Durley

A bit of lining added to finish off the livery. Blue is darker in reality than it appears in the photos. Getting close to completion.

Great to see Daisy at Southampton yesterday and great to chat to you Andy. As I mentioned when we spoke, you’ve inspired me to progress with my own dairy based layout in O gauge! Cheers, John

I’ve been making slow progress but have been remiss I’m not posting an update recently. I’ve been painting today and now have body and chassis mostly painted. Buffer beams, rods and crank still to be painted red and other details still to finish off. I’m going to do the wheels in body colour blue. I’ll be using the loco as a shunter for a small dairy layout. Express Dairies never had a Drewry to my knowledge (it’s probably a little big for that role in reality) so I’ve had to take some artistic licence with the colour scheme. I may yet add some lining in white. While the paints were out, I also painted the driver.

I do agree that the CSB approach appears complicated, mainly because it has generally been explained in terms of ‘beam theory’ using pseudo-scientific language. In reality, all the hard work has been done by others to develop the spreadsheet tools that are free to download on the CLAG site, making the process relatively straightforward for 4, 6 and 8 coupled locos, so covering the vast majority of prototypes. Firstly, only two things are needed to establish the positions of the fulcrum points (i.e. where to drill the holes in the chassis for the handrail knobs that the spring wire passes through), that is: 1. the lateral distance between wheel centres (wheelbase) and 2. the height of the beam above (or below) the wheel centres. The wheelbase can be entered in the relevant spreadsheet tool from the CLAG site. Selecting auto calculate will give offsets (i.e. the lateral distance from each wheel centre to its adjacent fulcrum points) which achieves (as close as possible) an even weight distribution across all coupled wheels. The spreadsheets do allow fulcrum positions to be manually altered if, for instance they are beyond the end of the chassis, clashing with a frame spacer or brake hanger mount, etc. This then recalculates the other fulcrum positions accordingly. It can also recalculate positions for an off centre weight distribution, but as Barclay points out, generally you won’t know the weight distribution until the model is complete. The vertical position of the fulcrum points is dictated by whatever method is used to connect the hornblocks to the spring wire, minus an allowance for deflection of the spring. In 4mm scale there are various proprietary solutions e.g. HighLevel, London Road Models, which use PE ‘tags’ soldered to the hornblocks. I have used the HighLevel ones myself on some 4mm scale builds, which provide for 3 different heights, which is useful so the beam height can be selected based on the geometry of your specific chassis. In 4mm scale, a beam deflection of about 0.5mm is commonly used. For 7mm scale, I have used 0.75mm. With fulcrum points determined and corresponding holes drilled in the chassis, it can then be assembled with hornblocks on all coupled axles, using jig axles and the coupling rods to align everything, as you would with a compensated chassis. It is important that the hornblocks use the same datum as was used to drill the fulcrum positions so that the correct offsets are achieved, HighLevel produces a jig to help with this which work with its hornblocks. Once the model is built (but no ballast weight added) it can be set on the tracks to see if it sits level and at the correct buffer height. Adding ballast weights at either end (e.g. in the smokebox or bunker of a tank engine) will normally be sufficient to get the model sitting level. Ride height can be adjusted either by adding (or removing) weight centred at the middle of the coupled wheelbase and/or by fitting thinner or thicker spring wires which will deflect less or more for a given weight. I just do this by trial and error. I have a range of steel guitar strings from gauge 8 to 24, with 10 to 14 gauge usually being used for 4mm scale locos. Is CSB worth the effort? I think so but as with everything it is down to personal preference. The main advantages are: 1. Smooth ride as the loco is properly suspended on the springs. It does even damp out some oscillations from any wobbly wheels 2. Compensation keeping the loco level over track irregularities 3. Coupled wheels maintain better contact with the track, good for electrical pickup 4. Even weight distribution so driven wheels will slip at the same time maximising traction (traction is always limited by the first wheel to slip which will be the least loaded one if weight distribution is uneven) 5. Hornblocks are retained by a single spring wire which is less fiddly than dealing with multiple hornblock retainers.

