Ian_C's workbench - P4 and S7 allsorts

Discussion in 'Workbenches, including workshop techniques.' started by Ian_C, 21 May 2017.

  1. oldravendale

    oldravendale Western Thunderer

    Wait for it......wait for it.....

    The F7 28XX will be coming, all in due course.

  2. SimonT

    SimonT Western Thunderer

    Nearly finished the tender instructions! Loco comes next.
  3. Threadmark: Smokebox liner rivets

    Ian_C Western Thunderer

    48142 had a liner plate fitted to the inside of the lower part of the front ring pressing and the rivet heads are visible on the outside of the smokebox. The front ring and the smokebox door are rendered in the MOK kit as a single brass casting. Quite a nice casting too, but it has no liner rivets. The positions of the rivets can be worked out/estimatedfrom the drawings on page 46 of the Wild Swan book. The chances of me marking the rivet pattern on the casting accurately by hand are about nuffink, so I sketched it all out on CAD and plotted the rivet positions relative to the centre of the door.

    0518 smoke box door rivets.jpg
    The rivets appear to be 1/2" diameter on the drawing. Googling rivet suppliers suggests that 1/2" rivets usually have a head diameter of 7/8", which looked about right on CAD. They're quite subtle on the prototype. You'd miss them if they weren't there but they don't stand out and I wanted them to be similar to the rivets already cast into the part. 0.5mm wire scales out about right for a 7/8" rivet head. The casting was set up on the milling machine with the door straps aligned with the x-axis and the centre found and set to x=0, y=0. The hole positions were cranked up on the DRO and drilled with the 0.5mm tip of a small centre drill. Which was OK, except that in my eagerness to get on with it I had the casting 180 degrees around and started drilling then pattern in the top half of the ring. Homer Simpson moment. Set it up the right way round and put all the holes in the right place. Short lengths of 0.5mm brass wire were soldered in the holes and then filed down to represent the rivet heads. The wrong holes were filled with solder and cleaned up (except one escaped - you can see it in the photo - I'll get it next time the soldering iron is hot) Residual solder was removed with a glass fibre brush.
    smokebox ring rivets.jpg
    Worked out OK. The outer rivets ended up a bit close the the radiused edge. I'd move them in a touch if I did it again. They'll hardly be visible amongst the ash, corrosion and flaky paint, but they're there and I can cross that job off the list.
  4. Threadmark: A distraction - lathe and DRO

    Ian_C Western Thunderer

    Sort of a second level subroutine - using machine tools to make parts for another machine tool to (eventually) make parts for a model locomotive. But it's all designing and making things so thoroughly enjoyable. New lathe arrived and I'd set my mind on fitting DRO to it. If you're a machine tool traditionalist this lathe, a Proxxon PD400, is heretical stuff. If has a die cast aluminium headstock for a start (hrrumph....splutter...etc). For the work I intend to do it suits just fine, and unlike most of the 'hobby engineering' lathes these days it's manufactured in Europe and not China and it is very nicely made and put together. It's a relatively small lathe though, so fitting DRO on the bed, and particularly on the cross slide, was a design challenge. For those wot's interested I used compact 5 micron magnetic scales and reader heads from Machine DRO (Machine DRO | Digital Readouts | Engineering | Machinery). Hope you can make some sense of the scrap book below.
    PD 400 compilation.jpg
    All set up on the bench now. Had a test run and it all works very well. Sooner or later the old lathe and the mill/drill head will be up for sale. Anybody wanting a first lathe?
  5. Threadmark: The missing drop link

    Ian_C Western Thunderer

    Not done much for a few weeks so it's therapeutic to get back to some modelling over the Xmas break. Thought I'd tackle one of the 'bogey jobs' . Not a set of weight relieving and guidance wheels, but one of those difficult jobs that you put off doing until inspiration strikes or you have no choice. This is the making of a crosshead drop link, a brass cast part that was missing from the kit. Sure I didn't lose it and it looks very much as if it didn't cast on the sprue, small feeder and possibly blocked by debris during casting. Either way, with no spares forthcoming making one from scratch was necessary.

