Sunday, 20 December 2009

Skiing in Dedham Vale




Dedham Vale is perhaps best know for the paintings of John Constable (Haywain, etc). Its a few minutes from my home, and for the first time ever I've been able to ski on the meadows leading down to the Stour. This meadow is on the north bank, Dedham church is visible in the background of the first picture. Sorry that the pictures are a bit blurry, there is only so much that a camera on a cellphone can do.






Tuesday, 24 November 2009

Couplings on 0-6-0T

This loco belongs to the East Anglian Scalefour group and is used on Coldfair Green.



I have fitted DCC controlled AJ couplings to the loco using my etched parts.

I built up the couplings as complete units on a piece of printed circuit board. This was then fitted to the loco using the chassis mounting screws. Wires from the board went up to the loco, via some small PCB plug/sockets which allow the couplings to be electrically removed.



Coupling in closed (normal) position. Some parts temporarily held with "whitetac", araldite (or similar) will be used when happy that no further adjustments are required.




Coupling in open position, with magnet clearly visible on bottom of flap.





Close-up of front of loco, counterweight is visible below buffer beam, but fairly unobtrusive. Loco now requires cosmetic 3-link or screw coupling elements to be hung alongside AJ wire.

Yes, I am bending AJ hooks the "wrong" way. I have started experimenting with Vincent de Bode's method which puts the double section downwards (the downwards part still pulls the train load). Vincent's method offers the possibility of finer wire (the train load is no longer taken by the 180 degree bend) which in turn means the wire can be shorter for the same degree of springiness. It also means that the lower tail of a coupling cannot catch on a burr (there is no cut at the lower edge). The "Vincent bends" are compatible with conventional couplers, so its possible to mix and match between different types. Its too early for me to tell whether I will swap over wholesale to Vincent's method, but it shows promise.

Thursday, 17 September 2009

Etched brackets for DCC controlled AJ's

I've just taken delivery of a sheet of etched bits for the DCC Alex Jackson couplings. There was a fair bit of interest in them at Scaleforum, and I will be ordering a batch of coils in mid October.

If anyone would like to experiment with these, get in touch. There are pictures of the prototype etches in a small 4mm scale 0-4-0 posted earlier. There is some video of the 2mm version in action, and a list of CV values on the 2mm Scale Association website.

One etch per coupling (more if you are creative with the parts).
Prices £2 each, 6 for £10, 12 for £20 (UK post free), 20 for £30 (UK post free). Small orders post is £0.50 in the UK.



The etch picture shows the components;
top left - box to hold coil for rotating Alex Jackson (and similar rotation designs). Approx 4mm cube with "ear" for rotating shaft above.
bottom row - three different flap sizes
top right - base for DG coupling (2mm/N)
centre right - tongue for DG coupling
centre top - alternative brackets to box for coil (for very tight places).


Folded up box with simple flap (no counterweight).




Folded up DG version.



Sunday, 13 September 2009

Snape diorama for Scaleforum


Scaleforum, an annual exhibition of 4mm scale models, is held on the last weekend of September.
The picture shows a diorama with some of the buildings from Snape Maltings (buildings and painting by Richard Kimberling) which will be used to illustrate my DCC controlled Alex Jackson couplings.
The diorama shows only a small section of a much larger building, set into a temporary roadway.

Sunday, 6 September 2009

2mm Scale Alex Jackson's

The Alex Jackson coupling is well known in 4mm scale, but less common in other scales.

Jim Watt is the main user in 2mm scale, and he sent me a few for testing with my 2mm loco with DCC controlled couplings. The couplings are very small and unobtrusive, here are a couple of photos of them fitted to a 21T mineral wagon. Since taking the pictures, I've noticed some clown has knocked the corner of the buffer beam from the wagon, so more repairs will be required. The buffer heads are my own steel turnings.




Saturday, 20 June 2009

ABC Stopping/Slowing

( Post updated 19th July with some clues on how BM2 works. Updated 21st Sept with CT Sound chip information).
Lenz and Zimo DCC chips (* below) support an automated stopping method called ABC (assymetric braking control). This allows a very simple trackside circuit to cause a train to slow to a stop. The train can be reversed away from the stop signal. There are differences in how each company has implemented ABC.

