How I used a DF Player Mini module as a simple soundcard

 Introduction

After my IP Engineering Plate Frame Simplex loco's cheap greetings card sound module failed (see How I modified a cheap greetings card module as a sound card), I wanted something similarly cheap but more reliable as a replacement. Enter the DF Player Mini.

The DF Player Mini sound module has been around for quite a few years. It is modestly priced (c£3.50GBP on eBay or less than £2.00GBP on AliExpress), and is very versatile.

Basically, it is a compact MP3 player which can access tracks stored on a Micro SD card. What makes it particularly useful for use as a soundcard is that it has a built-in audio amplifier which will power a 3W speaker. Its operational voltage range is 3.2v - 5v and so will need to be connected to a voltage regulator if used with locos with a higher voltage battery pack (or track power). 

 VCC is the positive (+ve) input and either of the GND pins are the negative (-ve) input. The speaker is connected to SPK_1 and SPK_2. 

Tracks stored on the SD card are played in the order in which they were uploaded to the card and need to be numbered 001, 002, 003, etc. They can be in .MP3 or .WAV format. .WAV files are better for our purposes as they can be accessed or looped with a hardly noticeable gap between them. The module can accept cards up to 32Gb and so can potentially access thousands of sound files. However, I only needed a couple of files and so used a 521Mb card. 

It has been designed to interface with microprocessors such as Arduino but, for my purposes, sound files on the SD card can be triggered by connecting various pins to Ground (ie 0v or the negative side of the battery used to power it).

For simple soundcard purposes, the most useful pins are:

IO1 - Short press = Play previous track - Long press = Decrease volume

IO2 - Short press = Play next track - Long press = Increase volume

ADKEY_1 = Play first track (and loop if held down)

ADKEY_2 = Play fifth track (and loop it)

 

However, by connecting resistors to the ADKEYs particular tracks can be played or functions selected.

 

Armed with this knowledge, I was now in a position to create a simple soundcard for my IP Engineering Simplex loco.

 

Creating the sound files

 Firstly, I needed to track down some suitable sound files. I would need the sound of a diesel loco similar to a Simplex running at a fairly constant speed, and also a horn sound.

There are various sound libraries on the internet where you can download sounds by paying a small fee. I used Sound fishing which is quite reasonably priced (eg 6 Euros for a month) - https://www.soundfishing.eu/premium.html 

Once I had found a suitable sound file, I then highlighted a section of the recording which I felt was most appropriate for my needs, .......

 .... and exported that Current Selection as a Mono WAV file.

I could now store this file on an SD Card (as 001) and then loop it using ADKEY_1 connected to ground to provide continuous sound for my loco.

 

 Adding a horn sound

However, I also wanted to add a hooter or horn sound, triggered by pressing the Bind button on any of my Deltang or Micron transmitters.

After tracking down a section of video which featured a typical Simplex loco horn sound, I exported the short segment from Audacity (as above) and then re-opened the file of the steady Simplex engine sound which I had saved earlier.

 I then imported .......

..... the short segment of horn sound which appeared on a new track below the steady engine sound track.

I moved the horn sound track slightly to the right and then trimmed the steady engine sound so it was slightly longer than the horn sound track. As you can see, the combined tracks last slightly under one second in total.

 The new combined sounds were then exported as a WAV file.

The reason for overlaying the horn sound on to the steady engine note is so that when the horn sounds, the engine sound continues in the background.

To allow the two files to be played and triggered, I had to make use of the ADKEYs ability to trigger separate tracks using resistor inputs. This also required the tracks on the SD card to be arranged accordingly.

I found that, not only did I need to delete the original WAV file, I also had to reformat the SD card otherwise the sound files wouldn't play properly. The horn sound was loaded first on to the card and renamed 001.wav . The steady engine sound was uploaded second and renamed 002.wav .

 The link from GND to ADKEY_1 was replaced with a 3k resistor (so Track 2 on the card would be played continuosly) and a link from ADKEY_1 to pad C on the Deltang Rx65b receiver was made. Pad C gives 0v output when the bind button is pressed on the transmitter (ie Channel 5)

So, when the loco is first turned on, the 3k resistor from ground to ADKEY_1 makes DF Player loop Track 2 (the steady engine sound) but when the bind button is pressed on the transmitter, the receiver connects ADKEY_1 to ground directly and so, while the bind button is being held down, track 1 (ie the horn sound) is looped.

Here's how it sounds.

