Sunday, October 25, 2015

Using output pads on Deltang receivers

Introduction

This article assumes you have a basic knowledge of installing Deltang equipment for controlling the motor of a battery loco, and have reached the stage where you'd like to make use of some of the output pads to operate lights or other accessories such as triggering the whistle on a soundcard. If you are unfamiliar with Deltang radio control equipment, then you might find Getting started with Deltang r/c in the garden useful.

To keep things simple, I will focus on the Rx60/Rx61 and Rx65 receivers as these are the ones I have used most extensively - but the general principles shown here apply to the majority of the other Deltang receivers on offer. Hopefully, once you have grasped the basics, you will be able to use the information on the Deltang website to explore other options.

This article is organised into the following sections
  • Soldering to the pads. How to solder to the pads without frying your receiver
    • Choosing a soldering iron
    • Tinning
    • Soldering
  • Which pad? Which pads do what on the receivers.
    • Using pads for lighting
      • Directional lighting
      • Using Pads A and B on an Rx65b for directional lighting
      • Interior lighting
    • Using pads to trigger effects on soundcards 
      • A cheap Greetings Card module
      • MyLocoSound
      • DigiSounds (MTroniks / Peter Spoerer)
      • Technobots programmable 
    • Using pads as inputs 
  • Conclusion

Soldering to the pads

Probably the most daunting aspect of using the pads, is having the confidence to actually bring a soldering iron close to the tiny circuit boards to fix wires on to the output pads - particularly on the Rx60 receiver. The Rx65 is only marginally more expensive than the Rx60 and so, if you have the space to fit it, you might find using an Rx65 the better option if your are not too confident. The Rx65 has larger pads which are slightly further apart, and so makes soldering a bit easier.
Deltang Rx60
Deltang Rx65
If you are already familiar with soldering techniques you can skip the next section, but I am assuming there are some who might find a beginners' approach helpful

Soldering iron

First, and most important of all, is to buy a soldering iron with a very fine bit. I use an Antex XS iron which has interchangeable bits (Maplin product code FR12N). I use the finest 0.5mm No. 55 bit (Maplin product code N85DS).

Tinning

Something I have learned from experience is the importance of having a well-tinned bit. The secret to tinning successfully seems to be to apply the solder to the bit while it is heating-up - before it reaches its full operating temperature.

Similarly, the wire and the pad need to be tinned before they are joined. I use multi-core solder, apply the well-tinned bit to the wire or the pad and almost immediately touch the interface between the bit and the piece with the solder. Once the solder has flowed, the bit is removed. Here you can see an Rx65 receiver with the supply and motor pads tinned (on the left hand side), Pads 2 and 3 (bottom right), pad C (upper left hand side), pad 5 (Upper right) and one of the negative supply pads (Upper left) tinned and ready to receive leads.

If you have bought a receiver shrouded in heatshrink sleeving, you will need to cut the sleeve to give access to the pads which need to be soldered. Make sure you know which pad(s) you will be using, so that you only remove the sleeving from the section(s) of the board required. (see Which pad? below)

Soldering to the pad

Provided the pad and the wire have been properly tinned with solder, soldering the wire to the pad is relatively easy and straightforward. The tinned end of the wire is held against the pad, the tip of the soldering iron is tinned with a small blob of melted solder and then touched against the wire and pad. The blob of solder on the iron helps the heat to flow quickly to the joint and so the iron needs only to be held against the joint for a second, thus reducing the risk of cooking the chips on the board.

Which pad?

Deltang receivers are supplied with their pads ready-programmed for a range of functions which might be needed by modellers. However, the function of each pad can be reprogrammed to tailor the receiver to the specific needs of the user (see How to re-program Deltang receivers).

