Printed from www.nzmeccano.com
A reliable Meccanograph
I'd not built a Meccanograph before, but on a couple of occasions I noticed that a fellow club member's Meccanograph always attracted the kids at any display. And it looked like fun, so I cast around for an easyish one to build. I came across John Ince's Modelplan 138 "Two Schmidt Coupling Meccanographs", which he sells for just $10 over here (which was about 3 quid), so I bought one.
The first Meccanograph is obviously very basic, but the second looked like it had a couple of interesting mechanisms in it.
The original design
The basic design of the original Modelplan is a simple framework measuring 12½ x 13 x 3½" made up with angle girders. From the operating point of view, the controls are on the front, and the handle on the right-hand side. The pen mechanism is a clever parallelogram whereby the pen moves according to the vector sum of the movements of the left and right hand cams. Separately, the table both rotates and moves from side to side, and the sum of these movements creates the pattern.
In order for the drive to the table to continue wherever the table is resting, there is a Schmidt coupling between the main gearbox and the table. More information on these couplings can be found here, albeit in German. The term is translated from Schmidt-Kupplung, which is in fact a registered trademark made only by the company of that name.
So much for the two basic tricks in this Meccanograph. It was a little disappointing as built, but even to a complete novice there were many items that screamed for attention...
As shown, the Meccanograph used a crank bolted to a faceplate to support the pin driving the two long arms. You were supposed to adjust the throw of these arms by unbolting the crank and then tightening it up when it was aligned as you wished. This broke my first rule of Meccano (never rely on a nut and bolt to stop two parts rotating), and was hopeless. In addition, the faceplate wasn't really tough enough for the job.
I have seen assorted bodges (including the use of a special strip Ashok makes for the purpose) to enable Meccanograph arms to be adjusted in ¼" increments. It seemed to me much more "Meccanoish" not to give up like this, but to use the obvious X-series strips as handles. Not only do I have quarter-inch adjustment, I even have very fine adjustment left and right!
It's a little tricky to see from here, but I improved the pen lifting arrangement too. I've seen a number of Meccanographs have trouble with starting and stopping, whether it produced untidiness on the pattern or required extreme skill and care. The entire pen assembly on this Meccanograph rocks on its central support. You can see (just) that the back of this central support has a 1" x ½" angle bracket at the back, and a long vertical axle goes through the slotted hole at the end of this bracket. As the pen arm rocks up and down this axle prevents any sideways movement.
The long diagonal axle in the picture above is the control arm, which is pivoted at the far end (rod/strip connector locknutted to an angle bracket bolted to the frame). Pulling down on this 'handle' (provided by a coupling) pulls the arm down because of a fishplate bolted to the top of the long angle bracket previously mentioned. A craftily placed pawl without boss on a strip is arranged such that this axle is 'caught' and prevented from popping back up. In the position you see (with the axle under the pawl), the axle is pushing down on the fishplate and rocking the pen arm down -- thus the pen at the other end of the arm is up (and off the paper).
To drop the pen, just move the control lever slightly to the left from under the pawl and gently release it upwards. The tension spring visible at the top centre holds the control axle up and out of the way ready for when you want to lift the pen again.
The traverse was also modified with a similar throw adjuster and a small X-series strip. This is particularly important as there is significant force on the table throw mechanism, and whatever adjustments possible are worth having here. Note that the X-strip and its attached crank are sitting loose on the vertical axles and can just be lifted off as required.
And, the most important part can be seen at the back of this picture. I always try to run all models from batteries (as you don't always have power at an exhibition, and if you do it's normally a pain in the bum having to dangle wires, remember to lug your transformer(s) along, and then have the health and safety police tut at your dodgy cabling. Forget it! I have standardised on one size of small 12V lead acid battery and bought a stack of them. They fit in pretty much any model (sometimes you can get two or three in if you wish), and are handy as counterweights. I have one or two on charge at all times and can just slip one into the model I want to run.
The battery shown is a 2.3Ah capacity, as you can see. It will run this Meccanograph continuously for at least a whole day (which can be over 100 pictures). It sits on the flat plate and is 'trapped' between the angle girder top right and the small angle bracket you can see on the 'top' of the battery. I find it easy to build in small spaces like this for a battery, but still have it so that the battery can be lifted out and changed in seconds without using a screwdriver.
At the top of this picture you can see the split of the main drive from the front cam (it arrives through the top left hand bevel). This drives the cross-shaft (holding the worm), which in turn drives the table translation (through the worm / 57t gear, then the small pinions and down to a contrate under the cam). To the top right, the cross-shaft also drives the main gearbox.
The Schmidt coupling is pretty much as is, and is certainly the highlight of this Meccanograph design. Three triangular plates are the key elements -- the outer ones bolted to bush wheels on the input (right) and output (left) shaft, and the central one floating. The corners of each triangle are connected together via 1½" narrow strips, all locknutted. The spacers and washers are simply to keep enough spacing such that the nuts and bolts don't foul each other. This is a lovely design that is a joy to watch.
I doubled up the 1½" narrow strips to prevent flexing, and you will be able to work out from the picture which nuts are tightened and which ones are the 'locknutted' ones.
