Stanier 8F in S7

Discussion in 'S7 Group' started by DavidinAus, 8 February 2014.

  1. PaxtonP4

    PaxtonP4 Active Member

    Unfortunately the design of the compensation system is fundamentally flawed. The "fixed" axle needs to be able to rock or the compensated four axle unit needs to be able to rock about the centre line of the tender.
  2. mickoo

    mickoo Western Thunderer

    Agreed, the fixed front axle is the cause and needs to be able to rock up and down on each side with a centre pivot for a true three point system.

    There should be a rocking bar running front to rear down the middle of the tender over the front axle.

    At the moment the fixed front axle is fighting the fixed compensated unit rocking bar.

    The above might clarify a bit better.

  3. DavidinAus

    DavidinAus Western Thunderer

    OK, here are the pictures, and the request for advice as to how I try to extract myself from this difficulty.
    I described how I was arranging the split-axle pickups in the tender way back on 22/3/2015 (p8 of this thread), and there is a clearer picture of the underside of the tender chassis published on 3/11/2015, but this is the situation now:

    Small Tender dismantled 001.jpg

    This shows the fixed axle at the front of the chassis. I used Steph's "method 3", summarised thus: 1. make a hole in a square of double-sided printed circuit board ("PCB") and put a top-hat bush (THB") through this hole (hole just large enough to take the THB); 2. put the THB through the frame from the outside and solder the PCB in place*; 3. remove the THB leaving the PCB attached to the outside, exactly lined up with the hole in the frame. Use a countersink or a larger diameter drill to open out the hole in the frame to make the THB, when reinserted, clear (and therefore isolated) from the frame.
    * when soldering the double-sided printed circuit board to the outside of the frame I de-laminated the PCB on one side (the nearer side of the above pic./the right side of the tender when upright and facing forwards) so I used epoxy to glue the PCB together and fix its position. Maybe this is where a the misalignment arose, although I cannot see how. Anyway, why solder and not glue, when single-sided PCB would then be adequate, anyway?

    I then soldered pickup wire into the THBs, to allow pickup from the split axles with no need for wire rubbing on anything, etc.

    I hope that this can all be seen above, but here is another angle:

    Small Tender dismantled 002.jpg

    My dilemma now is how do I re-align the front axle?
    Just for a start, which side to I move, up or down. It is the left wheel which appears to be high, judged by the fact that on the rails it is this wheel that is not in contact with the rail head. On a completely flat surface all six wheels are in contact, but pressing down on the rear of the chassis lifts the left front wheel off the surface. The left wheel is going to be difficult to move, but judging by the fact that the left wheel is high on the track, lowering this it might be better. How do I tell, though?
    Even if I move one side, how do I tell by how much to move it, or how do I align it with the axle about which the compensation mechanism rotates - seen in the picture below? The compensation axle, seen just above one of the red wires, and protruding through the nearer side of the frame, is simply not visible from the from of the tender, to allow visual alignment.

    Small Tender dismantled 003.jpg

    How much of the brake detail, etc., will I need to dismantle?

    I'm thinking at the moment that my plan must be:
    1. I will need to take all the remaining brake detail off.
    2. measure the distance above a flat surface or rail-head that the left front wheel lies when I press down on the rear of the frames.
    3. measure very accurately the position of the axle relative to the top (or the bottom) of the frame.
    4. somehow unstick all the epoxy around the front left THB (very hot water?) and remove the THB.
    5. replace the THB on the PCB and use epoxy to stick it in place, lower in the frames by the amount measured at stage 2.
    6. put right all the damage which I will inevitably have done in the meantime.
    7. cross all my fingers, reassemble the chassis and try it out.

    I am not optimistic that this will be any of easy, straightforward or accurate.
    Does anyone out there have a better way?


    Last edited: 13 July 2016
  4. DavidinAus

    DavidinAus Western Thunderer

    I was writing the above entry at the same time as "mickoo"!
    Please don't tell me I need to completely redo the front axle to introduce a separate compensation! Actually that is of course exactly what I am being told, but at the moment I am still grasping at straws and hoping that I may get away with less.
    If you look at the pictures from my previous post, there is precious little metal in the frame about which to construct any sort of compensation mechanism, and in S7 there's very little width.
    I'm beginning to wish I had thrown away all the lovely MOK etches and designed a six-wheel independent sprung suspension unit. Pointless wishing: it is too late now.

