Crab Comet conversion

Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Mon Apr 02, 2018 9:15 pm


Getting the right height relationship between the two was properly reliably achieved only by means of using the EMGS coupling assembly.
Correct height relationship

Edit. I have realised that this is still incorrect. The bottom steps should line up, not the higher ones.
Shows correct relationship of tender/loco steps
lnwrrm979.jpg (68.63 KiB) Viewed 799 times

On the loco it was screwed in with paper washers to provide bite and fixed with Loctite 243. The rear is a 7mm scale handrail knob upon which rests the suitably modified tender front coupling.
Loco coupling

Tender 'coupling'


The self tapping screw gave up the ghost and some plastic was glued behind the buffer beam to retain the chassis front - much easier for disassembly.
Front fixing
Last edited by Julian Roberts on Fri Apr 20, 2018 8:09 am, edited 1 time in total.

User avatar
Allan Goodwillie
Posts: 667
Joined: Sun Jul 20, 2008 7:00 pm

Re: Crab Comet conversion

Postby Allan Goodwillie » Sat Apr 07, 2018 8:50 am

Hi Julian, :)

I have been following your crab build with great interest as it will be good to see a crab performing again on Calderside after many years. Alan Clark still has the Japanese one he built for the layout in the 1970's. We had it running at Glasgow a few years ago when we had Calderside there in its original form. I am trying to sort out my garage space to get Grayrigg up and running again, it having been neglected while I was busy on the Wemyss layout. We can do a proper haulage test when the time comes and see how it copes with the 1:70 grade. Crabs did have very big and prominent cylinders, clearance for the abandoned platform might prove a challenge - my WD's occasionally give a rub! Always fancied one of these and the Carlisle ones came up into Scotland on a regular basis. There were also the Ayrshire based ones which pretty much stuck to their own area. Pity it will not be ready for Scalefour North as it will look and clearly run well on Kettlewell. An excellent addition to the fleet!

Allan :)

Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Sat Apr 07, 2018 3:07 pm

:D Cheers Allan. Alan Clark's Crab is Japanese?! What make is that? I think you had it at Scaleforum last year (or was it the previous year?) as well? This will be one of the Ayrshire locos for Calderside but will hopefully be allowable on John's Kettlewell ignoring the number and allocation, when there is a fully exMidland loco roster for running in the BR period.

Here are some further detail shots to illustrate the previous posts. This one shows all three sandboxes in better light if you zoom up. There is plenty of cleaning up still to do of course.

Flashlit side view

This prototype shot shows what happens to the reverser. It disappears behind the sandbox before emerging above the running board ( /footplate?). So there's not much point in fixing on most of part 36 - even in the unlikely event anyone was to notice such detail.
Reverser goes behind sandbox

A close up of how the model is at present there, and showing the enlarged slot in the frames which when painted will I think be more or less invisible.

Showing yet to be completed reverser detail.

Here is a prototype shot showing the centre wheel sandpipes in more detail. I expect there were variants. Not all prototype pictures show double pipes, so I don't know if that means they weren't there or just mounted more exactly together. Would there have been sanding without steam jets delivering it? - I am conjecturing one of each double pipe is for the steam delivery.
ELR Crab sanding mechanism detail

Anot her Crab with single pipes...or is it that the photo is not sufficiently sharp?
lnwrrm979.jpg (68.63 KiB) Viewed 941 times

And another prototype shot
lnwrns1654.jpg (69.6 KiB) Viewed 941 times

Still to be done after finishing the sandpipes is fixing on the cylinder draincocks, fixing on and trying out the valve gear in a running test, cleaning up/disassembly, painting, adding couplings and remaining tender wheels, then reassembly and retesting. Not to mention renumbering and weathering the whole thing. Tempting to hurry all this but my frustration endures from a year ago making a not wholly satisfactory paint job on my Barclay to have it ready for Model Rail Scotland, where however it was actually needed, not the case with this Crab.