 

Further DF Player Mini Projects

• A series of tracks triggered with the transmitter 
• Automatically triggered tracks dependent on speed - Part 1 (Bluetooth)
• Automatically triggered tracks dependent on speed - Part 2 (Arduino)

 

 

How I converted Bachmann Big Hauler 4-6-0 locos to battery power and radio control

 I was asked if I could convert a couple of Bachmann Big Hauler 4-6-0 locos to battery power and radio control using Micron radio control equipment. I hadn't realised, until they arrived, that there are at least six different "Generations" of this popular model - see https://www.girr.org/girr/tips/tips1/big_hauler_tips.html . It turned out that the two models I had been sent were Generation 2 (an Atchison, Topeka & Santa Fe loco) ......

 and a Generation 4 (White Pass) version:

Externally, apart from the colour schemes, there isn't a great deal of difference between them, though the quality of the fittings and the plastic used in the Gen 2 loco as not as good as in the later Gen 4 model. The main difference between them is internal. Over the years, it seems that Bachmann developed at least six different mechanisms for the loco - as will become evident during the dismantle and conversion processes below.

 

Dismantling Loco

Fortunately, Bachmann have made dismantling fairly straightforward on these locos. The first job was to remove the baseplate holding the driving wheels in place. The three screws are fairly obvious....

 

 

 A fourth, longer screw, is between the cylinders. This holds the cylinders in place as well as the baseplate.

 With all four screws removed ......

.... the baseplate could be taken off to reveal the mechanism. Here is the Gen 2 version which, as you can see, is a series of spur gears driven by a crown and pinion on the motor shaft. This mechanism is very noisy when being run, especially at high speed.

The Gen 4 mechanism, is much simpler. A straightforward worm and wormwheel.

As we would no longer need to pick up power from the rails, I snipped off the leads from the pilot bogie at this stage. The next job was to detach the connecting rods from the centre wheels. Plastic plugs hold these in place. They were carefully prised off. I used a pair of snips to very gently lever them upwards before using a pair of needle nosed pliers to finish their extraction. 

 The cylinders could now be removed.

At this stage, the wheels can be removed. As you will see, I did this on one loco, but left that until a later stage with the other loco. The wheels are mounted in slots on the chassis and simply lift out.

Also, the pilot bogie can be removed at this stage by undoing the screw in top of the pillar on which it pivots. Alternatively, it can be left in place.

 To remove the body from the chassis, six screws were then removed - two beside the footplate at the rear of the loco body ......

 .... and four towards the front, behind the low slung air tanks either side of the boiler.

 The two metal stays at the front of the loco .....

...... were then pulled out from above the cow-catcher (they are a push-fit).

 The main body could now be removed.

As I was going to completely rewire the loco, I snipped off the two wires feeding the smoke unit and the headlamp at this stage.

The screws holding the iron weight in place were removed 

 Note a slight difference in the way the weights were mounted in the two locos.

With the Gen 4 loco, I decided to remove the motor and its mechanism as a unit. This was held in place by a steel rod passing through the sides of the chassis. 

It was pushed through using a nail and some gentle tapping with a small hammer.

 With the Gen 2 loco, I left the gears in place and just removed the motor, by undoing a couple of screws.

The copper pick-up strips were soldered directly to the motor contacts and so these, plus the wiring, were removed with a soldering iron.

 The pick-ups on the Gen 4 loco were slightly more sophisticated plunger-style fittings. The two tiny screws holding each fitting in place were removed and the fittings taken off.

 The two screws holding the circuit board in the loco cab were unscrewed ......

 

 

... the unit was removed, snipping off the associated wiring.

 

Rewiring the loco 

 I decided to start at the front and work my way back. The smokebox door assembly was unclipped .....

.... and the wiring on the back of the smoke unit switch removed with a soldering iron.

 

 The headlamp was unclipped ......

..... and the filament bulb removed.

A 3mm warm white LED was put in its place - one leg was insulated and the other left bare, so the leads would fit into the tube which held the lamp in place.

The LED was bent to face forward before the lamp housing was put over it.

Ribbon cable ........

..... was soldered to the positive and negative contacts for the smoke unit switch and for the positive lead of the LED - the negative being shared with the smoke unit.

A 4-pin JST socket was attached to the other end of the ribbon cable (a 3-pin would have sufficed, but I didn't have one to spare).

 

More ribbon cable was connected to a 4-pin plug with an extra three cables (one for the chuff cam and two for the motor) leading to the rear of the loco (the negative input for the chuff cam was taken from the feed for the smoke unit).

The negative feed and return wires were soldered to the chuff cam contacts, ......

..... and the other two wires to the motor terminals, .......

..... before the motor was screwed .......

..... or pinned back into place.

The circuit board at the back of the cab was dismantled ......

..... and a piece of Vero stripboard, cut and drilled to the same dimensions, ......

..... before being bolted to the bracket.

The ribbon cable was then soldered to the Veroboard strips ......

..... and an 8-pin JST socket soldered on (a 6-pin socket would have sufficed)

The bracket was then screwed back into place at the rear of the chassis.

A slot needed to be cut into the back of the cab body to give clearance for the plug. I used a slitting disk in a mini-drill for this.

The loco body and wheels could now be reassembled.

 

 The tender - dismantling

 Four screws hold the tender body to its chassis. They are readily apparent when the tender is turned upside down. These were removed .......