To work out the function of each pad for a recently purchased receiver, you need to access the information provided on the Deltang website. For example, the information on the pads given for the Rx60 is:


1. DEFAULT SETUP
Item

Setting Details
P1 Front Light Auto action, LED2 enabled

P2 Rear Light

Auto action

P3 On/Off Direction switch (Ch3)
On when channel is Low, Momentary action
P4 On/Off Direction switch (Ch3)
On when channel is High, Momentary action
P5 On/Off Bind button (Ch5)
Start off, toggle when channel is Low, Latching action

So, what does this all mean? I've decided the best way to explain the outputs is to give some examples.

Using pads 1 and 2 for directional lighting on the Rx60/61 and Rx65b

Pad 1 (P1) and Pad 2 (P2) are set-up to automatically provide power for direction lighting (provided they are only powering LEDs). P1 will power a front light when the loco moves forward and P2 will power a rear light when the loco moves in reverse.

I used the P1 and the P2 outputs on my IP Engineering Lollypop railcar (see How I constructed an IP Engineering Lollypop railcar).

I could have kept things simple and had white LEDs in the lamps facing forwards and a red LED in the rear lamp, and wired-up both white LEDs to P1 and the red LED to P2, like this:

Bear in mind that, regardless of the input voltage (in this case 3.6v from three NiMh batteries), the output from the Deltang receiver is 3.3v. This voltage is what is used to calculate the values of the resistors which must be used in series with the LEDs to limit the current (usually to around 20mA). There is a useful calculator for determining the value of the resistors here - http://led.linear1.org/1led.wiz - but in general, I find 100 ohm resistors are fine for connecting to LEDs powered from Deltang pads.

For this little railcar, I actually used bi-colour LEDs so that the leading lights showed white and the trailing lights showed red whichever direction the railcar was travelling. The circuit for this set-up was (with the addition of 100 ohm resistors in the negative leads to each LED):


Using Pads A and B on the Rx65b for direction lighting
A slightly different approach needs to be adopted when using pads A and B on the Rx65b. As with Pads 1 and 2 on the Rx60 and the Rx65b, these pads provide automatic direction lighting but whereas pads 1 and 2 provide 3.5v outputs and so can be wired as shown with the Rx60 examples, Pads A and B provide 0v outputs when triggered. This means that when using pad A or B, an LED has to be connected the other way round - the cathode on the LED (ie the shorter leg) is connected to the pad and the anode (ie the longer leg) is connected through a resistor to the positive battery lead. The value for the resistor will therefore be determined by the supply voltage from the battery.

If you are uncertain about the parameters for your LED you can assume that:
  • the forward voltage or voltage drop across a red, yellow or green LED is generally 2v
  • for white or blue the forward voltage is generally 3v. 
  • while some white LEDs can operate safely with higher currents, you can assume that most LEDs operate satisfactorily at 20mA or 10mA. Bright white LEDs, for example, will give a relatively high level of light at 10mA or 5mA. Decreasing the current used for lighting will, of course, increase what's available for driving the motor.

My two larger diesel locos both have LED powered front lights which are connected to Pad A. Originally, I used Pad 1 for the front lighting and so used a 100 ohm resistor.

Lately, I have taken to using Pad A and, because I am now connecting the LED to my 12v supply, I use a 470 ohm resistor:

Note: For the purposes of the calculation, I assumed my supply was 12.5v (three li-ion batteries freshly charged can reach this voltage) and I knew the forward voltage of my LED was 3.3v. I went for 20mA as, although my white LED could take up to 30mA, I prefer to include a safety margin.

So why bother to use Pad A rather than Pad 1? Quite simply, Pad A provides an additional feature which is quite useful. The output from the pad mirrors the state of the LED on the receiver. As my receivers are hidden away inside the bodies of the locos, I can't see when the receiver is searching for the transmitter and when it has located it. If ever I want to re-bound the receiver to another transmitter, I can watch the front light until it starts flashing rapidly to show the rx is in bind mode. Once the rx has located the transmitter, the front LED headlight goes into its normal directional lighting mode. This LED2 feature wasn't available on the Rx60s when I bought them, though I see it is now provided by default.