  5. Dikitriki

    Dikitriki Flying Squad

    My solution;

    Leave the front axle.
    Sacrifice the rod holding in the compensation unit to get it out.
    Slightly enlarge (up/down only) the pivot points.
    Put in new rod.
    Hold down chassis on flat surface while soldering in new compensating rod both sides of the chassis.
    If body is nose up/nose down, use shim to level.

    Dog Star likes this.
  6. Overseer

    Overseer Western Thunderer

    Another variation would be to add a pivot to the centre of the transverse rod then shorten the rod so it doesn't go through the side frames. This would give three point compensation with the front axle staying fixed, although it is a complex way of achieving it. It could be done with the brake gear etc staying where it is.
  7. Steph Dale

    Steph Dale Western Thunderer

    I'm with Richard in that moving the fixed axle is likely to be a pain if it's insulated and has already been glued.

    Assuming the chassis isn't severely out of square I might be tempted to try something a bit lower tech and give the chassis a good twist. The final alignment can be checked by passing longish (9-12“) 3/16" diameter rods through the axle bearings and making sure they align with each other.

  8. DavidinAus

    DavidinAus Western Thunderer

    Thank you for all the above.
    I'm almost certainly going to try Richard's suggestion. The simplicity of just twisting the chassis to me is not attractive enough to overcome my dislike of deliberately changing the current dead-square frames (at least as well as I can measure it).
    If that doesn't work I might take heed of Paxton P4 and Overseer and see if I can construct a centre-line pivot for the compensation unit axle.
    I'll let you know what happens ....

  9. PaxtonP4

    PaxtonP4 Active Member

    As an alternative Ragstone Models do a sprung (CSB) chassis for a LMS tender: Tender kits

    You may well find that is the best and easier route - albeit some more expense.
  10. OzzyO

    OzzyO Western Thunderer


    before you start taking all sorts of bits off, try removing the wheels and axles and putting lengths of 3/16" rod in place then using some parallels to see how the axles line up. Use the fixed axle as your datum, the rear two axles should sit down. If not one or both of the compensation beams is not working right!
    This can be caused by a number of reasons,
    1] the slots in the frames not allowing the bearings to move freely up and down, if this is the case file the slots a bit.
    2] the pick up wires are stopping the compensation beams moving freely, use thinner gauge wire.
    3] the bearings could be too tight on the axles, run a 4.8mm reamer through the compensation bearings when one side is at it's top of travel and the other side at the bottom.
    4] how much play have you allowed on the centre and rear wheels (axles)? I'd allow about 0.005" if you don't give any then any the frames will almost act as rigid frames.
    5] check that non of the compensation beams are catching on any parts of the outer frames or body.

    I've built a good number of this type of tender frames, but I've always had the rear axle fixed.

    SteadyRed likes this.
  11. DavidinAus

    DavidinAus Western Thunderer

    If the inner chassis is just put down on the piece of polished granite that I use as a completely flat surface, all six wheels touch the surface.
    I can then wobble the chassis by pressing down on the rear end (either side), and the front left wheel comes off the surface while the front right wheel remains down on the granite.
    So if I buy and place 3/16" rod instead of the wheels I will just see the same thing, surely?

    I have measured the amount that the front left wheel is off the surface with feeler gauges and it is 0.4 mm. So if am going to change it, I suspect that I need to raise the end of the compensation mechanism bar by about that amount on the left.

  12. 3 LINK

    3 LINK Western Thunderer

    If you say it's 0.4mm, I would go with Steph's idea. I would still buy the 3/16" rods and use them front and back, as in hold the back rod down rigid, while slightly twisting the front rod. As the rods are threaded through the bushes you are only putting pressure on the parts which need adjusting.

    It isn't good engineering practice but it will save you loads of time, just a slight tweak is all you will need.

  13. Peter Insole

    Peter Insole Western Thunderer

    I know this probably seems a bit of an egg sucking point, and I know I'm probably the least qualified person to comment on this, but please let me describe a not dissimilar problem encountered with a friend's Aster Merchant Navy loco.

    It kept derailing on curves, once with a spectacularly expensive result! The problem was clearly the leading (compensated) bogie.

    Yet whenever it was examined, all moving parts were found to be perfectly free acting. There was more than enough free side to side, up and down and twist in the compensation beam!

