Simond

In the original of this post, I referred to an error in the servo program - fixed it in the attachment above. This post has also been edited to take the program out of the body of the thread and attach it. Now, as I said earlier, you need to know what the limits of your servo travel are. I think that the best results would be if the servo travels as much as possible, by using the holes in the horns as near to the axis as possible - but if you see my photos, link below, I haven't done this, so the action is a bit quick, and the settings are quite critical. To find the right settings for the servo end points, I used a very simple program, listed below, which uses a potentiometer on an analog input of the Arduino. The program makes a ratio between the voltage on the pot wiper and +5V, converts it to an output for the servo, and moves the servo accordingly. This is straight from the Arduino "examples", with the addition of the few lines to print the servo output to the serial monitor. This allows you to note down the correct positions, and to then implement them in the program that controls the servos. The program refers to the signal I posted at #22 in http://www.rmweb.co.uk/community/index.php?/topic/79052-porth-dinllaen-in-0/&do=findComment&comment=1272968 The servos should not normally be powered from the Arduino - but you will probably get away with it for this exercise, because you are only operating one at a time, and it is not likely to cause any malfunction, but if you have several, they will cause the Arduino to "brown out" and reset itself. Anyway, for the purposes of this exercise, connect one end of the pot to +5V, and the other to GND. Connect the wiper to pin A0. Connect the red wire on the servo to +5V and the black to GND, and the signal wire to pin 9. Run the program with servo not connected to signal. Centre servo. Connect to signal, use pot (gently) to find the end position, and write it down. Ditto other end position, write that down too. Use those positions in the operation program by inputting the numbers in this statement int sigbracket[2]={ 70,91}; eg, "70" is normal, "91" is pulled. If I had ensured that the operating wires were in holes nearer the axis, there would have been a bigger gap between these numbers (but I would have also had to make sure that the servo horn didn't hit anything as it would travel through a greater angle for the same linear displacement). It doesn't matter which is bigger or smaller, but do make sure you know which is which! Hope this makes sense! SD servo_setting_with_pot.ino

Jon, all You'll find a link to Arduino control in my signature below Hope it is useful SD http://www.rmweb.co.uk/community/index.php?/topic/82978-arduino-applications-and-programs/ Ok, link here, must have messed up again! SD

Following Andy's adjustments to the uploading permissions, I have attached the sketch, rather than embedded it in the body of the post. Servo control v4 is a very simple program to operate three servos from three single pole on-off (or change over) switches. In order to use the sketch, you need to download it into a directory whose name is the same as the sketch. You can then open it in the Arduino proigram and upload to the Arduino itself. In order to use this sketch, you will need to adjust the end positions of the servos. The sketch for doing this will be attached to the next post. best SD bracket_signal_v4.ino

Hi to one and all. over the last 5 or 6 months that I have been involved in RMWeb, there have been many productive and interesting discussions about Arduinos and their applications, including stepper motor driven traversers and turntables, and signal / point control. I promised to post some sketches (otherwise known as programs, source code, etc.,) and rather than have it in different threads, I thought I would start a new topic, and add pointers to the threads where relevant. I hope this makes sense, and doesnt cause the Mods a brownout. I firstly bought a Raspberry Pi when they first came out - I have had very little success with this, probably due to a lack of finding the right examples and bits of source code to copy from. Also, Linux is to me, at least, a bit of a black art, and my earlier programming experience was in Algol 68 (yes, I said "earlier"), Fortran and Basic. The Arduino is programmed in C, which is pretty much like Basic, without the line numbers, and many examples are included in the IDE which you can download here ;- http://arduino.cc/en/Main/Software The Arduino website has a full language and syntax reference, plus links to various forums (fora?) in which help seems to be readily available. http://arduino.cc/en/Reference/HomePage I have no connection to any of the traders or sites mentioned. The code I have posted is available to download on an "as is" basis with no warranty of any sort, and is free to use for non-profit - please do credit me (or any other contributors) where applicable. IF you are planning to incorporate it in something you plan to sell, please PM me and we can discuss crossing my palms with gold... first instalment to follow very shortly... best SD

"More reply options" & "choose file" Bon chance! SD

Ray, all Quickie note as I'm travelling - I couldn't get a Hall effect sensor in Maplin, so I bought an optical reflectance sensor, which I wired up with the aid of two resistors to an analog input. I glued a piece of white plastic card under the t/t which clears the sensor by about a millimetre. The input value changes from around 700 to around 50 when the card covers the sensor. I have set it up to act as a zero reference point, but the software needs to be finished. It works, anyway The sensor is tiny, around 3mm square, and cost £1.09 http://www.maplin.co.uk/p/photoreflector-ja26d Best SD