Using pads for interior lighting

On the Rx60, Pads 3 and 4 and on the Rx65b pads 6 and 7 respond to the direction switch on a Deltang transmitter (Channel 3). So, on the Rx60 Pad 3 comes on when the switch is flicked one way and Pad 4 when the switch is flicked the other way (usually Pad 3 for down and Pad 4 for up - but you may need to experiment to check this). I used Pad 3 of the Rx61 in my railbus to operate the interior lighting ......

........ (as well as Pads 1 and 2 to power the headlamps in each direction). (see How I constructed a pair of Ford type railmotors)

 Originally, I had four LEDs in each carriage for the interior lighting and so needed to use a transistor switch in the circuit to avoid overloading the receiver (see How I made a simple transistor switch). However,  four LEDs looked far too bright for 1930s interior lighting and so I reduced the number to two per carriage. This meant I could also do away with the transistor switch

I also used just one 100 ohm resistor for the two LEDs connected in parallel (assuming 3.3v supply, 2v forward voltage, 7mA current (because I wanted them to be dim)). However, connecting LEDs in parallel with one resistor is generally frowned upon as it assumes the LEDs in the circuit have exactly the same electrical characteristics - if not, one LED could become overloaded. However, as I had sourced the warm white LEDs at the same time from the same supplier and I was not running them at full power, I felt it would be OK. This online resource calculates the required values for resistors in parallel circuits - http://ledcalc.com/#calc

Using pads for triggering sounds

So far, I have had experience of four different types of soundcard (MyLocoSound, MTroniks/Peter Spoerer DigiSounds diesel soundcard, the Technobots programmable soundcard and a cheap sound module from a greetings card), but they all follow a similar approach. The cards all needed at least one input from the receiver to, for example, trigger the horn or whistle. In all cases, these triggers required a 0v input - in other words, they needed the input lead to be connected to the negative terminal of the supply battery to trigger the sound.

A Greetings Card sound module
The simplest system was using the module from a 'musical' greetings card. This was bought online from China for the princely sum of £1.29 (including postage).

After recording the sound of a horn (using the microphone and record button provided), the mike, record button, LED and batteries were removed and the board wired into the same battery supply as that used by the Deltang Rx65b receiver (a single 3.7v li-ion battery). The 'play' button was removed and the lead (not the one leading to the negative supply on the board) was soldered to output pad C on the Rx65b. Pad C is triggered by the bind button on any of the Deltang transmitters (Channel 5). As Pad C provides a 0v output when triggered, this was ideal for sounding the horn.

You'll see (or rather hear) from the video that I have also used another greetings card module to provide an engine sound. This is not triggered by the receiver (though it could easily be using another pad, but it would need to be re-programmed or inverted (see below) to provide a 0v output). I retained the 'play' button for triggering the engine-sound board (see How I modified a cheap greetings card module into a soundcard).


MyLocoSound soundcards
Four of my locos are presently equipped with MyLocoSound cards. Two have the older original type of card which uses small preset potentiometers to adjust the card's settings, such as volume, whistle tone and chuff rate. The other two are equipped with the more recent Universal card, which is adjusted by using a TV style remote control. The Universal card also has several more effects which can be controlled by the user, such as guard's whistle, safety valve and brake pump.