    For a long time I couldn't work out how I'd make one. It's quite a complicated shape, requires a degree of precision and I couldn't see myself filing it out of a solid lump. The eventual method and notes below...
    drop link storyboard 2.jpg

    1 & 2 - I find it helps to model problem parts on CAD. Gives you a good understanding of the part and the key dimensions and often suggests possible methods of manufacture. There's enough information in the Wild Swan book on the motion drawing on page 54 to construct the part. When you break it down to basic geometry it's not as bad as it first looks.

    3 - It looked like a possible way of making the part was to mill a brass section corresponding to the end view cross section. A length of 1/2" brass bar was big enough to make the section. Flat milled on top of bar in the vice, turned over and dropped onto parallels and flat milled parallel on opposite side. The reason for using round bar and not making the initial flats too big is that the bar can be rotated in the vice to mill the angled faces of the section. It's a whole bunch easier than angling the milling head over and re-setting it afterwards.
    milling sketch.jpg
    The sketch shows how the basic section is cut from the round bar. Most of the bar is held in the vice, either on parallels between jaws or in the locating groove in the sliding jaw. The milling is done on the overhanging section. Mill the top flat (section 1) and mill the vertical datum face (on the left). Mill out sections 2 and 3. Turn it over and mill off section 4 to give the required overall thickness of the part. Mill section 5. Then hold the bar in the jaws using the remaining cylindrical faces and rotate it to the angle required to mill the angled sections 6 and 7. Easy enough except for the small angled section 7. That's just over 1mm in width and sits between two other faces. I had a go with a 1mm milling cutter, and in spite of running at maximum spindle speed and feeding ever so slowly I managed to break it before the job was done, so that face was finished with a needle file after it was removed from the mill. Finish the milling by taking the other vertical face back to give the correct overall height to the section.

    4 - There's the basic section before cutting the angled faces. Hope it all makes sense.

    5 - With the work back horizontal the end is milled square and used as a datum for drilling holes. The three bolt holes for fixing to the cross head are 0.5 mm diameter and the union link pin hole is 0.8mm diameter. I made the section long enough to drill two hole patterns on the basis that I'd either mess one up, or lose one, or if not, then I could choose the best of the two.

    6 - Turned to vertical to drill the holes for the oil reservoir bungs.

    7 - The milled and drilled blanks are removed from the bar and cleaned up. At this point you get the feeling it might just work.

    8 - Blanks again, showing the cross section.

    9 - Marking out, piercing saw and about an hour of patient filing to make the finished part. Seen from the front - the kit casting is on the left, finished part in the middle, marked blank on the right. First time lucky and didn't need the spare blank!

    10 - Seen from the rear.

    That's the drop link job crossed off the difficult list.
    Anyway - I hope all you WT types had a good xmas and I wish you a peaceful and productive 2018.
    P A D, ScottW, Dog Star and 9 others like this.
  6. Brian McKenzie

    Brian McKenzie Western Thunderer

    Good going Ian, that's a neat trick - using round material to acquire the angled faces easily - without disturbing the *tramming* of the mill. ;)

    Breaking a cutter is a :'( moment. Do you recall if you were climb-milling? Although a 1mm dia cutter is never going to pull the work by drawing the machine table into itself, a light tightening of the table V ways (if available) can help smoothing the feed when using delicate cutters. A more likely reason contributing to breakage is run-out of the cutter, however minute. Tiny cutters are very susceptible to this, and if using ER collets, it's worth acquiring proper 1/8" bore collets (if using 1/8" dia shank cutters) rather than tightening down a metric collet that caters for shanks from 4 to 3mm.

    Best wishes to you and yours for 2018.
    Brian McK.

    *possibly an American term - used to describe the process of squaring up the milling spindle to the table surface.
  7. David Taylor

    David Taylor Western Thunderer

    Really good stuff! I find the worst thing about the tiny cutters is they're more expensive than the bigger ones :(
  8. JimG

    JimG Western Thunderer

    Very nice work. I sympathise with your small cutter short life span. :) I use 0.5mm, 1mm and 2mm carbide slotting cutters for most of my detail milling work and the attrition rate at the start was quite high and very expensive. :) However, I now have the luxury of occasionally being able to pension off a cutter for being blunt. :)

  9. Ian_C

    Ian_C Western Thunderer

    Thanks for the advice Brian. I think the problem was a combination of run out, too deep a cut, slower than ideal spindle speed, impatience and an ambitious (manual) feed rate. Wasn't climb milling, was cutting straight into a slot.
  10. Threadmark: "Obtain a running chassis" - part 1

    Ian_C Western Thunderer

    It’s been mentioned before on WT that sometime around now the MOK instructions simply say “…obtain a running chassis…”. I suppose if you take on a kit like this it’s not unreasonable to assume that you have the gumption to turn the thing into a working chassis. Still, it’s epically terse, and I’m sure there will be plenty of challenges to overcome before I obtain a running chassis.

