Fitting Radio Control to an Aster Bulleid Pacific
I fitted radio control to the regulator of my locomotive, and it is successful to the point that a
couple of people have requested information or help about doing the same to their own locomotive.
This summary for the Gauge 1 community may help others who are considering the same. The first few
sections are an overview of what has been done; and these are followed by a Technical Information
section, which provides some more detail for those who can use it.
Go Back to the Top
Go Back to the Contents Links
I decided to give this engine its own dedicated controller, rather than have it share a controller, as do my other engines; the image below shows this dedicated controller.
The controller is in a wooden box and, from left to right, the external items are the on/off switch; the indicator light; the regulator handle; the battery charging port; and the aerial. The photograph is laminated and is removable; other interchangeable photographs are shown.
Inside the box is a Spektrum DX2E 2.4GHz, two channel, surface transmitter; i.e., the cheap one made for racing cars. This system is more than adequate for railway use; and I feel obliged to comment that I never have understood why people use the, considerably more expensive, multi-channel aircraft systems. However, I have changed the regulator potentiometer. Racing cars seems basically to be a full-throttle/full-brake activity, and the throttle potentiometer in the DX2E reflects this: the control arc is quite small at 85 degrees. Initially, I fitted a standard 300 degree, 5K ohm, linear, potentiometer, which gives better control, and is readily available and not expensive. The section Using a Logarithmic Taper Potentiometer contains more on improving regulator control.
The installation uses just one of the two channels available since there is regulator control only. Also, after a trial with all the standard trim controls available, but, it transpired, not used, I cut-off the various control shafts, and hid the second channel control. All of this accounts for the simplicity of what is shown above. I made an acrylic support, tailored to the DX2E electronics board; made a regulator handle; bought some stand-offs for mounting, a charging port, etc.; and mounted the items in the wooden box with very little modification to the DX2E parts. The images below show the bits and pieces, including an acrylic support with a full complement of holes, a lot of which are not used now. The cut-off potentiometer shafts can be seen just under the holes provided for them.
I note that I am writing this summary as I make parts for a second installation. This situation accounts for any minor variations between images.
Go Back to the Contents Links
The next image shows the servo installation in the locomotive cab. As can be seen, there is not much room. This lack of room accounts both for using a Markus Neeser reverser, and for not trying to fit a second servo for reversing, which, probably, would not get much use anyway. The regulator shaft on the locomotive has quite a long cantilever and can be both stiff to move and somewhat sensitive with respect to steam control, which is a poor combination. So my preference is for a servo that has plenty of authority to provide fine control. The chosen servo is a Spektrum S6070, which costs quite a lot, and is a little noisy, but is low-profile (so it fits), has metal gears and plenty of torque. In operation the regulator shaft flexes transversely, because of its long overhang, and this can be a somewhat unsatisfactory control situation. However, the servo is up to the task, and fine control is good, better than manual operation. The two key reasons are that the controller handle is not hot, and the servo puts the regulator handle on the locomotive just where you want it, even on the move.
It is worthwhile to try to minimise any stiffness of the regulator shaft by checking its straightness, and bending it as required. I did improve mine in this way. The shaft unscrews and can be removed completely. However, there is a sealing o-ring in the shaft and the tube into which the shaft fits has a sharp edge at its entry, which can shear the o-ring very neatly. I removed the sharp edge, although the edge is a little hard to get at; and this provides an o-ring lead-in of sorts. Spare o-rings and lubricant are good practice.
The servo requires a bracket to be made, which is bolted to the cab floor with standard Aster screws, albeit a little longer than those specified in the assembly manual.
The crank on the servo is a standard plastic item that is supplied with the servo; it is shortened and trimmed to avoid fouling the cab footplate. The crank on the regulator shaft is the one supplied by Aster (the red one, painted black here), although the clamping screws may need to be replaced to ensure adequate grip on the shaft. The adjustable link connecting the cranks is made from brass strip and stainless steel screws. Two screws are mounted in a small brass cradle and then silver-soldered together to provide the necessary right angle configuration. This link, and its pivot connections, must have good structural integrity since failure means total regulator control failure.