As indicated above, with a Deltang Rx65, it is very easy to trigger the whistle or horn (or any other effect) on a MyLocoSound card. To trigger an effect on the MyLocoSound card the relevant input needs to be connected to 0v or the negative terminal of the power supply. Pad C on the Rx65 gives an output of 0v when it receives a low signal on Channel 5 from the transmitter (see table below from the Deltang website)
mSettingDetails
Purpose:Rx65-22Train with Tx22 transmitter
Red wire positive (+)
Black wire negative (-)
Battery3-18v
Observe polarities
H1 outputMotor
Ch1
Integrated forward/reverse ESC for brushed motors
Center off
F1 output 'A'
(P13)
Front Light
H1 (auto)
0v when on, disconnected when off; LED2 enabled
F2 output 'B'
(P14)
Rear Light
H1 (auto)
0v when on, disconnected when off
F3 output 'C'
(P15)
On/Off
Ch5
Start disconnected, 0v (on) when channel is Low
Momentary action
P1Front LightAuto action, 3.5v when on, 0v off
P2Rear LightAuto action, 3.5v when on, 0v off
P3On/OffCh2, Idle high, 0v when channel is Low, Momentary action
P4On/OffCh4, Idle high, 0v when channel is Low, Momentary action
P5On/OffCh5, Start high, toggle when channel is Low, Latching action
P6On/OffCh3, Idle low, 3.5v when channel is Low, Momentary action
P7On/OffCh3, Idle low, 3.5v when channel is High, Momentary action
P8ServoCh1, Standard servo
P9ServoCh2, Standard servo
P10ServoCh3, Standard servo
P11ServoCh4, Standard servo
P12ServoCh5, Standard servo

On Deltang transmitters, pressing the bind button does just this and so, if Pad C is connected to the whistle input on the card, the whistle or horn will sound when the bind button is pressed. To protect the receiver from excess current passing from the soundcard (which operates at 5v) it is advisable to put a 1k resistor in the connection from the pad to the trigger terminal on the soundcard.

As indicated above, the MyLocoSound Universal soundcard has several other effects and so, if the trigger terminals for these effects could be connected to 0v through the receiver, then they could all be triggered by the transmitter. At the time of writing, Pad C, Pad 3 (Channel 2) and Pad 4 (Channel 4), are the only ones on the Rx65 as bought which provide 0v in response to signals from the transmitter. This is OK if you are using an aero transmitter or a Deltang Tx20, as these can send signals on Channels 2 and 4. However, if you are using a Tx22, the only other channel which can be triggered is through the direction switch (Channel 3). Pads 6 and 7 respond to this channel but they provide 3.5v outputs rather than the 0v required by soundcards. One solution is to reprogram the pads to give 0v (see How to reprogram Deltang receivers) but another approach is to use a simple bit of electronic wizardry to change a 3.5v output into a 0v output.

Adding a transistor inverter circuit instead of reprogramming


Recently, I decided to use an Rx65b to trigger the five sound effects on the most recent MyLocoSound Universal Steam Soundcard. As I was using a Deltang Tx20 transmitter, I could use the buttons on the transmitter to energise channels 2, 4 and 5 which would send 0v from pads 3, 4 and C on the Rx65b to the soundcard but I couldn't use the direction switch (Ch3) unless I reprogrammed pads 6 and 7 to give 0v. Realising that many garden railway modellers would be daunted by the prospect of reprogramming, I  decided to explore another option; adding a small transistor circuit to the pads to invert their outputs from 3.5v to 0v (see How I triggered MyLocoSound effects with a Deltang Rx65b). 

With the addition of two transistors and four resistors (a total outlay of around £1.00), and a tiny bit of soldering, I now have a fully functioning remotely operated soundcard - literally with all bells and whistles!

The DigiSounds small / narrow gauge diesel soundcard
This soundcard sounds impressive, as it uses genuine digitised diesel sounds. In addition to the horn, the card includes start-up and engine-off sound effects. As with the MyLocoSound card, these effects are triggered by connection of the relevant inputs to 0v. I installed a DigiSounds card in my IP Engineering Jessie.

Triggering the horn was straightforward, I simply connected the input on the card to Pad C (or Pad 5) on the Rx65b. To trigger the engine start-up, I decided to reprogram pad 3 to respond to the direction switch on any of the Deltang transmitters (Channel 3) - see How I installed a Deltang Rx65 in my IP Engineering Jessie loco. Pad 1 was used for the direction lighting to power the front headlight (pad 2 is shown connected in the photo as I originally wired the loco in reverse).

Eventually I rewired her to make use of Pad A for the headlight

Here she is in action after the installation


The Technobots / Alan Bond programmable soundcard

I came across this soundcard recently as I wanted a different sort of sound for my other diesel (a scratchbuilt Fowler-inspired loco) rather than using another DigiSounds card.