The pen arm pivots on an axle in the fourth hole back along the braced girders (you can just see the end of this axle poking out of its pivot). The support for this axle was woefully inadequate, and I wanted to avoid building 'over the top' of the arm and hiding it with overengineering. As you can see, I used doubled-up angle girders, flat girders, and extra bracing to ensure this central pivot is absolutely solid -- a key part of the design to get accurate pictures.
The counterweights you can see on the right of the arm can be adjusted to give just the right pressure on the pen tip -- too much snarls the paper, too little doesn't give a continuous line. Ideally you want the lightest pressure possible that always draws a solid line. You can also see the rear support angle bracket on its axle again in this photo on the extreme right.
The original design suggested that the pen should be mounted inside the two collars at the far end of the parallelogram. Apart from needing a pen refill that is precisely 4mm in diameter (otherwise extra slop arrives in the mechanism), this seemed to me to be a complete dog's breakfast when changing pen colours. I don't want to have to say "I can't be bothered" when someone asks for a different colour -- and I've seen that at exhibitions before.
I sort of ignored the problem while redesigning the parallelogram, but eventually 'bodged' the solution you can see, which is two of the modern half-inch tyres (real Meccano!) which hold the pen inner hard against the far braced girder when hooked over the collars on the far axle. They are exactly the right size! Advantages are that the pen is held firmly but with a tiny bit of 'give', and that almost any size refill can be used and changed in seconds. It works superbly!
View from the front
The motor speed control is a complete cheat. I had a broken DeWalt cordless drill, an old 12V one. Screwdrivers, hammers, and nibbling away with pliers and knives released the central control for the drill -- a small box with a press button coming out of one side and a slider coming out another. The press button is of course what the trigger of the drill pushes against, and the slider is connected to the clockwise/off/anticlockwise control. This is a heavy duty 12V speed regulator of good quality, capable of dealing with plenty of current. It is mounted behind a bracket at the extreme front right hand side (behind the 2½"x2½" plate).
The controls are very simple. The push-button points upwards, and a cam sits on it, mounted on an axle with the yellow multi-purpose gear on it. Rotating the gear by hand rotates the cam, which pushes down on the speed controller button. About half a turn takes you from zero to full speed.
The left/off/right control has a strip hooked over it, which pokes out from behind the flat plate (supported by another strip behind it), and a threaded pin is in the end of this strip. As shown, it's in the 'off' position. Sliding the threaded pin slightly right turns the motor on, sliding it left turns it on in the opposite direction. Because this is a true 'off' switch, when the motor is off it is released, and so you can easily turn the Meccanograph by hand.
To the left you can see the usual things that make a difference between a model that's a pain to demonstrate and one that's fun. A heavily curved flexible plate holds spare pens, and there's a compartment to the right where a spanner and screwdriver fit exactly (the screwdriver pokes through the hold in the angle girder at the back).
Contrates swapped for bevels helps, as does swapping the drive in the right-hand gearbox to ensure (again) that the vertical shaft runs right through the model from top to bottom to avoid wobble. The only trickery needed in the right-hand gearbox is a swapped-boss plastic contrate (you can see it at the back) to enable this to fit, and be driven by the 25 tooth pinion far right which still keeps out of the way of the central vertical axle.
At the front, just behind the on/off switch, you will notice an axle running fore and aft with a small vertical axle poking up from it. This controls the secondary gearbox... we'll find that later. The lever is pulled back and forth to select the gears.
In this way, all control and fiddling and untidiness is done from this end, and the 'nice' side of the model draws the picture. I'm only letting you see this side because you're all friends...
This gearbox was a bit of a tight squeeze because it didn't exist at all in the original design. I wanted to be able to vary the density of the designs, by changing the overall ratio of the speed of the cams to the speed of the table. Note that the top bevel drives both the table rotation and translation, so it doesn't change the overall shape of the pattern.
In reality, it would be worth trying to get a three-speed gearbox into here (and I'm sure someone cleverer than I could do it). As well as the 2:1 and 1:1 speeds given here, 3:1 would be interesting, or even a little more.
The extra driveshaft and the pinion idler to the right are a result of trying to fit this gearbox in without changing the position of the cross-shaft (and hence the design of the main gearbox). It would be worthwhile biting the bullet and moving this to improve the drive throughout.
This was a fun Meccanograph to build, and one that demanded to be improved in many ways. It could probably be better (particularly in the area of the two main gearboxes, the first of which is a complete cop-out and the second of which is a bit of a bodge without enough gear ratios).
But overall, it's a small and simple model, very easy to carry about (it's always ready to go at the drop of a hat), not too hard to learn, and very quick and easy to produce good-looking patterns. It's not as complex as some, but has enough variety to keep you interested. It is utterly reliable, and produces standard six-way symmetrical patterns in about two minutes each. More symmetry costs proportionally more time, as does using the density gearbox (but I very rarely do that).
I would strongly recommend this model as a first Meccanograph, and it'll be one that you keep built for a fair while I'm sure. Mine isn't coming to bits anytime soon!
Charles Steadman, April 2010
potpot44 (at 5:34pm, Sun 7th Jun, 20)
très bonne article
Les Chatfield (at 9:03am, Sat 20th Oct, 12)
Would it be ok for me to build a copy of your meccanograph. I have watched your video many times and the more I see of it the more I want to build it. A very very neat piece of work.
Eric (at 9:19pm, Sun 22nd Aug, 10)
You forgot to show any drawings made by the machine. What do the drawings look like?