I've got interested in ABC for two uses. Firstly, an automatic shuttling railbus which can be set to run from fiddleyard to platform, and back again on a timed sequence using features within the Lenz Gold chip. This automation requires nothing more than two ABC stopping modules. Secondly, using ABC to protect the end of the running area for a small demonstration track. The ABC module will make it impossible for a visitor to drive a locomotive into the buffers or off the end of the track.

Both decoder makers offer an option of constant braking distance. This means a train will always stop in the same place regardless of approach speed, which means it will stop ahead of a signal.

Lenz's version of ABC has two modes, stopping and slowing. They sell modules (BM1 for stopping and BM2 for slowing) to trigger the effect, though they seem quite expensive. The stop module (BM1) is well documented as an array of diodes, and a DIY one is trivial to construct. The exact number of diodes seems to vary between different websites. My experiments have suggested six (diagram on left) are required for a reliable stop with a Lenz Gold.

I have been unable to find details of the slowing module. An email sent to me indicated that the BM2 works by sending the assymmetric signal on alternate cycles of the DCC signal. This would make a home-brew BM2 somewhat more complicated than the home-brew BM1. (I have tried a four diode matrix, and that seemed to cause my solitary Lenz decoder to slow correctly, but it might be an artifact I'm seeing, rather than correct operation given the comments of my email correspondant).

A "slow approach" could be deployed at a caution (distant) signal, and a "stop" ahead of a stop signal. A simple on/off switch is used to disable the stopping or slowing module.



Lenz' implementation of constant braking distance is a bit limiting.

  • Firstly, its an all or nothing setting, so it can be quite awkward to manually drive a locomotive with constant braking distance enabled as stops will always take the same (long) distance from the place the throttle moves to zero (!). Judicious use of F4 to turn on the shunting mode (disables acceleration and constant distance braking) can give some manual control. One can use Ops-mode programming to turn constant distance braking on and off, but is a pain in the neck to keep doing that. The net result is a loco has to be set for constant-distance stopping combined with automatic running.
  • Secondly, the manner of stopping seems to be an initial decelleration at the programmed decelleration rate (CV4) followed by a creeping crawl up to the stop line. If the speed is too high for the stopping distance, the loco decellerates following CV4 until it reaches the distance and then stops dead from whatever speed it was doing. Not nice in either case.
Lenz do have a unique useful "shuttle" feature which can be used to automate a railcar, multiple unit or push-pull train. There are a few minor glitches in the Lenz implementation, notably the "direction" of the locomotive can be reversed depending when manual control is taken back. But it does work and its a very low cost of building an automatic shuttle. Use of Ops-Mode CV changes to turn the shuttle on and off is recommended if the locomotive is to also be used manually.



Zimo doesn't offer a "slow approach" mode for ABC (they do have their own HLU system, which I think is considerably more powerful, but I think only supported on Zimo command stations), but in all other respects has far more user-settings which allow detailed customisation.
Zimo allows the voltage assymetry to be set in a CV, this allows a grid of only four diodes to trigger a "stop", which means the track voltage drop is lower. This does mean that if Lenz and Zimo chips are mixed, there can be a problem in Zimo chips slowing noticeably on crossing the braking zone margins.

Zimo also have a much better implementation of "constant braking distance".

  • Firstly, the user can set "constant braking" for only automatic stops (ie. ABC stops) leaving throttle stops at the rates specified in CV4. (There is an option for all stops to be constant distance).

  • Secondly, the deceleration is done far more effectively; the train drives at its initial speed until the decoder decides it is time to slow to reach the stopping point. The visual effect is far better, it looks as if the train driver has chosen the right time to apply the brakes depending on approach speed. Alternatively, a Zimo can be set to decellerate smoothly using a rate calculated from the speed when the loco enters the stopping zone. This mimics a driver applying gentle braking from a low speed or harder braking from higher speeds.

So, that's my quick summary of ABC Braking experiments. I'd like some more details of the inner workings of the BM2 if they are available.

(*) 21st September. For another project, I've been experimenting with a CT Elektronik SL75 sound chip. This tiny chip has Lenz ABC braking support, and it works on my test rig (6 diodes, diode sensitivity set to 15 in the CT CV's). I am editing the JMRI file for this chip, so this detail should appear along with a large number of other changes. I haven't found ABC to work in other small CT chips, such as the DCX74 or DCX75.

Friday, 19 June 2009

L&Y 0-4-0 Electric Nearing Completion

Progress continues on the L&Y battery electric, hopefully nearly done apart from lining (which was a bit over-the-top full-on swirly bits on the prototype!).