 The Technobots card generates the sounds electronically, rather than having digitised sounds. While some realism is lost, what is gained is programmability. There are ten different 'voices' provided on the card and each of these can be tailored to meet the specific needs of the user.

Initially, it appeared that this soundcard would be incompatible with the Deltang system as it is designed to fit between the receiver and the ESC, an option which is not normally available with Deltang receivers. However, undeterred, I explored different ways of enabling the card to be synched and triggered by my Deltang transmitters. Ultimately, I decided to use an Rx65b to control the motor on the loco and an Rx102 to communicate with the soundcard. The trigger for the horn was taken from pin 5 on the Rx102, through a 1k resistor to the Aux input on the card (the white wire in the photo below).

For more information on this project see How I integrated a Technobots soundcard with Deltang receivers. As you will see from the article, there was more than one way driving the sound effects, I opted for the approach which best suited my needs.

Using pads as inputs

In addition to driving devices, the pads on Deltang receivers can be used to input information to the card. The Rx65b includes built-in features which enable the motor to be controlled automatically. These features include - station stop mode (the loco slows to a stop at a station, waits, then continues on its way), buffer stop mode (the loco reaches the end of a siding, slows to a halt and does nothing until the operator moves the speed knob), and auto-shuttle mode (the loco reaches the end of the track, slows to a stop, waits and then moves off in the opposite direction). The length of time it takes to slow down and the length of waiting-time can be programmed by the user (see How to program Deltang receivers).

To enable the loco to know when it is required to stop, one of the pads needs to be connected to a reed switch (a switch which is activated when it comes close to a magnet). The other side of the reed switch is connected to 0v, so that when the loco passes over the magnet, the pad is triggered by being grounded to 0v (in the same way as the effects on the soundcards above are triggered).

So far, one of my locos has been programmed to operate in auto-shuttle mode on the 32mm gauge mine feeder line on my railway (see How I programmed a Deltang Rx65b to operate in shuttle mode).

I am planning to program another two of my locos to operate in station-stop mode, so that on balmy summer evenings, I can relax in a deck-chair with a cool drink while a couple of trains happily chug around my railway stopping at stations without my intervention.

Conclusion

In my opinion, the great virtue of the Deltang radio control system is its versatility - not to mention the diminutive size of the equipment, the quality of its design, its cost-effectiveness and the range of features it provides. In this article, I have touched upon some of the things which can be done with Deltang receivers. Pads can be used to operate a range of devices, such as driving a servo for remote uncoupling or playing a preset sequence of LEDs in an array.

In addition to using Deltang receivers within locos, I have also used them to remotely control points on my railway (see How I operate my points by remote control). With a little bit of imagination and ingenuity, they could equally be used to control virtually any lineside gadget or accessory.

I hope this little introduction has been helpful and that it provides you with sufficient confidence to have a go yourself.

How I trigger MyLocoSound effects with a Deltang Rx65 receiver

Having previously constructed a Manning Wardle 0-6-0 locomotive using a Piko motor block (see How I constructed a Manning Wardle 0-6-0T loco), I decided it was about time I provided her with sound. As with three of my locos, I invested in a MyLocoSound steam soundcard.  The more recent Universal card is not only adjusted using an infra red remote control rather than using onboard preset potentiometers, it includes sound effects in addition to the conventional steam whistle. These effects include a guard's whistle, a bell (or short steam whistle), the sound of a safety valve blowing off and an air pump. The effects can be triggered by the remote control or through a receiver and transmitter with the requisite number of spare channels.

Having recently constructed a Deltang Tx20 transmitter (see How I constructed a Deltang Tx20 transmitter kit) which includes two push-buttons to operate Channels 2 and 4, I decided it was worth investigating whether I could use these, together with the bind button and direction switches, to trigger the additional sound effects on the soundcard.