    I got this far before with the original Slaters wheels before deciding that they required improvement. As previously posted the Slaters wheels were converted to take home made telescopic axles. That worked out OK and the ‘Slaters shimmy’ has gone away. It’s worth noting that on the rear axle that carries the motor and gearbox the taper pin that holds the axles together has to be offset from the centre. I knew this but I hadn’t fully worked out the detail when I designed and made the axles, but as luck (I know, planning is better than luck) would have it there is just room to pin axle number 4 to one side of the gearbox. Just. One of the gearbox axle bearings had to be shortened slightly to make enough space. It’s a bit of a compromise but not tragic. The motor and gears are from ABC Gears - MINI-7E with a Maxon 352988 motor with a 30.2:1 reduction. Works well and fits in the chassis and firebox without any modification. I've had a go at some video and I'll try and work out how to post/link to it for those wot's interested.

    normal split axle.jpg
    split axle and gearbox.jpg

    There’s another problem that becomes apparent on closer inspection. The compensation doesn’t really work. The eight wheels are longitudinally compensated as four groups of two. All four compensation pivots are fixed to a chassis side plate. This produces the kinematic equivalent of a rigid four wheeled chassis. On an uneven surface one pair of compensated wheels won’t be carrying any load. That’s not quite true since the bearings on axle ends of the unloaded wheels are able to drop in the horn guides until the wheels touch the track. They only carry their own self weight though, not a good recipe for track holding.

    I had originally thought about building the chassis with CSB (continuous spring beam) suspension. Not wanting to complicate matters and wanting to press on I elected to go with the chassis as designed. Forum comments from other builders suggested that there was enough free play in the compensation to make it workable in practice. Now I have the chassis on wheels finally I can see that doesn’t work well. Adding cross chassis compensation to one pair of beams would work but would require major chassis surgery and would add parts in visible areas of the chassis. CSB likewise would require a lot of chassis re-work. The best solution I could come up with was to add spring assistance to one pair of beams. In normal running the loco sits on the compensation beams and they control loco attitude and correct ride height. When one beam becomes unloaded the springs press the unloaded axleboxes down so that the wheels contact the track and take a proportion of the nominal axle weight.

    Unsoldering the compensation beam pivots and extracting the beams is a faff and requires cutting and bending away part of an axle box guide. The beams are unsoldered and the laminated etches are separated. The inner etched beam is modified to allow a phosphor bronze wire spring to be fitted alongside the unmodified beam. The spring ends bear on the tops of the axle boxes. The compensation beams are now single thickness but still plenty strong enough to support the loco.

    There’s the question how springy the spring needs to be since I don’t yet know what the finished weight of the loco will be. I piled all the current components onto the kitchen scales and estimated how much else was to be added and came to the conclusion that the loco would weight between 1.0 kg and 1.2 kg when complete. I have spring calculations in a spread sheet, I know the length of the spring, I know the weight each spring should support and I have a choice of 0.5mm or 0.9mm phosphor bronze wire. End result is that 0.9mm PB wire set to about 1.5mm below the compensation beam level will deflect to support about 75% of the axle load when the loco is on level track with the beam ends in contact with the top of the axleboxes. It's approximate but it seems to work OK on the test track.

    Pictures being worth a thousand words and all...
    comp beam mod 1.jpg
    comp beam and spring 1.jpg
    comp beam installed 1.jpg

    ... and I will drop the unhygienic chassis back in the ultrasonic tank at some point to remove all the filing and fettling debris.
  11. Threadmark: Rolling chassis - first run video

    Ian_C Western Thunderer

    Bit new to this video thing, but here goes...

    Before anybody comments the sound isn't DCC, it's some steam loco sound effect pasted into the video.
    fenman, adrian, richard carr and 6 others like this.
  12. richard carr

    richard carr Western Thunderer


    One thing I often do with the taper pin on a driven axle, is to put that through the final drive gear instead of using a grub screw, then it just goes in the middle like all the others.