    After yet another frustrating session, it stood disgraced in a siding. During lunch I couldn't resist having a look myself. I carefully lifted the front end of the loco so that the leading driving wheel would just lift off the track and bingo! The entire bogie came up with it!

    Problem solved. The centre bolster pin was too short. All the flippy-flappy stuff worked fine, but in the WRONG direction!

    I still suspect that this might be a similar case with your chassis. The compensation beam is clearly able to move freely, but appears to be at the limit of it's travel. It must surely be either the left middle and conversely therefore also the rear right that is prevented from moving further upwards that will lift the leading left wheel off.

    All the best, Pete.
  14. Peter Insole

    Peter Insole Western Thunderer

    Perhaps I should add that there might be some obstruction that prevents the middle right or rear left from travelling downwards that is restricting the upward swing of the beam? I still fear that tweaking the axles might only produce a very small adjustment at the corners, not nearly enough to overcome the problem that is so evident in your pictures (centre left wheel is way too low).

    I do hope you find the answer soon and that you won't be tempted to reach for the "Fine adjustment tool"!

  15. PaxtonP4

    PaxtonP4 Active Member

    Unfortunately no amount of tweaking will take away the fundamentally flawed design. You effectively have a ridged four wheel chassis.
  16. DavidinAus

    DavidinAus Western Thunderer

    I have to say, having thought about this quite a lot (as you might imagine), mickoo and PaxtonP4 are of course correct in that the compensation mechanism only really helps to reduce a rigid 6-wheel chassis to a rigid 4-point chassis. The front axle is fixed and the rear four wheels rock (the two sides independently) about a point on the two rails below the compensation bar/axle. To make this into a proper three-point compensation system would require a major rebuild, so my approach initially is going to be to try to make this an accurate fixed four-point system with the four points exactly level !
    I note OzzyO's and Pete's comments also, and will start the process by trying to make sure all the sliding and rotating surfaces are smooth and free.
    Even all that requires quite a lot of dismantling and rebuilding, so I hope it's enough. if not I will have to go to making a central pivot for the front axle, or possibly the compensation bar.
    The fact that it ran well when it was just the inner chassis alone is encouraging and makes me hopeful that the simple four-point chassis will work adequately.

    Last edited: 18 July 2016
  17. OzzyO

    OzzyO Western Thunderer

    Hello David,

    what happens if do the same at the front? You may get the same response using the 3/16" rod but you may not. If you apply pressure at the rear end I would have expected the opposite front wheel to have lifted! If you apply pressure to the middle at the rear both front wheels should lift, if not one of the beams is not working right.
    Are both of the compensation beams free to move as in the photo below?

    To check your inner frames all you need is a pair of parallels on your granite work table and place the inner frames on them upside down one at the front and one at the rear (so that they are at 90Deg. to the sides). Then see if the frames rock if the frames rock a twist will sort them out.

    But I'm still inclined to think that something is not working right with the beams.


    PS I was writhing this reply when David was writhing his above.
  18. Yorkshire Dave

    Yorkshire Dave Western Thunderer

    Having seen OzzyO's picture and as an 'outsider looking in'. I have not built anything like this but if the compensation beams work independently of each other would it be logical to slightly open/elongate the axle bushes vertically to allow the axles to twist slightly?

    Or am I just barking up the wrong tree :confused:?
  19. markjj

    markjj Western Thunderer

    Having earlier dug one of my F7 tender chassis out which works and compensates exactly as it's meant to do I fear by turning it to S7 you may have caused a dilemma for yourself. There have been several F7 builds but I have never seen or heard of anyone having problems with the compensation. Have you checked that the axle bearings actually float as they are intended they rely heavily on lot of float. You may have made things too tight by stretching the frame width....
  20. DavidinAus

    DavidinAus Western Thunderer

    Yorkshire Dave - the MOK design has "loose link" bearings which allow the axles to change angle relative to the frame. otherwise you would be correct and the compensation mechanism simply wouldn't work. One of my tasks mentioned in the my last posting is to make sure all of these are as free as possible.
    markjj - that is a possibility, not because the frame width is increased, which isn't needed to turn this kit into S7 as there is an frame outside of the wheels, but because of the printed circuit board stuck to the outside of the loose links in order to make up the split axle pickups, as described by Steph on his website and illustrated by me on p8 of this thread. I hope it isn't this, though, because there is no room between the wheel faces and the outside frame in S7 (just like the prototype, I suspect!).