What I wanted was:
  • To sound the whistle when the bind button was pressed (Ch5 low)
  • To start and stop the safety valve effect when button 1 was pressed (Ch2 low)
  • To sound the guard's whistle when button 2 was pressed (Ch4 low)
  • To sound the bell when the direction switch was flicked left (Ch3 high)
  • To start and stop the air pump when the direction switch was flicked right (Ch3 low)
A quick consultation of the chart showing the Rx65b's features on the Deltang website .......
Source: http://deltang.co.uk/rx65b-22-v611.htm
 ...... revealed that:
  • Pad 3 gives an output of 0v when Ch2 goes low (ie the button on the Tx is pressed)
  • Pad 4 gives an output of 0v when Ch4 is energised (ie the other button on the Tx is pressed)
  • Pad 6 gives an output of 3.5v when Ch3 goes low (ie the direction switch on the Tx is flicked down)
  • Pad 7 gives an output of 3.5v when Ch3 goes high (ie the direction switch is flicked up)
The only problem was that the inputs on the MyLocoSound card are triggered when they are connected to 0v (ie to 'ground' or the negative lead from the battery). This meant that the outputs from Pads 3 and 4 would be fine for triggering the card's inputs, but the outputs from pads 6 and 7 were the inverse of what was needed.

I could have reprogrammed the pads to give 0v outputs (see How to program Deltang receivers - pending), but decided that I would experiment with another way of achieving the same outcome, as I would be posting my findings on this blog and I know many people are daunted by the prospect of programming. Fortunately, I discovered that the output could be inverted using a simple transistor circuit:
Source: http://dlb.sa.edu.au/rehsmoodle/file.php/282/kpsec.freeuk.com/trancirc.htm
After consulting David T at Deltang, who advised me that the circuit would be fine and that I could use the voltage from the input terminal of the soundcard (which is around 5v) provided the 1k resistor was connected between the transistor circuit and the soundcard. He also suggested that all connections between the Rx65 and the soundcard should be protected from the risk of excess current by connecting a 1k resistor between the outputs from the pads and the terminals of the soundcard.
As there were only three components required, I decided to solder them together directly ......

.... and then shroud all the connections in heatshrink sleeving.

The leads from the 10k resistors (on the base leads of the transistors) were then soldered to the pads 6 and 7 of the Rx65, and the emitter leads of the transistors were soldered to two of the negative pads on the Rx65.

The rx with its attached transistors were then shrouded in an additional layer of heatshrink sleeving for added protection.

Flying leads from the 1k resistors (on the collector leads of the transistors) were then connected to the relevant screw terminals on the soundcard. 

With everything connected, the loco and card were ready for testing. And then the loco was reassembled and ready for test running.

With most aspects of garden railway modelling, there are often many solutions to any given problem. I could have reprogrammed any of the pads on the Rx65 to provide the 0v required, I could have connected a switcher unit to the servo output from pad 10 or I could have even used a completely separate receiver for controlling the soundcard, as I have done on another of my locos (see How I used Deltang receivers to trigger sounds on a Technobots soundcard). There are probably many other solutions, but we all tend to stick with the areas of knowledge with which we feel most comfortable, and as long as it works .....


Sunday, October 18, 2015

Installing a Technobots programmable diesel soundcard with Deltang receivers

Since converting my semi-scratchbuilt Fowler diesel loco from track power to battery powered radio control (see How I converted my Fowler from track to battery power), I have been trying to decide which sound system to install.

I have previously used the rather fine MTroniks DigiSounds module from Peter Spoerer, which uses genuine digitised diesel sounds (see Adding a soundcard to a diesel loco), but I really wanted my two diesel locos to sound a bit different. I then stumbled across the Alan Bond / Technobots programmable soundcard. Although this doesn't use real digitised sounds, I was impressed by the realism of the synthesised sounds it creates in the videos which I found online (eg https://www.youtube.com/watch?v=vsIhu18zT_E ).

The card is neatly housed in a splashproof case as it was originally designed for use in boats.

However, the circuit board was easily removed by undoing four screws.