The loco weighs around 400g now plus around 100g coming from the 160g tender so I hope it may master a 1 in 70 with a decent load! - but the pony truck takes a good bit of the weight at the front, about half and half with the front wheel. With the tender attached the C of G is over the centre driving wheels. Let's hope those massive cylinders don't destroy too many platform edges...but my main satisfaction will derive from it never derailing! - yet to be proved except in the tests shown already. So look forward to trying it on your WCML layout! :thumb I will attempt to make the draincocks clear the pony truck swing for 3 ft 6in curves. Next post awaiting a Roun Tuit will show my son's spreadsheet for working out the clearance required!! 8-)

User avatar
Allan Goodwillie
Posts: 667
Joined: Sun Jul 20, 2008 7:00 pm

Re: Crab Comet conversion

Postby Allan Goodwillie » Sat Apr 07, 2018 5:27 pm

Hi Julian, :)

I can always move the platform!

Should be good to see it go and you are right about racing to finish something for a target, it means getting everything just right first go and that seldom happens if at all. We are under pressure this weekend to try to get the buildings finished or as near so, on Burntisland. Lindsay and I had a session with the builders yesterday and a lot of ground was covered - we have six weeks to get all the buildings finished before Railex and I think we might make it with another couple of painting sessions. Interiors will have to wait until the next outing, but I am already thinking about mine so that I can complete the blacksmiths shop to be able to get on to my own workshop complex for the Wemyss line.

Not all the crabs had steam sanders and you are right about the second pipe.

I brought Mike Gilgannon's buildings home and have also brought the trackbed of his layout for possible inclusion on some future project.

Will see you at the weekend at Scalefour North

Allan :)

Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Sun Apr 22, 2018 6:24 pm

A professional sounding post would read thus:

"The front and centre tender wheels that bear no load as per the Sharman arrangement were installed, with the brake gear. Springing them adequately necessitated some extra weight in the water space."

Tender with brake gear. Still sitting a little high. The lower steps should line up, not the higher ones.

I'm not sure if the following fuller description of the tribulation involved would be better left unsaid. That brief paragraph absurdly minimises the effort and hours involved, the best part of 5 days modelling time. Any remaining readers can see what to avoid which might be helpful. Here goes, for better or worse.

A diversion to the tender was prompted by the arrival in the post of a package from a kind friend with a lathe, who had turned down the original Bachmann tender wheels to be flangeless.

As a reminder, my approach has been to arrange the tender Sharman style, so it sits on the back of the loco a bit like an unbalanced caravan might onto a car. The idea is that the tender weight is transferred to the loco increasing the adhesion and hence haulage power. Additionally it balances the weight of the loco where, as in this case, it is easier to add weight towards the front than in the firebox where the motor is located. In this mode only the rear wheels of the tender are actually needed. The 1st and 2nd wheelsets contribute nothing, simply trundling along to look right.

To this end the tender has been piled high with lead at the front where coal should go and where the extra weight has the most effect on the loco , but it is quite light at the rear.

I should mention again that my test bed is an approx 5ft length of track with a reverse curve of 4ft radius, made with Exactoscale Fast Track, not gauge widened (which, in theory at that radius, it should be). Nothing at all to look at but it really seems to test out any issues. There is always one place or another that catches out any fault with whatever rolling stock item is being made. I am not satisfied with any item unless it can stay on the rails of this little test bed, and it must work backwards and forwards and both ways round. This tender proved a very devil to pass the test.

Greybeards probably would have seen what is coming. It has been very difficult to arrange the 1st and 2nd tender wheelsets to cope with the 4ft radius curve test bed. (I even had hopes of a 3ft 6in curve being practical, which the loco centre wheel sideplay should be enough to allow.) Yes, it's easy enough to have the rear wheelset trailing along. But the front wheelset needs a lot of sideplay, the centre wheelset almost as much. With the length of this loco much more sideplay is needed than my previous experience, the relatively short CR Class 812 0-6-0. Additionally the Bachmann plastic model side frames are appreciably thicker than those on the 812 etched brass kit.

Seeing the problem of the sideframes I had asked my friend from the club to turn down the flanges of the Bachmann wheels. He did a perfect job but even these 00 gauge wheels are not wide enough to stay on the rails, for the front wheelset at least, allowing for the slewing of the track on the reverse curve.