Ray You put your sensor on the outside of the pit, and your magnet on the table - logical! - but I wonder how far away you could have sensor and still get a reliable switch point. The Metalsmiths t/t is mounted on an MDF pit, and I don't really fancy digging up the reasonably-well scenic-ed bit to install it. It's about 6 mm thick. Will the sensor work reliably at that distance, plus the one or two mm of clearance? I can hide it under the centre of the table reasonably easily, but then it's angular discrimination will be much less than if it were mounted at the edge. I wonder if I could just stick it down in the pit somewhere and paint it to match the gravel, and hope nobody notices.....? I'm planning to use something like this http://www.farnell.com/datasheets/608704.pdf Sound reasonable? SD

Hi Anthony now I understand!!! it is indeed a 4-bar linkage or mechanism. Comments as follows; the lengths of the bars (yellow) is absolutely critical, as is the position of the pivots. I really would not use threaded rod as a pivot - it would be worth going to see your local blacksmith / engineering job shop and getting some really nice pivots made, as anything running on the threads will wear and go sloppy and then your lifting bit won't lift level - well,. it won't anyway, as you've shown in your sketch, but it will become a bit unpredictable and may even jam. This is particularly likely when the bars are parallel or very nearly so, and there is a risk of them getting crossed - please do look at the available analyses on google - http://en.wikipedia.org/wiki/Four-bar_linkage is a good starting point and this may also help http://www.mekanizmalar.com/fourbar01.html the risky area is when the lifting piece is just going back into place and the bars are parallel and particularly when one bar crosses the pivot of the other; If you like, it would be like a steam engine without quartering on its wheels. Obviously, one pair of bars needs to be outside the other pair, if they are going to cross, and you've shown this in your original post. I think I would suggest having a bracket below the lifting part, and below the fixed part to which one of the two pairs of bars can be fixed - this will prevent them "lying on top of each other" as they do in the first of your pair of pictures. This way, there is better separation and they can be in the same planes. hope this is clear and helps! Simon

Anthony sorry if I'm being dim, but I can't visualise it - any chance you could post a sketch of a side view - end on to the rods? SD

Paul Thanks for your diligent efforts in posting the details of hard and software - makes it much easier to follow what you gave done and what you have achieved. I wish you well, good luck in your efforts and keep posting! SD

Anthony Trying to understand your plan for the lifting section: your diagram appears to be a plan (ie bird's eye) view of the section, with the hinge line in the left - are you planning to use the threaded rod as a hinge pin? I'm not sure what the other rods are for - are they to stiffen the lifting panel? If so, I think metal tubes (eg aerial tube 30mm ally) would be both lighter and stiffer, particularly if you glued it into the framework - would also provide comfortable handles as a free extra! I think maybe you are planning to have the section pivot on two pairs of parallel pivots - a "four bar mechanism" in which case Google is your friend - you'll find lots of worked examples so you can determine the geometry. You will need to be pretty accurate in your drilling or it'll not quite behave as you want - but given what you've achieved to date, a non issue I guess! I think I would recommend using proper hinges, rather than pivoting things on threaded rod. If it wears, it will get misaligned, and that'll be a PITA. You could use the threaded rod to provide adjustment of the lengths of your links. Following with interest SD

Hi all, An enjoyable & stimulating thread! I have been toying with a Minories-based track plan for my intended "might have been" harbour terminus, Porth Dinllaen - link to thread below. I imagine the addition of the access line from the down main to platform 3, with a further siding and a kick back to form a head shunt/siding alongside the down, and a double to single junction to feed the freight part of the dock. I don't recall the CJF version of this, but may have seen it, so I claim no original thought there! Loco release crossovers to be provided. I'll post a sketch later today all being well. I'd be interested in the opinions of the cognoscenti regarding the likelihood / reasonableness / utility of such a track plan in such a location. It seems to me that reclaimed / built-up land within harbour walls may be as difficult, if not as expensive, as land in a city, so the compact aspects of Freezer's design appeal, and the operation potential appears to be excellent. Best, SD