It’s starting to look like a Drewry shunter now. I’d been having problems getting the cab and bonnet to print properly on my 3D printer which I thought was my error but turned out to be a bug in the slicer software I was using that was randomly inverting some layers meaning it was trying to print the voids and not the model! Anyway, I updated the software and the problem went away. I’m happy with the cab but will be making some modifications to the bonnet as I’ve made the sides a bit too thin causing the print to distort slightly, which is why the bonnet sides are angled where they butt up to the cab. I also want to refine some of the details. I tend to find it takes me 2 or 3 test prints on larger items before I get a result I’m content with.

Scale crank now added from a 3D print. I managed to make it a sufficiently tight fit over the functional metal crank that it is push fit and retained by the crankpin bush. Allows me to easily disassemble for painting and any subsequent maintenance.

I’ve elongated the holes in the connecting rod to give circa +/- 0.3mm movement fore and aft, which appears to be sufficient to allow the suspension to work through the range of travel. I’m still to my draw up and make the 3D printed cover to sit on top of the jackshaft crank, that will include a representation of the balance weight. That will probably be todays task!

A bit more progress to share. I’ve been drawing (in Fusion360) and printing in resin (on my Creality Halot Mage) the details for the chassis including footsteps, air cylinder, brake shoes & hangers, sandboxes and the channel guard iron. I’ve milled the coupling hooks and printed the extended hook plate that was a feature of these locomotives. I found some whitemetal buffer housings in my bits box, left over from a kit of a BR Class 03, so have used these in place of 3D printing my own as I’m sure they’ll be more robust in metal rather than resin. Final clean up still to be done on the prints but I’m quite pleased with the results, especially the mesh guard for the rear steps which is just about fine enough to look convincing whilst retaining some strength.

Jackshaft drive assembled and torque reaction post from brass rod added to the gearbox which sits between two brass tubes, allowing the driven axle to move up and down on the suspension. Test run with sound enabled on the HM7000 DCC decoder using a LiPo battery as the power source. It’s the 08 sound profile so not particularly prototypical sounding but the closest currently available from Hornby.

Slow progress due to work commitments but I have managed to make the cranks and some spacers for the jackshaft drive. These are designed to fit on a Slaters square ended axle which will then set the quartering. I’ll 3D print a cover to represent the crank and counterweight of the prototype, which will fit over this functional crank.

Hi Ruston (it feels weird typing that as Ruston is my username on most forums!) Good question, I’d say it does improve the ride somewhat as the loco tends to glide over track imperfections with CSB suspension, rather than the sudden jerky movement when the fixed axle on a 0-4-0 encounters a bump on a three point suspended loco. However my main motivation with this build was to try out some ideas and push the capabilities of the CNC as a learning exercise for future builds.