    Ian@StEnochs likes this.
  13. Threadmark: "Obtain a running chassis" - part 2

    Ian_C Western Thunderer

    Fitting the coupling rods for the first time was something that I was both looking forward to and not looking forward to. If successful then it's a big step towards a running chassis. If not, then something's gone wrong and it's backtracking and problem fixing. In my 4mm experience it's usually the start of a struggle with quartering and rod centres. Surprisingly the rods dropped on and it all worked first time. I shouldn't be surprised really, a lot of careful work went into chassis, rod and wheel sets, so why wouldn't it work?
    rods fitted 1.jpg
    The chassis trundled up and down a gently inclined test track with only an intermittent stiff spot. Some close examination showed the stiff spot to be caused by the inside end of the crankpins occasionally contacting the heads of the 16BA screws holding the axle box keepers to the chassis with the wheels at one extreme of side play.
    Thinking back, the crank bosses were reduced in thickness (see an earlier post), which left the original S7 Group crankpins threads slightly longer than the thickness of the boss. The crankpins were taken out and had a couple of threads filed off the inside ends before reinstallation. That cured the problem but the screw heads were still closer than you'd like. The eventual solution to this problem wasn't what I had imagined and will be covered in the next episode.
  14. Threadmark: "Obtain a running chassis" - part 3 - springtime

    Ian_C Western Thunderer

    Thinking some more about the axle box keeps. The instructions suggest soldering small nuts inside the chassis and fixing the keeps with small screws. There are half etched recesses in the chassis side plates for the nuts. They are very small and the only nuts that fit are teeny weeny 16BA nuts (Eileen's Emporium - there are none in the kit). This is how it works out...
    keeps plan A.jpg
    ...and as I found in the previous episode the screw heads are a nuisance in the limited space between wheel and chassis. I figured a better solution was to put the keeps on the inside of the chassis plate, solder 14 BA nuts to the the keeps and fit them with 14BA screws. Voila...
    keeps plan B.jpg
    Although the 14BA screw heads are fatter than 16BA screw heads moving the keep inboard means that the screw heads don't project as far. And a bonus is that 14BA is slightly less of a fiddle than 16BA.

    Well, that would have been that, except I remembered about the cosmetic spring castings and wondered how they would fit around all this. Turns out that with the spring castings fitted to the chassis you cannot remove the axle boxes. I couldn't see a practical way of making the spring castings removable. After sleeping on it I decided that the axle boxes probably didn't need to come out again (painting? I may regret this, but press on) and I may as well just tack solder the keeps in place and dispense with the nuts and screws.

    The spring castings are of brass and quite nicely done. They do need to be modified to get the spring hangers to clear the compensation beams. This is shown in the instructions. You need to remove fully half of the hanger bracket and put a decent chamfer on the end to clear the compensation beam in it's extreme position. Make sure you do 4 of each hand!
    springs prepared.jpg
    Another challenge becomes apparent now. The spring hangers are to be soldered to the inside of the chassis plate. The castings have small round spigots on the hangers, presumably to locate them on the chassis. Firstly, we've just filed most of the spigot off one end, and secondly there aren't any holes etched in the chassis anyway. I'm guessing they were just missed off the etch. Locating and soldering the springs 'freestyle' inside the chassis would be an absolute a**e of a job. So we must have at least one of the locations holes in the chassis for each spring. Measure, sketch, arithmetic and marking out to locate the holes.
    spring hole.jpg

    I drilled the holes a little small and opened them up to a tight fit on the casting spigot. They pretty much stay in the right place as you tack one end, adjust and tack the other then solder both ends. There's a little spigot and solder to tidy up on the outside of the chassis. Easy enough in the end.
    spring fitted.jpg Eventually they are all fitted and the chassis starts to look a bit more like the real thing. We have snowdrops, no daffodils yet, but it's definitely springtime here.
    all springs fitted.jpg
    Next task is to refit the wheels, rods and the motor and gearbox and see how it runs under power driving from the rear axle.

    Attached Files:

  15. Scale7JB

    Scale7JB Western Thunderer

    If I ever need to put a bolt through the frames to hold something internally (mostly to hold leaf Spring castings) I usually tap the frame material, screw in the bolt and solder in place.

