can use this to read or write the decoder’s CVs (configuration variables). 1 to 99, add leading zeros to pad the value to three or more digits. Long addresses are displayed with five digits using leading zeroes. There are a few safety interlocks in the code. You cannot set two tabs to use the same decoder address, and you cannot change an address without reducing the speed to zero first. That helps to prevent conflicting control commands and runaway trains! The REV, FOR, STOP and F0-F3 buttons control the commands that are sent to the addressed locomotive. F0 has a toggle action, since it is usually used for controlling a light such as a headlight. You might see it referred to as FL for this reason. F1 and F2 are momentary-action and are typically used to control a horn or whistle. The four indicators above the buttons show the state. F3 is provided with a toggle action, so there is another latching control available. Switch on the DCC track power (ON) if it is not already. To operate a locomotive, enter its address at upper left, then select the direction (FOR or REV) and switch on the headlight (F0) if needed. Drag the slider to change the speed, which is displayed with the direction on the top line. Pressing the yellow STOP button will set the speed of all locomotives to zero, while OFF can be used to shut off power in an emergency. The current display will be green if it is below the limit, or red if it has tripped. You will also see the MAIN LED go out when a trip occurs. Screen 3 shows the page used for CV programming. This is accessed through the PR button on the main page. DIR, PAG, PHY and OPS refer to direct, paged, physical and operations mode programming, respectively. DIR, PAG and PHY modes occur on the CON2 programming track, while OPS packets are sent to the CON1 main track to the currently active locomotive, as selected by L1-L5. Power is only applied to the PROG output when a read or write is occurring on the programming track, so you will see the PROG LED light up during these times. Pressing BACK during programming will cancel the operation. The CV to be programmed is entered with the CV# button. It can be read with the READ button, with the value shown below it if the read was successful. A value to be written is entered in the box below WRITE and pressing the WRITE button performs that action. The status of the last or current action is shown at the top of the page. The LONG address is actually a pair of CVs (17 and 18). They can be edited separately, but the LONG button manages the value of both of these together when reading or writing a long address. Press LONG instead of entering a CV# to access this mode. Note that you will need to set the long addressing bit (CV29, bit 5) to activate the long address once set. Other information The Software panel overleaf has more information on the libraries used in writing the software for this project, so you should have a look at that if you wish to compile the sketch yourself. Our separate feature article has more depth on using our Decoder and Base Station as a complete system. We recommend reading it if you are new to DCC. The Decoder article from last month also has information about the most common CVs, including all that are implemented by that Decoder. That article also includes a glossary of DCC terms. Figs.9 & 10: with rectangular holes abutting the top edge of the case, it is not difficult to make the cuts needed. Once the PCB assembly is complete, you can use it to judge whether any of the holes need trimming. siliconchip.com.au Australia's electronics magazine Conclusion The DCC Base Station is a simple but complete control unit for DCC Decoders. Once you have built it, adding Decoders to the locomotives on your layout will provide most of what is needed to convert a layout to DCC operation. January 2026  43 Arduino library, software & screen calibration This panel provides a bit more background on the libraries and other code that are used for anyone interested in compiling the Arduino code, either to make some tweaks or perhaps create your own version. We’ll also discuss how to calibrate the touch panel. The following assumes that you have an assembled Base Station PCB connected to a 3.5in LCD touchscreen or, at least, the same wiring between a Pico 2 and the LCD panel. A solid background using the Arduino IDE would help. Many of the functions used by the main sketch are in the util.h file. Near the top of this file are some defined colours, so you can adjust the colour scheme easily. The dcc.cpp and dcc.h files contain the DCC-specific drivers. LCD driver The LCD driver library is the main external library we have used, and this is the TFT_eSPI library. It can be found at https://github.com/Bodmer/TFT_eSPI or installed by searching for TFT_eSPI in the Arduino Library Manager. We also use the TFT_eWidget library (https://github.com/ Bodmer/TFT_eWidget) to draw the GUI elements. These libraries are quite powerful and offer anti-aliasing on the fonts and GUI elements, so the display looks very nice. You will need to install these libraries and any dependencies they require. Rather than using a configuration within the sketch, this library uses a global (library-level) configuration for the display pinout and driver selection. You will need to set this up before doing anything else with the library. The code for this configuration is noted in the util.h file. It requires creating a profile in the “libraries\TFT_eSPI\ User_Setups” folder to suit the display type and wiring; this is the PICO_ILI9488_DCC.h file that you will find in the software bundle. Then edit the User_Setup_Select.h file to include the PICO_ILI9488_DCC.h file as the active configuration. This configuration will now be used for all sketches compiled with this library, so you can try any of the example sketches using the Base Station display hardware. You can subsequently change configurations by editing the User_Setup_ Select.h file. The Examples → Generic → Touch_calibrate sketch can be used for touch panel calibration. Upload this sketch and open the serial monitor. Run the calibration, and the results are displayed on the serial monitor. The updated values can be used to set the calData array in the main sketch file before compiling. That’s all there is to changing the calibration. tool at https://vlw-font-creator.m5stack.com, we converted this into a VLW file using a size of 36pt. We then used the HxD hex editor program to convert the file data into a byte array that could be embedded in the sketch; this is the asimov_36.h file. DCC code The DCC code has been written with the Pico/Pico 2 architecture in mind, so you will need the arduino-pico board profile installed. The DCC code depends on this profile and the Ticker library that calls the DCCcallback() function every 58μs. The code provides functions to create all manner of DCC packet types with both long and short addresses. The main track DCC output implements a short queue that can be filled with the queuePacket() function. If the queue is empty, the code will produce idle packets to keep valid data on the track. As the DCCcallback() function consumes the packets, new packets can be added. For the most part, the software updates an array of the 10 packets that are needed to control the state of the five locomotive outputs. As the queue empties, the software cycles through the array and delivers each of the 10 packets in turn. There is a loco_t data type that can be used to hold the information (speed, direction, address etc) about a decoder. Most packets can be created directly from the loco_t object using straightforward function calls. For the simplest implementation of a mainline track output, create a loco_t object and set its various elements (address, speed, direction etc). Create two dccPacket_t objects and call the speedPacket128() and F04Packet() functions to load these packets with speed and function data, respectively. Use the packetQueueSpace() function to see if there is space in the queue and, if so, queue the packets with the queuePacket() function. Keep updating the loco and packets, and continue queuing fresh packets as needed. The DCC output can be switched on and off with the dccSwitchOn variable. The programming track output works slightly differently; it doesn’t have a queue. It is expected that the packets for the programming track are managed from a tight loop that produces the specific packets as needed at the correct times. SC Fonts The anti-aliased fonts used in this project require a different data format than we have previously used. The asimov_36.h file contains the font data we created to suit this display. There are online tools to create custom font data from computer font files. We started with the open-source Asimov font in the OTF font file format. Using the The RJ45 socket on the right-hand side of the Base Station can be used to connect extra controllers. Next month, we will introduce our design for a Remote Controller which includes a display, speed potentiometer and six buttons. 44 Silicon Chip Australia's electronics magazine siliconchip.com.au