CSB Suspension The principles of CSB suspension are meticulously documented here: http://www.clag.org.uk/beam-annex3.html. CSB offers a number of benefits, combing both springing and compensation in one, giving improved ride and wheel to rail contact when compared with a rigid, independently sprung or compensated chassis. My loco is battery powered, so doesn’t rely on wheel to track contact for electrical supply. However, CSB does ensure close to even weight distribution across the wheels which gives optimum traction, something that is important for a small loco intended to pull heavy loads. I have built few locos in 4mm scale with CSB suspension using Highlevel Kits hornblocks, CSB tags and drilling jig, which all work well. Unfortunately nothing similar is available for use in O gauge, so this was where I started the project by designing some parts in Fusion360. Holes for the CSB pivot points in the chassis need to be positioned with high accuracy. Small inaccuracies in the position of these mounts can have a significant impact on the functioning of the CSB system, leading to uneven weight distribution. That is why in 4mm scale I use the Highlevel Kits jig to position the holes. Using the CNC means I can have the machine drill these holes without the need for a jig, with precision of 0.1mm or better. I laid out my hole positions in Fusion360. Vertical positions need to be set based on the distance between the axle and the springy beam, allowing for deflection of the beam when loaded. I designed in 0.75mm deflection to hopefully achieve the correct ride height when the chassis is carrying the full weight of the loco. Horizontal positions are easy to determine for an 0-4-0, with one hole centralised in the middle of the wheelbase, and the outer holes equidistant from the centre holes. I had to move these slightly outboard from the optimum position (as detailed on the CLAG site) to avoid a clash with the brake hangers on the rear axle. To connect the hornblocks to the springy beam, I designed a simple part the could be milled on the CNC to fit over a standard Slaters brass hornblock. The beam passes through a hole in a small tab the is bent at 90 degrees, using a 1mm milled pocket as a fold line, in a similar way to a half etched fold on an etched kit. The hole in the tab is 6mm above the centre of the axle, the holes in the chassis to mount the beam pivots being 5.25mm above the centre of the axles, thereby allowing for the 0.75mm deflection. To avoid the need to mount separate hornblock guides, I decided to use slots in the frames to act as the guides, something only possible because of the accuracy afforded by the CNC machine. I did this by making an inner and outer frame, the inner frame has a narrow opening into which the hornblock is slotted, this being a loose fit. The hornblock is held accurately in position and prevented from rotating by the opening in the outer frame. After a bit of trial and error, I found an opening width that was 0.15mm wider than the Slaters hornblocks gave a nice sliding fit. The pivot mounting holes where drilled into both inner and outer frames, which then acted as a register to accurately align both parts together, using the mounts (re-purposed handrail knobs) as alignment pins. The spring beam itself is a metal guitar string. The exact gauge will be determined when the loco is close to complete and I know the all up weight. For now I have used a 22 gauge string. Hopefully these pictures explain what I’ve attempted to describe.

Hello Having mentioned this build on another thread, I’ve started this topic to document my build of a Drewry diesel shunter in O gauge, intended for use on a small shunting layout I am slowly developing. My model of the Drewry is based on the General Arrangement in the Drewry Car Company catalogue from the 1950s. I’m not modelling any particular prototype loco, but photos of the now preserved Elizabeth on the Alderney railway have been a useful reference. This class of loco is available as an O gauge kit from Connoisseur Models. Having built a few of Jim’s kits in the past, I’m sure this would make up into a great model but I fancied trying a few different ideas and, having recently acquired a 3D printer and CNC milling machine, I thought it would be an ideal candidate for a scratch build using these new toys. My intention is as follows: - Brass/nickel silver chassis cut on the CNC with Continuous Springy Beam (CSB) suspension, using cutouts in the frames as the horn block guides. - 40:1 Roxey Mouldings gearbox and Highlevel Kits 1626 coreless motor. Although intended for 4mm scale use, I’ve found these motors to be plenty powerful enough for use in small to medium size locos in O gauge. - Slaters wheels and crankpins. - Hornby HM7000 Bluetooth DCC control (made possible in O gauge by the smaller motor ensuring stall current is under the rating of the chip), powered by an onboard LiPo battery with battery management system (BMS) to prevent over discharge. - DCC sound using whatever is the closest profile from Hornby, currently a Class 08. - 3D printed cab, bonnet and details. - Removable bonnet to provide access to the battery for charging/swapping. All design work will be in Fusion360, using its inbuilt manufacturing features to produce STL files for the printer and NC files for the CNC machine. Some progress already is shown on this thread:

I’ve completed the basic chassis on my Drewry shunter project and given it a first test run. All parts cut on the CNC. Testament to the accuracy of the machine, it ran perfectly with no fettling needed. The loco is DCC Bluetooth controlled and powered by an onboard battery, but that’s a story for another thread!