    Then the bolt head can be filed away on the outside so it becomes flush and invisible.

  16. Threadmark: "Obtain a running chassis" - part 1b - side play

    Ian_C Western Thunderer

    Kinda forgot to mention this - axle side play and negotiating curves. Back tracking a little then.

    Plotting a 6ft radius curve on CAD and superimposing the axle centres suggested that if the side play on the front and rear axles was zero then axles 2 and 3 would need about 0.9mm side play either side to negotiate the curve. Side play on the front axle needs to be as little as possible to maintain clearance between the back of the cross head and slide bars and the coupling rod and leading crankpin. Side play on the rear axle should be minimal in order to keep the loco aligned with the (straight) track. Axles 2 and 3 can float laterally and find their own positions on straight or curved track.

    0.5 mm spacer washers were made to fit between the inside of the front wheels and the face of the axle boxes. 0.6 mm spacers were required on the rear axle. You’d think that front and rear spacers would be the same wouldn’t you? Maybe there are some small errors or accumulated tolerances in the chassis. I’m not going to worry about it. With that done, and with the help of a straight edge, I found that axles 2 and 3 had about 1.0 mm side play each way.

    One thing to note; there’s quite bit of side play between the axle boxes and the axle box guides in the chassis. That means you can’t completely eliminate side play on axles 1 and 4. Hopefully axle 1 is well enough controlled to clear the crosshead- we’ll see shortly. Altogether that arrangement should work on a 6 ft radius curve with a bit to spare.
    Number Cruncher likes this.
  17. Ian_C

    Ian_C Western Thunderer

    Hadn't thought of that. Thanks for the tip. Next time....
  18. Threadmark: "Obtain a running chassis" - part 4 - coupling rods and under power

    Ian_C Western Thunderer

    The chassis was stripped and all the parts cleaned in the ultrasonic tank to remove flux, filings and general bench debris. Back together and lightly oiled. Ran just as smoothly with the motor driving as it did freewheeling. That's a relief.

    I'd considered a number of methods for motor mounting. However you do it you want to constraint the motor from fore/aft movement but allow it to move from side to side to accommodate the rear axle sway on uneven track. There are three motor restraint brackets in the kit, one of which fitted the Maxon motor nicely. I arranged the bracket to engage the plastic end cap of the motor to reduce noise from vibration of the motor against the mount. As I had offset the motor and gearbox slightly to make room for the axle taper pin I had to elongate the bracket mounting holes a little to centre the bracket on the motor. There's also a small relief cut out to keep the wires clear of etch edges. Also worth noting that the motor restraint bracket on the chassis has been cut away to provide clearance to footplate and firebox casting but hasn't had any reinforcement added. There has been a bit of debate on one of the MOK 8F threads about whether it needs reinforcing when the side ribs are cut off like this. It doesn't. The unreinforced etch is plenty stiff enough to resist the motor reaction torque. motor restraint 1.jpg

    Here's some video of the chassis with coupling rods fitted running under power.

    It is very smooth and very quiet, which is a good thing when I come to add the DCC sound. Can't decide whether there's any benefit to be had from adding a flywheel, there's a bit of room in the firebox for one. The big shiny lump on top of the chassis is a tiny toolmaker's vice just to add enough weight to compress the springs added to the front compensation beams. There are no pick ups as the plan is to pick up from the tender (which means that I can happily kick that challenge down the road for a bit longer). The motor leads are croc clipped to wires from the controller. You can see them dragging along behind and trying to snag every obstacle on the bench. . The pony truck is just going along for the ride held down by gravity; maybe there'll be a spring when I get round to thinking about it. I'll add that the tractive effort is astonishing, an order of magnitude more than my 4mm expectations.

    Sorry about the glaring, flickering background and some ghosts where the clips fade in and out. Still trying to get my head around iMovie and this video thing. It's not as easy as I thought!
    Last edited: 12 February 2018
  19. Scale7JB

    Scale7JB Western Thunderer


    I don't think an 8 coupled chassis gets any smoother than that, i'd be over the moon to make something as well running as that with so many wheels.

    Nice little concept to the video too...


    Len Cattley, 7mmMick and adrian like this.
  20. Lancastrian

    Lancastrian Western Thunderer


    I made two coil springs for the pony truck on my Finescale MOK 8F. I'll be doing the same on the current S7 build.