Firstly, some sprung buffers. The photo shows an original whitemetal casting (rear), and a bored out casting (middle) and a couple of steel heads and a buffer spring. All done on a lathe, the castings were bored after making a split collar to hold the casting in a collet. The spring is 36SWG phos-bronze, wound around a 1mm dia drill bit. The fat end of the buffer is 2.4mm dia. The buffers don't yet show the cross-drilling (0.35mm) in the thin tails (0.8mm dia) for the locking pin which fits behind the buffer beam. There is one pin for each end of the loco, which goes across the buffer beam, and through the shanks of both buffers.


Secondly further video of the loco uncoupling with the DCC controlled AJ's. As shown on earlier blogs, this is the first 4mm loco built with the couplings in mind, which was a lot easier and neater than retro-fitting. Chip is a Zimo MX620, activating F3 causes the entire "backup, uncouple, pull away, stop" routine. There is still some more electronics to add; a bank of capacitors to provide power over dirty track, and I've started to consider lamps fed by fibre optic.




Thursday, 21 May 2009

Decoder Comparison; TCS vs Lenz

Lets begin with the prices. Its an unfair comparison. A TCS MC2 costs under £20. A Lenz Gold plus USP module costs nearly £70. So, those with a restricted budget, stop reading now.

I have a 4mm scale railbus, its an Airfix kit with a Branchlines mechanism and interior. A Mashima motor drives one axle, the other has simple compensation. Wiper pickups on the top of the wheels. For a while this has run with the TCS chip; its seemed adequate for the job, but there have been shortcomings.

The biggest problem is not down to the chip, its the light weight four wheeled vehicle. The pickup can be erratic on all but the cleanest track. So, as an occaisional use vehicle on an exhibition layout, its a problem if it stalls on the crossing into the platform ! The TCS chip does exaggerate this problem to an extent; once power is lost, the chip restarts from speed zero, doing all the acceleration again. The TCS chip control is OK, the motor runs a little noisily, but its OK. The start speed is a little on the high side, but again, its OK.

So, why replace a £20 chip with a pair of devices costing £70 ?

One way round a pickup problem is a UPS capacitor to store power when there is a break in power to keep the chip alive. The store need only last for a tiny fraction of a second to overcome a microscopic bit of dirt. Zimo have solder pads for this and instructions on how to add the cheap components, some other chips can be end-user "hacked" to add a capacitor (your warranty is void when you start soldering bits onto most chips!). Lenz have a fairly expensive solution in the UPS module which fits Gold chips. It contains far more capacitance for the volume than is possible with the Zimo method.

The Railcar has potential to be used as an automatic train. If the layout operator wants a rest, or there are questions from a visitor, a train which shuttles into the platform, waits, then departs is very useful. DMUs and Railcars are ideal for this service. The Lenz Gold and Silver chips support automatic shuttles using ABC braking modules. ABC braking is very simple - four or five diodes per brake section - so a shuttle could be created on a layout for the price of a couple of on/off switches and about 20p worth of diodes, and the shuttle won't break anything else on the layout.

So, for both UPS and Shuttle features, the Gold is the chip for this particular loco.

Having fitted the Gold, the changes are significant.

The motor in the loco does run more quietly, though the slowest speed isn't massively better than the TCS chip (and its nowhere near the slow speed which CT and Zimo can achieve). I have tried various of the Lenz motor options (though there isn't any documentation to explain what they do, so changes are blind), and have gone back to the defaults as the best option.

The UPS module is the biggest improvement. I had expected this from my experience of similar units in Zimo equipped locos. Its now pretty much impossible to stall this locomotive.

The automation is taking a while to setup; the basics are fine - ABC plus Shuttle behaviour. Distance related stopping does what it is supposed to do, though it doesn't appear to be as well implemented as that on Zimo chips.

Getting the distance related stopping to stop smoothly seems to need tweaking of the decelleration (CV4) and possibly a custom speed curve, otherwise the loco decellerates quite rapidly (following CV4?) and then inches along to the stopping point. The Zimo method where the loco maintains its running speed until it needs to decellerate to the stopping point seems so much better (or optionally the Zimo can work out a braking curve depending on the approach speed). There may be a few CV's I haven't understood, but, in part, that must be the weak Lenz documentation compared to that from Zimo.