Flangeless wheel experiment. Weighted with lead, not sprung.

However it is possible to thin the plastic sideframes sufficiently to cope with P4 wheelsets and it is practical to allow them plenty of sideplay.

Then the issue becomes, how to spring them sufficiently forcefully downwards to stay on the rails given the quite extreme sideways forces exerted on them - without negating the whole point of having them NOT bearing the front of the tender weight, which is needed to balance and add to the weight of the loco.

First go at springing the front wheelset was successful, and the wheels stayed on the track completely reliably. So that was fine, but the spring force was coming from the centre of the tender and reducing the weight on the loco. Getting the spring force further back was needed.

The water scoop is visible here before being
removed. Also the centre fixing screw, first springing experiment for front wheels, and to left an extra spacer on which the AJ can be soldered.

That was OK - but any problem and the tender lurched over rather than the wheel derailing. Clearly the idea of concentrating the weight at the front was insufficient. So all remaining space (the water space) was filled with lead, quite a fiddly job with the various plastic bits and pieces for the body assembly. The tender now weighs 267g! - with the C of G over the centre.

Springs and everything to do with them is my least favourite aspect of modelling. I am always in the dark about them and am just reporting my blunderings and probably half baked ideas. Guitar/Piano wire is the material that best retains its spring I think but is harder to solder on reliably. I bent a right angle so that it could be fixed on two perpendicular surfaces. 0.5mm wire was used, 2 springs for the front wheelset and one for the centre. The spring force is adjustable by bending the wire.

Springs fixed further to rear, an extra piece of square tubing has been attached for the front wheelset springs.

Springs slide in guides

A constraint was that the chassis is fixed to the body with a single screw above the centre wheelset, which I had assumed would be unreachable once the axle was positioned there. It now proves possible though tricky to un/screw with the wheelset in place. Thus there shouldn't be much need to attach or remove the centre wheelset from the chassis.

My idea had been to retain the wheels against the springs with the brake pull rods, with brakes removable at their top from their fixings. However I didn't foresee that there was insufficient space to do this at the top of the brake yokes (? correct word?) but having soldered up the assembly they were too well attached to the cross shafts and pull rods at the bottom to be unsoldered without risking damage. So the brake gear is a bit of a mash up with a piece of strip to be unsoldered if the centre wheelset needs removing, and cross shafts cut away to give room for the springs.

With brake gear

Anyway from the side, with the wheels in running position, the springs are pretty much invisible. I did away with the water scoop, though if I was feeling terribly thirsty for ultimate fidelity something of it could be reinstated around that rear spring. Compare first photo with this side view of water scoop - it is not very evident I think.
Water scoop before removal

A further post will be added soon.

Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Sun Apr 22, 2018 6:51 pm

This is in addition to the post submitted a few minutes ago.

On return from Scalefour North I find that the weight transfer to the loco is only around 40g. It is difficult to measure on kitchen scales as the slightest unsteadiness affects the reading dramatically. But it makes sense considering the spring force needed to keep the wheels on the track.

Also I find the wheels make the tender level. So I have filed the axle slots upwards and somewhat reduced the strength of the springs. The only other thing I can think of doing is to pile even more lead into the coal space. There is room for a higher load of coal.

I don't know if the idea of moving the spring fixtures back was any use or even perhaps counter productive. In fact I can't see any way round the problem of springing these wheels adequately without negating the adhesion gain - this tender now weighing as much as at least 5 wagons.

However my aim is simply a loco that stays on the rails, not a heavy hauler. Next task, before piling on more "coal", and having completed the sanding gear, is to fine tune the spring at the front of the loco above the front driving wheels, after filling up any remaining voids with more weight. The loco may then be better balanced within itself and not rely on tender weight.

All these additions need to be tested before the final step of adding the previously completed valve gear and finally fixing on drain cocks and front step, for which the pony wheel sideplay spreadsheet I posted last week here will be invaluable.
Centre wheels sanding etc assembly. Only one sandpipe is fixed to the loco.
Front driving wheel brake and sanding gear etc assembly

Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Fri May 04, 2018 8:15 pm

Now that the chassis assemblies were complete, first task was to load every little space with pieces of lead. Here are two parts of that process.