A quick update with some parts I have made on the CNC machine. These are for an O gauge diesel shunter that I am scratch building, incorporating continuous springy beam (CSB) suspension. I’ve designed some CSB tags that I can use with Slaters hornblocks. I’ve then made an inner frame to retain the hornblocks and position the CSB mounts (repurposed handrail knobs). This will all be mounted on the inside of the outer mainframe (yet to be made). I intend to use slots in the outer mainframe as the guides for the hornblocks. I’ve also made the coupling rods. Brass parts are 28 thou (0.7mm) and the nickel silver rods are 64 thou (1.6mm). I’ve measured the finished parts to be spot on intended dimensions with my digital vernier callipers which measure to 0.01mm.

For jointing motion I’ve used these Prime Miniatures flat head brass rivets in 0.3mm and 0.5mm diameter which can either be used as rivets or just used as a pin and soldered to one rod. https://www.prime-miniatures.co.uk/content/03mm-dia-flat-head-brass-rivets-50

Hi All I’ve recently acquired the new CNC milling machine and thought I’d share some thoughts. I’ll start with a disclaimer, the machine was bought paid for by me. I have no relationship with the manufacturer other than being a satisfied customer! I had been looking for a CNC machine to allow me to accurately make parts in brass or nickel silver for various railway modelling projects. Mainly I was looking at 2D profile milling in sheet material for things like locomotive frames, valvegear parts, W-irons, etc in various sheet thicknesses from 0.3mm to 2 or 3mm or so. Until recently, the choice of CNC machines capable of cutting metal has been limited to professional steel framed machines priced in £thousands. Cheaper aluminium extrusion based machines, typically diy home builds or from Chinese manufacturers, have been available but weren’t really rigid enough to cope with machining metal. These machines have been improving however and the recently upgraded Genmitsu 3020 Pro Max V2 CNC promised to fit the bill, being marketed specifically for machining in metal and priced at less than £500, it looks good value. With a bed size of 30cm x 20cm, it’s large enough to cover the sort of jobs I wanted to use it for, but still small enough to be used on a desktop. Machine Setup The machine comes mostly built, but requires some final assembly to attach the z-axis to the bed and install the spindle motor. This was achieved in around 30mins following the excellent instruction manual. The machine comes with an offline controller, so machining projects can be undertaken without the need to connect a computer (although this is an option and a usb cable is also included). Projects are loaded to the offline controller via a (included) SD card. The card comes preloaded with a few example engraving projects for use with the supplied vee bits, so I was able to try it out the machine on a wood off cut, to get used to its operation. For precision cutting in metal, a level bed is important so I acquired an MDF spoil board and surfacing bit, to ensure the x-y bed is perfectly perpendicular to the z axis. G code files These machines use GRBL firmware, requiring the project to be machined to be produced in G code (.nc) format. There are many software products available to produce the required G code files. I used Fusion360 as this allows both CAD (design) and CNC (manufacture) to be undertaken in the same software. I was already familiar with Fusion360 as that is what I use to design models to 3D print. There a loads of video tutorials for Fusion360, making it accessible, and it is free for hobby use. Metal cutting end mills There is a bewildering range of milling bits available so finding an appropriate milling bit compatible with both the milling machine and material to be cut can be a daunting challenge. Having done some research online, single flute end mills looked like the way to go with this kind of machine for cutting brass. The single flute allows cutting at a sensible feed rate with the relatively low spindle speed of 10,000rpm that the machine provides. I bought 1, 2 and 3mm end mills to give me a few options. The machine comes with a 1/8 inch collet so the bits need a 1/8 shaft. Buying other diameter ER11 collets allows for a wider range of bits to be used. Feeds and Speeds Milling requires parameters to be specified covering most importantly the spindle speed, feed rate and depth of cut. There are calculations that can be done to determine the range of these variables for a given end mill and material combination. To be honest, I just copied the parameters from this video on YouTube and have adjusted from there. These settings are very conservative so make a good starting point. Work holding There are various ways the material to be machined can be held on the bed. The machine is supplied with 4 screw clamps which are fine. I however found by far the easiest option for sheet materials was to use the tape and glue method. This involves applying masking tape on both the back face of the material and the surface of the bed. The two can the be glued together using CA glue, ensuring the material is firmly held on the bed but allowing the finished parts to be removed cleanly by just peeling them away from the masking tape. Machining in Brass So how well does it machine brass? With the correct settings, amazingly well I think! Finished parts are cleanly machined and I am getting within around 0.01 to 0.02 mm tolerance when measured with digital callipers, which is plenty accurate for my applications. The specs for the machine state 0.1mm tolerance which I suspect maybe closer to what is achieved on larger parts using the whole bed. Here are a couple of videos showing my machine in action. The first is 1mm brass sheet machined with a 2mm bit and the second is the same brass sheet machined with a 1mm bit. Feeds and speeds I used are included in the video descriptions. Pictures of the finished components are also attached. I’ve successfully cut thinner sheet (0.3mm, 0.6mm) and up to 1.5mm which is the thickest I currently have in stock. In summary, I’ve been really impressed with the machine, its capabilities and ease of use. I now have a stack of project ideas I want to use it for!