The final item is the electrical connection. This is the block on top of the servo, immediately above the servo shaft, and showing white scuff marks. This, and its mating part, were taken from an old computer motherboard. The connector is glued to the top of the servo, which is not a particularly good arrangement: I have yet to find a glue that will do a study job holding together the greasy plastics. It works, but will not withstand a knock. It is important to ensure that the servo crank can not hit the connection during operation; that will cause the glue to fail too.
Go Back to the Contents Links
What remains is to mount the receiver, batteries, charging port, and on/off switch. I put them in the tender and the electrical connection to the servo runs underneath the tender floor next to the fuel and water lines. I cut acrylic pieces that clip together to form a box inside the tender. The fuel tank slides into this box, and the electrical stuff is scattered around in the gap between the acrylic box and the metal tender sides. The on/off switch, and the charging port are mounted on the tender front. These two items need to be carefully positioned because there is not much room inside the tender, and they must not foul placement of the receiver or batteries.
The image shows my installation of the on/off switch; charging port; and electrical connection to the servo, which is sheathed in heat-shrink tubing. At the bottom of the image can be seen the water pump return, fuel quick disconnect (non-standard), and water supply line from the tender pump.
The next two images show the acrylic box, both free-standing and in-place. The box is not fully self-supporting, as can be seen by careful examination of the free-standing image; it requires support from the tender. Other than the box installation, the thing to note from the in-place image is the electrical line coming from underneath the tender bottom. This is the other end of the line to the servo. In use, this line passes through the opening at the bottom of the acrylic box, left in the image, and plugs into the receiver. With that exception, this image shows, in essence, how the installation looks when complete.
The next image shows what remains to be installed. There are five rechargable batteries in series, which gives 6 volts. These batteries are divided into two packs, the pack of three goes on one side of the tender (on/off switch side in my installation); the pack of two goes on the other side, together with the receiver. The tack putty is how I prevent the receiver bouncing around: I stick it to the side of the tender as far forward as possible, with the ports forward and at the bottom. The leads coming out of the tender are from the on/off switch; one connects to the batteries, the other plugs into the receiver BIND port to supply power to the system. I have had no problem with simply letting the aerial dangle inside the tender; but, if long range operation is needed, this might be a problem.
The next image shows a set of new parts ready for installation. The energy flow can be traced in the image, from the batteries, through the on/off switch and the receiver, with grey aerial, to the servo plug. The charging port also connects to the on/off switch, but the junctions cannot be soldered until the port is installed in the tender. The servo connection has been painted with plastic paint because it is on display when the locomotive is operating; underneath the paint it is the same as the other wires. This is different from my original connecting wire, which is sheathed in heat-shrink tubing to hide its colours; currently, I prefer the painted version because it is more flexible. The servo plug is a scavenged computer plug, conveniently having three connectors. Once the junctions were soldered the plug was potted with epoxy to give the plug and junctions some mechanical durability.
Go Back to the Contents Links
As a result of making a second set of parts for someone else, I have been forced into re-addressing details that otherwise would have disappeared in the mists of time. So I am creating this set of notes for any future need. But first some encouragement.
Words to the Wise: I think that I have described a way to fit these parts that works. However, it did not come easy, and it might not for you. Indeed, those of a religious bent, attempting the detailed work, might discover clear evidence for the existence of the Devil. All I can do for you is quote a famous man:
"Nothing in the world can take the place of persistence.
Talent will not; nothing is more common than unsuccessful men
with talent. Genius will not; unrewarded genius is almost a
proverb. Education will not; the world is full of educated
derilicts. Persistence and determination alone are omnipotent.
The slogan 'Press On' has solved and always will solve the
problems of the human race."
Calvin Coolidge
Go Back to the Contents Links
The support for the Spektrum DX2E electronics board and the parts for the box that contain the tender fuel tank, I produced in CAD software and cut with a laser machine. Laser cutting is a very satisfactory way of making the parts since it avoids splits, rough edges, etc., and is trivially repeatable. If anyone would like these parts for non-commercial use, then I shall be happy to supply them at nominal cost. If someone with commercial interest would like to have the CAD files, then that can be arranged. Contact Information.
My guess is that the tender box will be useful to anyone doing an installation involving putting radio parts in the tender. However, the other part is not only highly Brand-specific, but may not work if Spektrum changes the design of the DX2E.