20180423_131212 (Small).jpg
Spaces filled with lead.

20180423_131226 (Small).jpg
Lead...rather poor photo sorry

Now to test the newly nearly completed chassis before adding the valve gear. Miraculously the sandpipes were not creating any short circuits - but I am familiar here with the need for a rather greater clearance than strictly prototypical.

I had been caught out though - the D rings were fouling the coupling rods. The clearance must have been very minimal previously. Now that they were more substantial with a full depth backing piece there was no "give". Perhaps I may have bent them outwards minutely. Work with a file and strip of sandpaper was all that was required. Smooth running is needed before I fit the valve gear.

Now to the crux of the matter. With all the weight installed would the loco suspension work as intended?

On further thought it had become apparent to me that the Centre of Gravity needs to be half way between the pony wheel and rear driving wheel, not over the centre driving wheel as I had previously imagined. Given this argument, the loco is now perfectly balanced. It weighs 400g. The tender is no longer needed to balance anything. This weight, I expect, is perfectly adequate for haulage, even though some is not adhesive weight. Unlike previously, the loco on its own now proves to ride as perfectly backwards as forwards. I ascribe the change to the extra weight installed around the C of G. None has been added towards the rear.

So at this point if I took myself seriously, I would delete from this thread any reference to the Sharman style tender! But I don't. All this is for fun :? and part of the fun is the learning process. It's part of life's rich pattern that I spent 5 days on the tender suspension, for no reason as it turns out. I always think though that such things are never a waste of time and always something is learned. There are worse ways to use 5 days...

So the issue is now whether the weight on the pony wheel just exceeds the strength of the front wheel springs. It does. But with the tender hanging on the back, it does not. Partly this due to the cantilever effect whereby the tender rests on the loco coupling which is some way back from the rear wheels beam fulcrum point. If the rear wheels had been in fixed bearings the effect would have been less.

To remind anyone still out there of the suspension arrangement: the 3 legged stool principle applies - one leg is a central roller above the pony truck, the other two legs are the beam fulcrums between the rear two driving wheels. The front driving wheel is sprung with the kit's springs which have been pulled outwards - lengthened - (and the chassis slots for them deepened an equivalent amount) as described earlier around the New Year.

20180423_131022-1 (Small).jpg
Original size spring is the small one.
20180423_131022-1 (Small).jpg (27.91 KiB) Viewed 638 times

The test track, since the time I made it about 10 years ago, has a dip filed into one of the rails to enable the behaviour of each wheel to be seen. Now to increase the challenge the track gives I have put a half millimetre thick piece of brass under the edge of one of the sleepers on both the curves, to create some momentary negative cant. I don't subscribe to the idea that track can be perfect. I take the Sharman viewpoint that the challenge is to build a loco that will cope with inevitable track faults.

20180427_114612 (Small).jpg
Brass shim to give track fault

20180427_114313 (Small).jpg
Negative cant side view

The video shows the loco negotiating the track at full speed with no difficulty.

The low speed videos show the behaviour of the wheels over the rail dip. There is a difference that is hard to discern between the pony falling into the dip because of its own weight (and the effect of the springs installed in it), from when it goes into the dip with the loco falling with it. In the latter case the loco front can be seen to slightly jolt downwards, and the wheel seems to move less far from the loco. This is what I prefer. In the former case the wheel is very lightly loaded and can derail, as I found earlier this year, while in the latter case the weight of the front half of the loco is on it - 200g, minus whatever load the springs are taking...ideally the pony would take about 110g and the front wheel 90g. But as I found last autumn, the spring takes more weight as it compresses, which is what makes me think compensation is a better principle. Whatever the overall weight and whatever the vertical position of the wheel the load proportion will remain consistent. But in many cases, as here, a spring is just simpler to fit into the space available.

On the video showing the left hand side the pony satisfactorily takes the loco weight. But the front driving wheel does not fall into the dip in the same smart way as the following two. So it would seem the spring on the LHS could be made a little stronger.