So, overall, for this locomotive, the Lenz Gold + UPS does the trick. Compared to other methods of building an automated shuttle train, the Gold/Silver features are quite cheap. The UPS is moderately pricey, but its the only way for a Gold. But, if not requiring the shuttle feature and only needing ABC braking, then I think a Zimo is a better chip than the Lenz; the low speed control is better, the distance braking control is much easier to understand and setup, the UPS capacitor is £1 or so, rather than £36. etc..

Tuesday, 14 April 2009

Latest 4mm Developments and ScaleFour North 2009



At ScaleFour North in early April, I showed a few of my most recent developments.

To begin with, the "Coronation" has had its "up/down" coupler replaced with a rotating version. This has several new developments;
The rotation is in the opposite direction to previous designs. This works because the pivot point is below the deep buffer beam, so the coupling swings clear of the matching coupling on a wagon.
The pivot and flaps are from my test etch of parts; they have been modified further to fit the tight space, notably at the rear of the loco where brake rods interfere with the obvious positioning.
The coils are from Plantraco as discussed in earlier posts.

The chip in this loco is a CT DCX75, the coils are wired in series to the white function output. (The slow process of decyphering the behaviour of CT chips moves on; it seems that the white output is reliable for this operation, but the yellow still erratic. If the chip has firmware above v66, then both wires work correctly, though with different values for the CV's. The JMRI files are being updated again. ).




Secondly, I built a "demonstration frame" to show the etched bits and coil, these two photos might help illustrate where I am heading with etched components:

Tuesday, 24 March 2009

2mm 04 progress and more CT decoder issues

Firstly the 2mm scale 04, mentioned before in the blog. This is a Farish model, which has 2mm scale wheels fitted to it (modified 2mm Scale Association castings and I have a lathe!). It finally has the coupler fitted to the front and it works nicely.

I did encounter a couple of problems along the way, one magnetic and one software.

The magnetic problem is quite simple - the Farish motor is in the bonnet, so near the front of the loco there is some magnetic field from the motor. This proved enough to just cause the magnet in the coupler to "stick" in the open position. The eventual solution was a larger counterweight on the shaft. Video below shows operation, and the counterweight. The filling of the big gap in the bufferbeam is not yet complete.



The software problems were a bit more fraught, its yet another undocumented feature/bug in CT decoders. Without getting too far into CV numbers, it revolves around CV152 and CV153 in decoders with version 56 (DCX74) or 59 (DCX75) firmware (might be other numbers below 59 as well). In theory those CV's specify the output wire to activate during an uncoupling sequence. A value of "2" works correctly for the white wire. The documentation (German language versions) say its a "bitwise" allocation, so in theory "4" would work for the yellow wire.

Yellow does not respond to a value of "4". However, a value of "1" does work for Yellow, but with a minor glitch.... If the loco is set for Yellow at one end, and White at the other, then the sequence to operate White will also cause Yellow to momentarily activate whiles the loco changes direction. Its not a massive issue, but it might cause some systems to disconnect unintentionally whilst operating the uncoupler at the other end of the loco.

In the end, I rewired the loco so that both coils are in series, rather than independent function outputs. With the rotating action, the "wrong" end will uncouple and re-couple, so there is no disadvantage from operating both together, and arguably a little bit of wiring simplification.


Further bench testing has shown that a version 66 decoder (DCX74) has yet another change to CV152/153, which looks like a bug fix. The values are now "White=1" and "Yellow=2" (logical!) and the behaviour is correct at both ends. For testing, I use a home made decoder tester, which is a collection of LEDs and resistors connected to a small screw terminal block. The decoder motor and function leads can be connected to these, the decoder red/black go via crocodial clips to the command station/programmer. In my case a Sprog which is driven from my computer using JMRI. Using JMRI its quick to flip between different CV settings and observe the results.


So, to the recommendation for anyone else trying this stuff; if using a CT decoder, check how CV152/153 works in advance for your firmware. If in doubt, use the White output wire.

I'd like to see reports from others using the CT decoder and CV152/153; what happens with your decoder, what is its version number. It will help with development of the JMRI decoder definition files.

Tuesday, 3 March 2009

Plantraco Coils in L&Y 0-4-0 loco

The pictures in this post illustrate the use of the commercial Plantraco coils inside a scalefour L&Y electric loco (another High Level kit). I've used two slightly different arrangements, partly because of interior clearance issues, and partly to try out the options.