Hello Does anyone have a copy of the instructions for the Backwoods Miniatures L&B Lyn kit in 009? I recently acquired the kit second hand but alas no instructions were included in the box. Reproduction and postage costs happily paid, or beer money in exchange for an electronic copy!

Agreed, this 7mm axle box has the LSWR lettering at 0.35mm tall and the writing is perfectly legible. The individual parts of the letters are around 0.1mm wide.

The new Creality Halot Mage would fit a 4mm coach at an appropriate angle. They look like a bit of a bargain compared with other similar size machines. https://store.creality.com/uk/products/halot-mage-8k-resin-3d-printer?utm_source=Creality&utm_campaign=7d1fc81503-EMAIL_CAMPAIGN_2023_03_06_08_26_COPY_01&utm_medium=email&utm_term=0_-24b15189fe-[LIST_EMAIL_ID]&mc_cid=7d1fc81503&mc_eid=5f07354477

@drduncan is the distortion you are experiencing on a wagon that includes the floor as an integral part of the print? If so I would suggest trying printing without the floor which may be part of the problem by generating an enclosed volume during the printing process. I have a similar but not so extreme issue on my 3 plank wagon. The lower wagon is printed with the floor included and shows slight distortion to the headstock and in the line of the side planking. The upper print is the same except with the floor removed, it does not have any distortion evident. In my case I decided I could live with the distortion which is not that noticeable to the naked eye.

The calculation for optimum angle is: Arctan (layer height / pixel width) Note this is the angle that produces the smoothest finish on an angled surface by minimising the step size in the print. It doesn’t address distortion, tearing or other issues, it is just about optimising print fidelity.

Wagon as supported.

I have found angling wagons at 45 degrees gives the most consistent results. 45 degrees is commonly considered the maximum overhang for each layer that is self supporting, and both the sides and the ends print with the same amount of overhang, minimising drooping or tearing between support attachment points. On my Creality Halot One, the screen pixel width is 0.05mm and by matching the layer thickness to this, you end up with the horizontal and vertical surfaces having the minimum step size which minimise visible layers in the print. This video is a good explanation of angle and print quality https://youtu.be/Qs2Rb0ExnIM I have had problems with getting slicer generated supports in the optimum place so have drawn up my own angled support ‘wedge’ that I combine with the wagon in CAD. I then use breakaway supports to attach the wagon to the wedge, similar to the supports Stevel has shown, either along the top edge of an open wagon or bottom of the solebar on a van. I then add some light slicer generated supports to any details on the overhanging end of the wagon to avoid any protruding details from failing to print due to islanding. The attached photo shows a print of 3 plank wagon that I have scaled down to 2mm/ft scale. The lattice structure is the wedge that I printed with the wagon and have since detached. Note this model is drawn to exact scale so has very thin sides in 2mm/ft scale, circa 0.35mm. O gauge version is in the background for reference which was printed exactly the same way.