On the RHS the opposite happens. The pony flops into the dip, but the front driver falls in with a more vigorous motion than on the other side. However I think that in reverse the loco weight goes with the pony rather more. It might seem the front driver spring is stronger this side.

Counter intuitively however the spring strengths are the opposite, and the RHS one is if anything slightly weaker. This baffled me but on further investigation it was apparent that the RHS wheel could tilt more easily.

This tilting aspect of model loco suspension is what Mike Sharman emphasizes so much in the first chapter of his book and which it seems to me is rather ignored in more contemporary hornblock bearings. The MJT bearings have the slot design as per Mike's drawing but the Comet bearings, and others I have, have simple parallel slots.

I had filed the slots wider on some of the bearings of this loco but was too impatient to be bothered to do them all, leaving them for a Rountuit session, and thinking that the design is probably perfectly adequate. I suspect this accounts for the difference here.

Whether I will bother is doubtful as the loco behaves perfectly well enough already to judge by the full speed test. The track dip is a very exaggerated test of what will be happening on most track faults.

The tender suspension and coupling will be modified to normal versions.

Detail remaining includes the drain cocks (and three more balance weights). A good photo that shows the below cylinder detail clearly is not something I've so far seen and I may have to confect what I can from Keith's clear but not side-on view. As no one looks much at this area I won't worry excessively about perfection here.

Philip Hall
Posts: 1047
Joined: Mon Aug 10, 2009 7:49 pm

Re: Crab Comet conversion

Postby Philip Hall » Sat May 05, 2018 10:20 am


I have just come across your late April post about the tender and have a few observations. I have built a few ‘weighted tender’ engines over the years, including modifying a number of RTR ones, and although I wanted to increase the haulage capacity my principle aim has to make the engine have the same weight, approximately, on each of the driving wheelsets. I do this very simply by picking up the engine by each wheelset in turn and comparing how heavy they feel. It has been relatively easy to detect an engine which is light at the rear or heavy at the front. If it seems a bit too much out of balance I consider hanging the tender on the rear dragbeam of the engine.

Now I used to do this in the usual Sharman/Rice fashion, by allowing the front and middle wheelsets to do their own thing and trundle along under their own weight, which as you found, isn’t terribly reliable. I tried light springing, or top acting pickups with vary degrees of success, or more often, lack of it. So my present system is to have the compensated tender underframe as a complete unit, which pivots under the body which has the weight in it. I usually put in about 80 grams, biased a bit towards the front. The tender underframe is supported by a transverse beam (0.6 or 0.7mm wire) on the top of the frames, midway between the middle and rear axles, and this rests on a little ledge either side of the tender body.

The effect of this is that all the wheels on the separate underframe are firmly held in contact with the rails, maybe a little less than the front but not enough to cause problems. The weight is transmitted down through the little beam to the rear wheelsets and to the midpoint of the compensation beam between the front and middle axles. The front bit of the weight bears down on the rear of the engine. The coupling between the two is a simple bar of wire, about 5mm long, on the rear of the engine, and a hook on the tender. The hook can be bent up or down to get the ride height of the tender front end correct. Pickup drag on the wheels is not a problem as there is some weight on the wheels.

Two further things: this can just as easily be done on a rigid tender, so long as there is a separate inner chassis, or a separate bottom half, as on the Hornby T9) and here I allow the middle wheelsets some upward play. I’ve not had derailments here. I imagine it could also be done on a sprung tender, although I’ve never tried it. I was a bit mystified by your problems with the tender on reverse curves, because most of the ones I’ve done in P4 and EM have gone around 3 ft or less with no trouble. The front and rear axles need next to no sideplay but the middle ones do need a bit. Maybe yours derailed because of the lightness of the wheelsets, or maybe the tender coupling did not swing enough? I think that if I ever did have this problem I would turn up a heavy brass sleeve to slip over the middle axle. The bar and hook arrangement does enable the tender to move around a bit.

I am about to have a go at a variation on this arrangement on a very lightweight 4-4-0 which has come to me for some remedial work. Previously I have got a Hornby T9 to waltz around with nine or ten heavy carriages, which seemed enough.