Although the card has a fairly small footprint, space was tight in my Fowler loco so the first task was to decide where the card could be located. I toyed with the idea of placing two of the 18650 li-ion batteries in the cab as I had done on a previous conversion (see How I converted my Peckett loco to battery power),  however, they would have been more visible in this loco's cab unless I opted for the smaller 14500 batteries - which, of course, would have less capacity and hence would give shorter running times.

Eventually, I decided to install the card in the roof of the cab, where it fitted quite snugly.

The strange protuberances from the ceiling were actually not that noticeable and so I felt I could live with it.

Normally, I locate the speaker under the roof of the cab, to allow the sound to escape easily, but this option was now not available. Fortunately, when I constructed the loco, it was originally equipped with an inexpensive sound decoder and the speaker for this was mounted behind the radiator grille. A suitable 8 ohm speaker was purchased from Maplin (Code L68AA0 and squeezed into the space.

The next decision was how to trigger the soundcard. The Technobots card was designed to fit between a receiver and the ESC (Electronic Speed Controller), the lead from the receiver being plugged into the unit and then another lead (provided) connecting the soundcard to the ESC. The Deltang Rx65b, which is now my preferred receiver, has a combined receiver/ESC on a single circuit board and so the interface between the receiver and the ESC is not accessible. I figured I had four alternative approaches, three of which I tried and one I considered and rejected:
  1. Use a Deltang Rx102 receiver and a third party ESC (I had a spare ESC to hand) and connect the card between the two as per the design
  2. Find some way of dynamically outputting the motor settings from the Deltang Rx65b to the card
  3. Find some way of inputting the PWM signal for the motor from the Deltang Rx65b to the soundcard
  4. Use a Deltang Rx102 receiver to control the soundcard and the Deltang Rx65b receiver/ESC to control the loco motor, making sure both were bound to the transmitter.

  1. Using a Deltang Rx102 and a separate ESC

This was by far the easiest option. The Rx102 was bound to the transmitter and a standard servo connector taken from it to the soundcard. The ESC was wired into the system and connected to the battery and the motor. The soundcard drew its power from the ESC and so everything was installed quite quickly.
However, although the soundcard operated well, I was very disappointed with the level of control provided by the ESC. It was nowhere near as precise and controllable as the Deltang Rx65b, particularly at low speed. As I really enjoy slow running and shunting operations, this was not really something I could live with. I did consider buying another ESC to see if it provided more precise control, but that would risk shelling out more money and still maybe being disappointed with the outcome.

2. Outputting the motor settings from a Deltang Rx65b to the card

As well as its controllability, another great virtue of the Deltang Rx65 is the number of output pads it provides (12), all of which can be re-programmed. Pad 8 on the Rx65b comes ready-programmed to output Channel 1's servo settings. This was therefore ideal, as Channel 1 is used to control the motor speed.

However, in practice I ran into a difficulty. While the output from the pad was sufficient to control the sound when the loco was running forwards, in reverse the card emitted a monotone sound unrelated to the speed of the loco. It looked as if, maybe, the 3.3v output from the receiver was insufficient to be detected by the card when running in reverse.

3. Finding some way of inputting the PWM signal for the motor from the Deltang Rx65b to the soundcard

This solution was feasible. The soundcard is designed to receive input from a potentiometer and so an electronic circuit could be designed to take the PWM (Pulse Width Modulated) output for the motor from the Deltang Rx65b and convert it into something which the soundcard could recognise. However, whilst I had access to expertise (Alan Bond (the soundcard developer) and Greg Hunter (of the Sandstone and Termite Railway)) both of whom volunteered to advise, I felt this was making the project too complicated for replication by others. It would also require a separate battery supply (unless I made the circuit even more complicated) and with space at a premium, it looked as if this approach was going to create more problems than it solved.