Jon I think you can be proud of that! Simon

Chris the loco's lovely. You mention that you have three point compensation using the two front axles, so I suspect a rocking beam. What did you do with the truck, which I suspect would have been radial on the real thing - indeed as it appears on your model? Is it actually a well disguised pony truck, or is there some other means of suspension/control employed? best SD

Anthony I remain staggered by the scale of this - it is outstanding, and I'm very much looking forward to seeing some video, but I just wanted to commend you on your track diagrams - every bit cartographic works of art! You have given me something to aim for in my own musings!!! best SD

Hi Don You are right, of course, that there is a "side thrust" (in this case up and down) from any crank and slider mechanism, and this is reacted by the crosshead on the slide bars. Going forwards, when the crank is "up", the steam pressure is on the back of the piston and the crosshead is pressed against the upper slidebar, and when the crank is down, the steam pressure is on the front of the piston, and the side thrust again is pushing the crosshead against the upper slidebar. The magnitude of the side thrust is determined by the angle of the connecting rod, which depends on its length, and the crank throw (and the angle of the crank and the steam chest pressure, of course), and any offset, such as the one we are discussing. The bigger the offset, the bigger the side force, and so, the upwards offset would reduce the side force on the "pull" stroke, and increase it on the "push" stroke, which is what I was suggesting in the last paragraph of post 30. The friction that would exist between the bars and cross head slippers would be related to the instantaneous loading, so the thrust might have been more even in the two parts of the cycle as a result of the offset, however the wear is related to the average loading, and I don't think that the offset changes anything in this regard, as the reduction and increase would effectively cancel one another out, the average load on the upper (or lower) slidebar would be the same with them offset, either way, or central. Your suggestion of this side thrust being countered by the weight of the parts is interesting. Clearly this is only going to work when the side thrust is upwards - which it clearly is when the loco is going forwards (irrespective of the offset). There is another consideration. As the loco accelerates, the forces on the connecting rod are augmented by the inertial loads (sort-of-centrifugal force) caused by their sort-of circular motion. It's obviously possible to get the weights of the parts, and the dimensions & loco performance are also available. So it should be possible to work out all the forces, deduct the weights, and see whether we end up with a lower overall side thrust, thus lower friction & losses, and lower wear. All we need now is enough energy to sit down and do the calculations (which are not trivial or I'd do them!). That, or a couple of engineering students... Of course, even if the calculations confirm that, at some reasonable speed, it all looks good, and would benefit the efficiency and/or longevity of the loco, it doesn't prove that this is the reason GJC did it... Fun speculating though! SD

and just for the record, 28xx (scratch built) and 47xx (Martin Finney kit) both showing slight angle on connecting rod, confirming that the cylinder axis is above the wheel axis. best SD

Further info; - Page 62 of Holcroft's "Locomotive Adventure", he talks about his early work after transfer to the DO at Swindon - he was working on a scheme to rebuild the 80 single-wheeler locos into 4-4-0's - "he (Churchward) was prepared to see the cylinder centre lines pass 3 1/2 in. above the driving centre as against 2 1/2 in. in his standard types". This was some time between 1906 when Holcroft transferred from Wolverhampton, and 1908 when they started to scrap the single-wheelers, so presumably, the "standard" was in place before 1906. This would agree with Russell above. Reading further in the Russell book, there is an interesting reference to the County tanks - page 60 - the final ten (2241-50) of these locos did not have the 2 1/2 inch offset between the axes. These were apparently built to Lot 188 in 1912, and also had the curved drop to the front frames - which apparently was designed by Holcroft. The same change appears to be the case on the County 4-4-0's the later (1912), curved frame, versions appear to have cylinders aligned with wheels, whereas the earlier (1906) versions have an offset. The first of the 43xx class was designed and produced in 1910, and appears to incorporate the offset - it's certainly there in all the drawings of Moguls, and of their rather larger cousin, the 47xx, in Russell's book. Holcroft merely reports that he used "a No.4 boiler and brought in all the standard parts he could" as instructed by GJC. Holcroft moved to Ashford in 1914. So, in large-wheeled engines, the offset between the cylinder axis and driving centre was discontinued in new builds after about 1912, but remained, for everything with smaller than 6'6" wheels. Presumably, it worked, and they had all the jigs, and tooling, so perhaps a case of "why change it?" The 4-cylinder engines (Stars, the Great Bear, Castles and Kings) all had the cylinders in line with the axles, and a quick inspection of the other drawings I have suggests all the inside cylinder loco's did too, although these were typically inclined. One other speculation - perhaps it was an attempt to improve efficiency when running forwards, at the expense of it in reverse - the effective areas of the piston are different, by the area of the piston rod. The offset, upwards on the cylinders, would favour the forward going piston, ie lower area, stroke. Not impossible, but abit difficult to work out whether the effects cancel. Unless someone knows, I leave it as a speculation! best SD