The Crab is looking good and runs very well, and I would be very pleased if one of my engines behaved as well as yours does on all your lumps and bumps!


Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Sun May 06, 2018 7:40 am

Hi Philip

Thanks for all your reply! Well, as you see, I realize to my relief I don't need the weighted tender for this Crab. But my P4 loco that is in temporary abandonment, a 4-4-0 LMS Compound, I do, as that has all the weight over the front bogie and very little where the motor is, and relies on the tender to balance itself. It has fixed rear drivers so there isn't much cantilever effect. But I had the same problem, of the tender wheels derailing. As I had bought the kit semi-made, and the tender was near complete, I put it back while I did more urgent, or as I thought in the case of the Crab, faster projects, as I could see a more radical approach was needed.

So when before too long I get back to the Compound I may follow your suggestions. But I think I'll come and find you at Scaleforum if you're doing a demo desk again as I don't fully follow what you've written here - are you there this year? What I found with the Compound too was that there is very little space in P4 between the outside of the wheels and the outer frame, so I would need to see what you have described I think to see how there is room for the chassis to act as a bogie.

I can not have been quite clear though: the coupling at the back of the Crab is totally fixed, there is no swing at all. So this means the tender is "tracking" (if that's a correct way to use the term) in a totally weird way, especially on reverse curves, and it's the front wheelset that needs such a lot of sideplay, in the way I made it.

Yes, the tender if made and coupled conventionally won't have any problem with sharp curves as it's not a long wheelbase and just the middle wheelset needs the sideplay as you have said. I'll write up what I do when it comes to that.

Meanwhile here is the valve gear with temporary fixings on dieblock and reverser set up ready for testing when the 243 has set on the return crank.

Thanks again

Philip Hall
Posts: 1047
Joined: Mon Aug 10, 2009 7:49 pm

Re: Crab Comet conversion

Postby Philip Hall » Sun May 06, 2018 1:03 pm

Julian, No demo at Scaleforum, just taking the photographs as usual. So I will be around the hall somewhere, and will be pleased to meet up and have a chat. I shall also be at Railex, not sure what day yet.

The tender weight idea is, as you say, not easy to describe. However, it is simply that instead of the tender body and chassis being 'hinged' on the rear wheels, with the front two flopping downwards, the whole inner chassis sits firmly on the track, but the body is 'hinged', midway between the middle and rear wheels. So there is weight all over the wheels but a large chunk of body weight can be born by the rear of the engine. I showed this in a T9 conversion article in MRJ many years ago; although the tender was an eight wheeler the principle is the same.

The fixed nature of your tender/engine coupling is, I am fairly sure, the reason for your derailments. If the front of the tender cannot swing at all then as it enters curves it cannot move; in effect the front axle is a continuation of the engine chassis, so it seems to me to be a very long wheelbase 2-8-0. To allow the tender to 'steer' an engine I would try a sprung coupling bar, so that it returns to centre as the curve is left behind.

No doubt we can have a look at the engine in September. Good luck!


Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Sun May 06, 2018 5:15 pm

Thanks again Philip. Yes if I can be there I'll search you out.

Just to clarify, I did manage to get this tender derailment free even with the fixed coupling but at the cost of so much spring force that there was almost no weight transfer to the loco. End of that approach!...

The Crab now can almost be said to be in steam. I'll write up what was necessary to achieve this separately.

Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Mon May 07, 2018 7:14 pm

For anyone who like me finds Walschaerts gear a not undaunting prospect, here is a list of the problems I encountered:

Capture Crab fouling points 2.JPG
Capture Crab fouling points 2.JPG (80.56 KiB) Viewed 318 times

Taking the picture to which I've added red lines which indicate the problem area. In clockwise order from lower right:

1. The return crank – I put a washer on top of the connecting rod to bring the return crank to the correct point for glueing with Loctite 243. It was filed till it was the right thickness. As this is not the final assembly, I have not checked from photos etc that the angle is absolutely right. The further it is inclined from the centre, the more movement there is.