4. Using a Deltang Rx102 to control the soundcard and a Deltang Rx65b to control the loco motor

In the end, this is the approach I adopted. I get the best of both worlds - the precise control and the host of features which the Deltang Rx65b brings, together with the ease of installation which the Deltang Rx102 provides. Both receivers are bound to the same Selecta channel on my Deltang Tx22 transmitter and so the Tx simultaneously controls both receivers. A 5v regulator circuit (see below) was needed to provide power for the synthesiser part of the soundcard, but this was a lot less complicated to construct than the circuit needed for option 3 above.

Installation

As indicated above, the card fitted neatly into the roof of the cab and the speaker sat snugly behind the radiator grille. All I needed to do now was wire everything up.

The sound synthesiser section of the soundcard normally uses the BEC input from the ESC to power its electronics and so, as I was not connecting it to a standalone ESC, I needed to provide the card with an alternative 5 volt supply. I constructed a simple circuit using an LM7805 voltage regulator and a capacitor following guidance provided online. Greg suggested adding a diode to the 5v output and then inserting a 1000uF capacitor in place of the 0.1uF ceramic capacitor for additional protection, but I didn't have a suitable capacitor to hand.
Source: http://www3.nd.edu/~lemmon/courses/ee224/web-manual/web-manual/lab8b/node6.html
The components were soldered to a small piece of Vero board, together with the wires for connections (red = 12v input, black = 0v (negative), yellow = 5v output).

The board was then shrouded in heatshrink sleeve to avoid accidental short circuits.

The yellow wire from the regulator board was then soldered to the middle pin marked Ch1 on the soundcard board and the black wire inserted into the -ve terminal on the board. Another black wire was also taken from the -ve terminal and soldered to the outermost pin marked Ch1. The JST servo lead provided with the card was inserted into the other Ch1 socket on the circuit board .......

..... and connected to the Ch1 pins on the Rx102 receiver board. A wire was also soldered to the uppermost (signal) pin for Ch5 (ie pin 5) on the Rx102 board and the other end soldered to the innermost (signal) pin nearest the DIP switch block. This carries the signal from the Bind Button (Channel 5) on my DeltangTx22 transmitter to trigger the horn on the soundcard.

The red wire from the regulator board was inserted into the +ve terminal on the board, together with a red wire leading to the on/off switch on the loco. Another black wire from the loco battery negative supply was squeezed into the -ve terminal on the soundcard alongside the other two wires. The pink and purple wires (I like to colour code my wiring) from the speaker were inserted into the speaker terminals on the board.

 The wiring to the speakers and the battery were taped to the back of the cab, the regulator and Rx102 receiver were tucked into the cavity in front of the driver and the board was positioned beneath the cab roof.

After checking that the card and loco were reliably operational, the loco was reassembled, and then taken out into the garden for a test run.

From some angles the soundcard and its wiring are barely visible (I've just noticed that the nameplate has fallen off ...... must be lurking in the undergrowth somewhere).

The wiring was a little more obvious from another angle ..... so maybe I will site the batteries in the cab after all.

Yes, the spaghetti is somewhat in your face (well the driver's) while a couple of 14500 batteries will be barely visible.

But, for the time being, ...... the proof of the pudding ..........

I turned the volume on the card down to about a quarter of its full setting in the video. The unit is designed to be heard across a large boating pond and so the sound does carry quite well across the garden (and into the neighbours'). Technobots market a smaller unit (at about half the price), which doesn't have the full programming features but does seem to have a similar sound to the one I selected on this unit. It also can be tailored to have 2-6 cylinders, which makes it sound (sorry) quite attractive should I decide to build another diesel loco (I have another motor block sitting on the shelf....).

In the meantime, I am more than pleased with the way this unit behaves. And if ever I get tired of this particular engine sound, I can always reprogram it at no extra cost!

UPDATE - 20/10/15
 I decided that the clutter in the cab was unacceptable and so re-positioned the card and associated electronics in the engine compartment. I left one of the batteries in the compartment  .....

...... and placed the other two in the front corners of the cab, painted dark brown to match the console.

The batteries are hardly noticeable .......

..... a lot less obtrusive than the electronics and wiring were. (Still not found the missing nameplate!).