Further thoughts on this topic - the Halls and 28xx shared boilers and cylinder castings, amongst other things, and the shared them with quite a range of other locos too - the Moguls and Prairies particularly - Mr Churchward was very keen on standardisation, which was economically important for a number of reasons - reduced tooling costs, reduced drawing office costs, reduced spare parts counts, reduced test and development costs, amongst them. He was a very able engineer, and supported at various points in his career by Messrs Stanier, Collet and Holcroft, none of which were intellectual slouches either. The theory that the cylinders were placed above the wheel centre line to avoid fouling the loading gauge seems somehow spurious to me - the first reference I can find to it is in Jim Russell's book, "A pictorial record of GW Engines vol2", and whilst I hesitate to disagree with such an esteemed and respected historian & modeller, it doesn't make sense to me. He writes, on page 34, referring No 98, "In order to avoid any possibility of fouling the low narrow part of the loading gauge, the centre line of the cylinders was 2 1/2" above that of the driving wheel centres". He goes on to refer to numbers 97 (the forerunner of the 28xx's) and 99, (the large prairie), both of which also had the same cylinder castings and the same 2.5" offset between cylinder axis and wheel axis. I suspect that this has become "an accepted fact" and part of the lore of our hobby and interest, but I think it's a wrong assumption. Why don't I accept the "loading gauge" rationale? Simply because numbers 98 and 100 had 6'6" drivers, placing the cylinder centreline at 3' 5.5" above rail level with new wheels, and nominal weight, whereas, number 99 with 6' drivers, would have the cylinders 3" lower, and number 97, with 4'7" wheels would have the cylinders at 2'6" above rail level - nearly a foot lower!!! So, whilst fouling the loading gauge might have been a problem for the small-wheeled freight loco, it certainly would not have been a reason to raise the cylinders on any of the larger wheeled locos. And, to add to the entertainment, the wheels certainly got turned during the life of the loco, so they got smaller too, which lowered the entire loco. And the springs sagged, so the cylinders would have got lower, relative to the wheel centreline as well. Turning to the loading gauge itself, there appear to be two ways to foul it - I have not yet found a dated drawing of the GW gauge, and to complicate matters, there are several; There is a "static" gauge for each specific vehicle, a "dynamic" (or "kinematic") gauge for all vehicles, and a "structure" gauge. There has to be clearance between the dynamic and structure gauges to prevent collisions between vehicle and lineside furniture, and the dynamic gauge might be speed dependent (although I believe not, with the exception of cureve radii). In principle, you could foul the gauge by dipping into the zone below the 6" line, or you could go outside the width. The "low foul" would appear to be possible if the cylinders were low, and the drain cocks protruded below the line but even with the springs down an inch, the wheels down an inch (two on diameter) and the lowest cylinder centre line, the underneath of a cylinder of 26 inches diameter over its casings would be 15 inches above the rails - certainly leaving room for the drain cocks. Width-ways, it is much more of an issue, particularly dynamically, as the loco has clearances between flanges and rails, between wheels and frames, and will also flex - all of which will vary from loco to loco, with loading and between locations. Russell's book shows on page 41, a composite loading gauge diagram for the 4-6-0 locos, apparently a Swindon drawing reproduced, and this gives an 8'11" dimension over the cylinder casings, and that this dimension appears to be allowed from just above the bogie axle centreline - unfortunately no vertical dimension for this on the drawing - but this would certainly allow for the 2-8-0 lower cylinder centre line, right up to the 3'6" line, and 9' above.. This of course exceeds the current 8'8" dimension but appears inevitable if you have outside cylinders - they have to be farther apart than the outside of the wheels, plus coupling rod thickness, plus bearings. (strictly, you could put the connecting rods inside the coupling rods but this wasn't common on "big" engines). I therefore conclude that the loading gauges were somehow different but I don't have access to dimensioned, dated ones. Surely someone will! Perhaps they can clarify. Given the data I can find, I conclude that the reason for the offset cylinder centreline was not loading gauge, which begs the question of "why did they do it?". In a piston / crank assembly, it would be normal practice to start off with the axes aligned. There might be some advantages in arranging a straighter "push" on the power stroke, thus offsetting one axis relative to the other to reduce side forces, thus friction and wear, and I am aware of some IC engines that do have offsets between bore and crank axes for precisely this reason, but a steam engine "pushes" and "pulls" - it's double acting, (a "one-stroke", if you like!) so what you would gain on a straighter line of action on one side, you would lose on the other stroke. So that doesn't offer an immediate answer either. I have Holcroft's books, which I will dip into again, perhaps there is some insight therein. One theory I can suggest is that the cylinders needed to be at a minimum height to clear the gauge for the small wheeled locos and this may have been the deciding factor in setting the lowest cylinder axis relative to rail level. If the castings & boiler were common to both small and larger wheeled locos, the necessary clearance between wheels and boiler would locate the boiler relative to the wheels, and thus axles, given maximum wheel diameter, and given clearances. I suspect that this sets the distance between the cylinder and wheel axes, calculating which might have been a drawing office exercise right at the beginning of Churchward's career. Meanwhile, I remain curious! best SD