2 and 6. The eccentric rod just fouled the connecting rod until I arranged the expansion link to be situated as far outwards as possible in the generous envelope available for it, by fixing spacing material on both on the dieblock assembly and the expansion link itself. Both these extra bits involved very careful soldering. The fouling was minimal on straight track but on a RH curve became unacceptable. This adjustment means that the radius rod is not quite parallel to the footplate, it goes inwards towards the combination lever.

20180505_094650 (Small).jpg
Washer soldered onto expansion link with cocktail stick.

20180505_125713-1 (Small).jpg
2 layers of 0.5mm thick material soldered on with 70 deg.
20180505_125713-1 (Small).jpg (28.9 KiB) Viewed 318 times

If I had soldered the rods using pins instead of rivets and filed carefully there may have been less problem. The back of the rivet, I assume, can't be filed away to the same extent. The rivets came with the kit but I wonder if pins would look less obtrusive even if they are more difficult to do.
Fouling occurred between eccentric and connecting rod at this position

3. (The bottom of the D ring was fouling the coupling rod at the previous stage.)

4. More filing to be sure the top of the combination lever etc doesn’t foul the stretcher assembly.

5. It was not fouling, but I bent the reverser into a cranked shape as per the prototype.

7. The expansion link on the prototype goes behind the footplate valance at the furthest back position. I had to file a recess in the Bachmann footplate.

Capture motion 1.JPG

20171119_103759 (Large).jpg
The expansion link goes behind the valance when at that position of its travel.

Regarding point 7: The expansion link, if following the instructions, will not foul the footplate. But the instructions say to use the hole in it that is not central. I filed that hole away to make the upper portion a more correct length. I suppose the link is in total the correct length but the compromise of having the hole more than half way up is to avoid the necessity of filing the mazac with the possibility of going right through it.

Edit - please see following post.

Regarding the video posted above, running-in should smooth out the low speed movement. Stickiness is not coming from the valve gear - more likely dirty wheels and pickups, as they were not cleaned for this performance. The top speed in the previous videos seems rather high for its intended purpose lumbering along with and shunting heavy freight trains. I may change the ratio to 120, that will help.

Two pictures show the full horror of this contraption that obviously will need a good clean up before painting. The cranked reverser can be seen, and its connection to the driver's control still needs to be done (though most of it goes behind or through the sandbox so won't exist). The missing pipes from the front sandboxes are not going to cause insomnia. Proper fixtures for the dieblock and reverser have still to be added. The gap at the top of the motion bracket has been filled though not fully made good.
20180505_160058 (Small).jpg
20180505_160044 (Small).jpg

Julian Roberts
Posts: 571
Joined: Sat Jan 09, 2010 11:33 pm

Re: Crab Comet conversion

Postby Julian Roberts » Tue May 08, 2018 1:45 pm

A clarification on Point 7 in previous post regarding the expansion link.

Here are two pictures that show the two etchings that together make up each link.
Capture Expansion link original.JPG
Capture Expansion link original.JPG (27.49 KiB) Viewed 230 times

20180508_104556-1 (Small).jpg
20180508_104556-1 (Small).jpg (28.48 KiB) Viewed 230 times

It can be seen that the one with just one hole has that hole not quite in the central half way up place. Thus the top half is shorter than the bottom half. So, if made up that way, the top will clear the Bachmann footplate.

The one with several holes is for three options of having the loco in various positions of the forward gear (depending on the cut off selected on the real thing).

What I did was to put the central pin one hole lower than it is meant to be, and filed away the hole of the one with just one hole.

Of course I don't know for sure if all this is pukka - I'm just going from the pictures as I see them, I'm an amateur. And all this is headbanging stuff that probably most people won't be bothered with, don't know why I am, there's something wrong with me clearly, it just gets to me with a buzz thinking this kind of thing through, even if what I conclude is total b...ocks.

Here is a pic of the recess needed. I'm in danger of going right through.
20180508_120041-1 (Small).jpg
20180508_120041-1 (Small).jpg (17.84 KiB) Viewed 230 times

The top of the link needs to be properly clear from the frame, If jiggled with tweezers at that furthest back position it needs to move as freely as it does when clear and further forward.

Return to “Steam Locomotives”

Who is online

Users browsing this forum: No registered users and 1 guest