Go Back to the Contents Links
Installing a 300 degree arc linear potentiometer is an improvement over the standard Spektrum potentiometer, as indicated in the section on the Controller. However, by fitting a potentiometer with a logarithmic taper (aka audio taper) the control over small regulator openings can be spread-out over a larger controller arc. There is a small squish-in sacrifice at large regulator openings, but this is barely noticable. Using a logarithmic instead of a linear taper gives a useful improvement of control over locomotive performance. The reason is that the electronic spread-out effect works counter to the mechanical squished-in effect. In other words, at small regulator openings a small change in regulator setting has a relatively large effect on steam flow; but at those small openings a relatively large potentiometer setting change is needed to effect regulator movement. Another way to look at this is to see the electronic logarithmic spread providing precision to the mechanical regulator movement where it is needed, with the precision fading away as it becomes less needed.
However, on installing and testing a logarithmic potentiometer I found that there was too much dead movement at the beginning of the controller arc. I looked into this, and what follows is a record of my investigations.
The standard potentiometer has an overall resistance of 5K ohms and the connections are to orange(O) and green(G) wires; the centre tap is connected to a yellow(Y) wire. I found that the resistance range of O-Y from initial servo/regulator movement to full movement was 1K-3.8K. This is a phenomenological result, obtained by disconnecting wires and using an ohmmeter; I have no knowledge of the controller circuitry. The initial dead movement of the controller handle, for any potentiometer, is explained by the O-Y resistance going from 0 to 1K. But a 5K logarithmic potentiometer takes a bigger arc for this change than a 5K linear potentiometer; hence the exaggerated dead arc with the former. I decided that I could use the potentiometer arc to resistance range mapping information to modify the controller.
The modification is to have a fixed resistance of 1K on the O side of the potentiometer, and another fixed resistance across the potentiometer O-G connections, sized to reduce the overall O-G resistance to 4K. This sizing is calculated using the usual equation R = 1/(1/Rp + 1/Rf), where R is the required 4K, Rp is the potentiometer resistance 5K, and Rf is the required fixed resistance. A practical value for Rf is 18K. Another practical pair of values is an O side fixed resistance of 1.2K, and Rf = 15K. For simplicity, no account of the 3.8K full movement resistance is made.
A specific modification recipe is as follows.
- Record, in writing, the wiring connections, in detail, before unsoldering anything! You know why.
- Unsolder the potentiometer.
- Establish which end of the potentiometer increases its resistance O-Y as the shaft is rotated in the "open regulator" direction. This is what is referred to above as the orange(O) end. The other end is the green(G) end. You can measure this resistance with an ohmmeter, or you can risk it and assume one end. If the controller still has a large dead movement when you have finished, then you guessed the wrong end; change it. Note that there may be some dead movement because of the way the potentiometer is made.
- Remove the standard potentiometer and replace with a logarithmic potentiometer with the same overall resistance. If this resistance is not 5K, it will be necessary to re-calculate the values of the fixed resistances to be installed using the equation above. If you keep the overall resistance the same as the original, you are not likely to run into trouble. The values that follow assume a 5K resistance.
- Re-establish the O connection by soldering-in a 1K resistance. This resistance is not a special item, but is a standard electronics part with brown-black-red bands around it.
- Solder a 18K resistance (brown-grey-orange) across the potentiometer O-G. Re-establish the G connection by soldering the wire to the potentiometer at the same time.
- Re-establish the Y connection by soldering the Y wire to the potentiometer centre tap.
Go Back to the Contents Links
The image below is of a finished controller installation, but with the photograph, handle, and other external parts removed. The view is through the acrylic mounting plate, everything shown is underneath this plate, except the potentiometer shaft and charging port, which poke through the plate, and two mounting screws in the corner of the box. The image shows details, e.g., circuit-modifying resistances and stand-offs, described in various sections of these notes.