The first shows the coil mounted on a cylinder. This proved to be fairly troublesome to make, and I probably won't repeat the exercise.




The second shows a more normal "hinged flap" arrangement, similar to that on my 2mm scale Class 04. This was much simpler to build, and is the basis of a prototype etched component. The "nose" sticking out is a piece of lead which ensures the flap falls when the current is turned off, and the horizontal wire at the top limits rotation to 90 degrees.





The loco isn't finished yet; lots of details to add, and I want to make sprung buffers and other extras which require work on the lathe. So video in operation will be a while coming. But it does all work, the couplers are wired in series to a single function output of a Zimo MX620 chip. The chip is in a paper sleeve below the motor. Both couplers rotate 90 degrees, which is far more than I need to release an Alex Jackson. The Alex Jackson's will be fitted onto the outputs of the mechanisms with a bit of fine bore tube.


Friday, 30 January 2009

Zimo "stay alive" capacitors

Zimo offer the option of adding "stay alive" capacitors to their chips, which will keep the chip alive when the loco is not connected to the track. This can be a bit of dirt or similar getting in the way. Its a cheaper DIY equivalent to the Lenz "UPS" add-on for Lenz Gold decoders. Lenz use "superCaps" which have massive capacitance and can keep a loco running for several seconds, but superCaps are expensive and need extra circuits to work on 12v.

I have a Simplex in 7mm scale, from an old Impetus kit. Its rigid underframe, 4 wheels, very short wheelbase. As its rigid, there is a fair chance only three wheels are on the track at any moment in time, so its a bit sensitive to tiny bits of dirt.

As there is lots of space inside, I have tried a 4700uF capacitor connected to the Zimo MX63, with the additional diode, resistors and choke recommended by Zimo. This seems to keep the wheels running for about 1/4 of a second before the chip decides there is insufficient power for the motor. The chip stays alive for about 12 seconds before the capacitor has discharged through a resistor. This is a massive improvement in practical running - if a loco has bad pickup for 1/4 second its not going to run at all. It will sail over turnouts where it previously would stutter or stall. Obviously 1/4 second doesn't allow silly stunts, such as running along a workbench, but practical considerations suggest that is an easy way to break a loco !

4700uF is a big lump for anything other than 7mm scale. In smaller stuff, I've tried 100uF and 220uF capacitors and it does help even if its only keeping things alive for tiny amounts of time.

Saturday, 10 January 2009

Commercially wound coils for DCC controlled couplings

The biggest drawback with the DCC controlled couplings I've been experimenting with is the need to wind coils. Its time consuming and difficult work (I need a close up TV camera and it takes over an hour of continuous concentration to wind a single coil at the lathe). So, the hunt has been on for ready to use coils.

( Fanfare....)


Plantraco Microflight of Canada sell a "Nano Actuator Kit", which is a 3.8mm dia coil and a couple of tiny magnets. Not only does this save winding ones own, but it comes with a few neat side-effect benefits. When I enquired, it was 10 coils minimum for shipping to the UK, but I recommend checking before ordering.


The coil is 75 ohm and designed for 4.2v. So, two in series is 150ohm and nominally 8.4v. Bench tests suggest they are fine with two in series driven from the 12v of a standard DCC function output for durations of up to 5 seconds (I think they'd be fine for longer, just not tried). Alternatively, they also seem fine on half-wave (using a track pickup for the positive supply rather than blue wire). Compared to my home-wound coils which all needed current limiting resistors, this is a big benefit; no longer need to find space to hide a resistor.

If really worried a 56ohm or 68ohm resistor in series with a pair of coils would bring the current back down to within the maker's specification.


Having an internal diameter of 3mm, they will slide fit over brass tube of 3mm dia (the tube needed a tiny amount taking off with abrasive paper), this makes mounting easy, and also the construction of a hinge pivot quite simple, see the diagram below.


The brass tube is cross-drilled to take the shaft (0.35mm dia), and then the "ears" carefully filed to shape. The shaft is inserted, and fixed to the screwhead (solder/glue to choice), and then the magnet glued to the back of the screwhead (if the screw is steel, the magnet self-sticks!). Finally, the coil is slid into position - ideally a bit further towards the pivot that the diagram shows.
In deciding the shape of the "screwhead" part, give thought to whatever counterweight is necessary to return the coupling to its rest position - a piece of 2mm square bar with a slit might be a better shape in some cases, in others the counterweight needs to be away from the shaft movement.