Hi Unclebobkt Thanks for the note - I can confirm it is a trick of the light, or the angle, or something! The loco is a Springside kit of a 49xx unrebuilt Hall, which I built as 5903, Keele Hall, built pretty much as the instructions. As you can see in the rather poor photo below, the cylinder axis is dead in line with the axle centres, which agrees with all the drawings I have. Best SD

Thanks all, I do agree, working signals are almost as important as working trains! If anyone is interested, I have posted some details of my current layout at http://www.rmweb.co.uk/community/index.php?/topic/80443-great-windowledge-railway/ enjoy! SD

Anthony Happy New Year! For sure, the plasticard version is detailed, striking & credible, but it may detract from the foreground if too much so. Also, as you have to "break" the scene somewhere, doing it at a sharp demarcation between painted backscene and top-of-wall might be a bit too harsh. You could perhaps try to separate the wall from the painted flat by a few mm, or a few 10s of mm. This looks ok unless something casts a shadow on both, but shadows on backscenes are a pain anyway! I suspect that given the distance between the front scene and the back, I would tend to want to lose the detail - either by not having the plasticard at all, or by toning it all down so there was no contrast between wall and mortar. Hope it helps SD

First attempt at an O gauge signal Kit from Scale Signal Supply about 10 years ago - been sitting, part built, looking at me for most of that time! Driven by servo motors, in the style of Jon Fitness' models of this parish, to whom thanks for inspiring me to get my finger out - if imitation is indeed the sincerest form of flattery, I hope the imitation is good enough! The servo motors are controlled by an Arduino Uno, I'm just playing with interlocking using that. It seems to me that this is a cheap and effective means of signal control, at about £3 per servo, and £20 per Arduino, which can drive at least 7 motors. Learning points - don't wire the lamps in series - they are too dim, that said, it is convenient to drive them off the 6V supply for the motors, and I hadn't planned that when I fitted the lights. Difficult bits - handrails!

Don, Couple of thoughts Reaching over corners - you could move your t/t and have a gap behind the backscene into which you can "duck under" - not something you'd want to do very frequently, but certainly a possibility. Whether you'd be able to get the t/t in somewhere else, I'm not sure. Curved turnouts - I have been doing some planning for the loco shed for Porth Dinllaen, and I started in my usual TurboCad, doing board outlines, etc., then went over to Templot to do the track plan in detail. This rather surprised me - curved turnouts take up much more space than I expected, but I guess you would still save space by using them - my problem was to keep a minimum radius that my 2-8-0's & 4-6-0's will go round - about 1800mm. I'd really recommend Templot - it is not easy to learn, indeed, it can be downright frustrating at times - but once you get your head around it, I think it will allow you make your errors "virtually", which is considerably cheaper and less frustrating than finding that your layout doesn't work when you've spent time & money on it. And it's now a free download... I like the through station in your upper plan, but the lower plan, as modified by Stu, does offer the opportunity to do terminus to fy via through station all on one level. I think you have enough length in the upper plan in order to be able to get the height difference you would need to cross one line over/under the other. Hope it helps Merry Christmas SD

Thoroughly enjoyed reading this forum - I really cannot offer adequate words - the layout is magnificent Where will it be exhibited in 2014? Best, SD