Go Back to the Contents Links
The image shows new components made for the servo installation prior to painting. The electrical connections at the back of the servo are shown after soldering, installation of heat-shrink tubing, and potting of the wires in epoxy. The angling of the white connector is intentional: there is not much room available, and the angle-offset avoids having the regulator actuation link clout the connector at full servo displacement. The mounting bracket was made from 1/16 inch sheet brass; the servo mounting holes are tapped, but the holes for mounting to the footplate are clearance holes for the Aster screws used for attachment. The mounting bracket can be made from other materials, angle rather than sheet is rather obvious; however, it must be strong and stiff enough for the job. The bearing at the end of the brass actuation link is two washers soft soldered to the link; the hole is reamed to accept the Aster regulator pin; mine is 0.093 inch (#42) diameter. The adjuster-cum-right-angle, which attaches to the plastic servo crank, is made from two stainless steel screws and a couple of washers, silver soldered together. Hard solder is used on the latter component for strength; failure of this joint means complete regulator control failure. The screws are 0-80 unified, which is close to 11BA, and 1.2mm diameter metric. An assembly of similar components is show in this cab image.
Go Back to the Contents Links
There are a number of issues to consider under this heading. One is the mechanical adjustment of the link connecting the servo to the regulator handle. Another is the mapping between the controller potentiometer movement and the servo movement; closely associated with this is the attachment position of the controller handle on the potentiometer shaft.
A specific adjustment recipe is as follows.
- Bind the transmitter and receiver with the controller potentiometer shaft fully counter-clockwise, and the servo shaft fully clockwise. These are the normal "closed" positions: the regulator opens counter-clockwise because the regulator thread is right-handed; human-centred controls usually open by turning them clockwise.
- With the transmitter and receiver turned on so that the system is working, turn the controller potentiometer shaft clockwise whilst watching the servo shaft. Find the maximum servo movement position. Now install the controller handle. The controller handle should be in the position corresponding conceptually to maximum open regulator. Ideally, the controller handle should be against a mechanical stop.
- Turn the controller potentiometer handle counter-clockwise whilst watching the servo shaft. Find the minimum servo movement position. This action puts the controller handle in the position corresponding conceptually to closed regulator. If needed, control of this position is changable by altering the resistances installed under Using a Logarithmic Taper Potentiometer, which will alter the arc length of the controller handle.
- Leave the controller position in the "regulator closed" position, and close the locomotive regulator. Install the link between the servo and regulator handle, adjusting the link as required. The link is shown in the regulator closed position in this cab image.
- The adjustments are complete and the regulator should be controlled effectively by the controller handle; the effect of the logarithmic taper will be evident, and may be a little disconcerting! The servo movement arc is around 90 degrees, and that is about all that can be used because of the limitations of the type of linkage used. This amount of movement normally is adequate for locomotive control.
- Check all locknuts, etc. Lubricate the regulator link using plastic-compatible oil.
Go Back to the Contents Links
- If you choose to use a Spektrum DX2E, you might like to know that there are 10 screws holding together the controller, 3 of which are hidden behind stick-on labels. Remove all ten, and it comes apart easily.
- It should be noted that the controller potentiometer modification in section Using a Logarithmic Taper Potentiometer does not change the total servo arc. If it appears that this has happened, then the modification resistances are wrong for the specific hardware in use.
- The installation recipe in Control Adjustment assumes a normal control direction; but you are not bound to follow this convention. You may prefer to have the controller handle mimic the regulator handle, for example. The Spektrum DX2E has a reverse operation switch which will effect this change, but it may be necessary to re-wire the O and G resistance connections referred to in Using a Logarithmic Taper Potentiometer.
- If more than about 90 degrees of regulator movement is desired, then note the availability of servos for model sailing boats which provide a few turns of movement, i.e., winches. On the face of it, these offer a lot of possibilities for regulator, and other, control using gears, chains, etc.
Go Back to the Contents Links
This radio control installation is well worthwhile in my opinion; fine regulator control is superior to that under manual control. However, it must be acknowledged that I am heavily inclined to support actual control of moving artifacts. Opening a locomotive regulator and watching what happens is a little too much like lighting the blue touch paper and retiring for my liking. Also, my railway has significant gradients, which more or less demand remote control, driving the work reported here.
For those considering radio control of their own locomotive, I think that what is here gives a fair idea of the amount and type of work required: not a huge task, but not trivial either, and some technician skills are required. If there is no reason or preference to do otherwise, closely following what is here is likely to be successful.
If questions arise that I may be able to answer, then I can be contacted per the Contact Information. I shall be interested to hear or read of other radio control installations based on these notes.
Go Back to the Contents Links
last-modification-date: 7 Sep 2013