 .... and the body simply lifted off.

 

 I decided to leave the speaker in place, but the soundboard was unscrewed and its mounting pillars taken off with a razor saw.

 The contacts for the battery  .......

..... were slid out, ......

.... and the mouldings ........

.... removed with a razor saw.

The chuff-cam lead from the sound card was snipped off and removed.

A mounting bracket for the switch and charge sockets was designed using Tinkercad and 3D printed (the .stl file for this is available as a free download on the gardenrails.org forum.

A 4-pin JST plug was soldered to another piece of Vero stripboard .....

... and a 3-position SPDT switch and a DC charge socket acquired.

The switch .......

.... and the DC socket were wired-up ........

.... before being inserted, along with the 4-pin JST plug, into the mounting bracket.

 

The lead from one pin of the switch was then soldered to the positive terminal of the DC socket.  

The Micron MR603c receiver/controller was wired-up - with leads connected to the positive and negative input pads, the motor pads and the C, D, P1, P2, P3, P4, P5 and P6 pads. P1 and P2 are for the headlamp and the tail lamp (for future attachment) and pads C, D, P3, P4, P5 and P6 are for the triggers for the sound card.

The trigger wires were then soldered to the trigger inputs on the Micron Mini Sound Module, together with shared positive, negative and motor power leads (see circuit diagram below for more detail).

The mounting bracket was then screwed to the tender chassis, being carefully positioned so the switch and sockets would be directly below the flap previously used as the opening for the 9v battery compartment.

 

It was screwed rather than glued in case any additional adjustments were required at a later date. 

 A battery pack was made up from three 18650 li-ion cells, connected to a 3S BMS (Battery Management System) board and to the balance charge socket (see circuit diagram below). NOTE: Care must be taken when connecting the cells together to prevent accidental short circuits. Li-ion cells are extremely volatile and should be handled with caution. The battery pack was held in place with a couple of cable ties passed through holes drilled in the floor of the tender.

An 8-way JST plug was wired up to the postive, negative, motor power, chuff cam input to the sound module, and to the headlamp output from the MR603c (with a 150R resistor connected in series) (see circuit diagram below), with the wires passing through holes in the front of the tender and the floor

 

Once the connections had been made, the last few centimetres of the cables were shrouded in black heatshrink wrap.

NOTE: Initially, I passed the cable linking the tender to the main loco, through the original hole in the front of the tender, however, because the cable is thicker it sometimes pulled the loco or tender off the rails, and so I made a new hole in the centre of the tender's buffer beam, which works more reliably.

Circuit diagram

Click to enlarge

 

 Reprogramming the MR603 receiver/controller

 The owner of the locos had also purchased a Tx22X transmitter from Micron, which provides a series of outputs using two-way switches and push buttons. 

 

 Making the most of these outputs, I needed to repogram the receiver so it would trigger the soundcard as follows:

•  S1 - A - "All Aboard" 
• S1 - B - Bell
• S2 - Safety valve
• S3 - C - Coal shovelling / Injector
• S3 - D - Water pump
• Bind button - Whistle

 This would entail changing the default outputs from the receiver pads:

• P1 - Front light (Retain)
• P2 - Rear light (Retain for future connection)
• P3 - 0v when Channel 3 is low
• P4 - 0v when Channel 3 is high 
• P5 - 0v when Channel 4 is low
• A - Front light (Retain for future)
• B - Rear Light (Retain for future)
• C - 0v when Channel 7 is low
• D - 0v when Channel 7 is high

For more information on reprogramming Micron receivers see - How I reprogrammed a Micron MR603 receiver

For more information on reprogramming Deltang receivers see - How I  reprogrammed a Deltang Rx65 receiver

 

 Re-assembly, tweaking and testing

The tender was then reassembled.

and access to the switch and charge sockets through the tender tank filler cap was checked.

Following advice from https://www.girr.org/girr/tips/tips1/big_hauler_tips.html, I added some lead weights to the pilot bogie cut from lead flashing.

It seems that, as supplied, the bogies have a tendency to derail. In addition, on the White Pass loco, I replaced the original spring with a stronger one - made by wrapping phosphor bronze wire around a piece of brass bar of the right diameter.

 I also noticed that some of the brake blocks were missing from both locos, so I designed some replacements in Tinkercad and 3D printed them (Files available as free downloads from the gardenrails.org forum).

 

The locos were then given a few test runs

 Conclusion

 On paper, the conversion of a Bachmann Big Hauler ten wheeler is relatively straightforward. However, the differences in the design of the mechanisms add some complications. In addition, the need to transfer connections from the tender to the main loco adds another level of complication. This could be reduced if the smoke generator and chuff cam aren't used.

Overall, the models are quite easy to dismantle when compared to some other G Scale locomotives and there is plenty of space in the tender to hide the batteries and control equipment. Those who are wary of lithium batteries may prefer to use the space for a NiMh battery pack.