Gina

CAD model of the getready/start striking control shown coming up to the hour and just about to start the striking sequence. In this position the striking is held stopped and the rack pawl pushed away from the rack. Next the triangle will drop on the minute shaft cam (bottom right) and stop and pawl released, letting the striking start.

I have a green observatory and though it warms up inside with the roof closed, it soon cools down again once the roof is open. I haven't found a problem with overheating.

Reversed the main gear train to accommodate the striking mechanism on the left. To make use of gravity on the pawls the auto-winding motor will now go in the middle. This has one advantage that it can auto-wind both the main clock drive and the striking mechanism, if I can work out the gear ratio suitably.

Just before the hour, a cam on the minutes shaft applies a stop to the striking main shaft and then the rack pawl is lifted releasing the rack which falls until the probe hits the large snail cam. On the hour the cam on the minutes shaft releases the pawl and striking shaft stop and the striking mechanism springs into motion. With each revolution of the striking shaft the rack is raised one notch and the gong or bell is struck by a hammer. As mentioned above, striking continues until stopped by the lug on the rack engaging with the small snail cam. The process then repeats around the next hour.

This diagram shows the dimensions of the 3 shafts relevant to the getready/start which is controlled from the minutes shaft.

Rack pawl added.

These diagrams show the rack and small snail cam that lifts it plus how the lug on the end of the rack stops the action when the required number of strikes has be achieved. The small snail cam shown in the second diagram rotates until it contacts the lug on the end of the rack and striking stops.

Another attempt - this time for the RHS.

However hard I try, including some "thinking outside the box" there seems to be no way to have the striking mechanism on the right and that includes the rack lever with peg, without complicating it (more levers and pivots). The probe (whatever sort) cannot clear the 1 o'clock part of the cam from the 12 o'clock position. I even tried cutting back the cam but then the 1 o'clock wouldn't work. Oh well, I've had a lot more CAD practice! ?

First diagram shows the small snail cam that drives the rack upwards and pawl that stops it from dropping back when the cam disengages from the rack. Second diagram shows the rack in the 1 o'clock position.

In working out the design of the rack lever I found I could have the lever in the same plane as the large snail cam which would make things easier. Bit rough ATM but I can tidy it up easily enough.

I like both the green ones.

Constructing model for rack lever and small snail cam. This is looking a possibility. Although the small snail cam looks close to the frame the centre is 30mm away so a 50mm OD gear would fit in easily.

Realised I was thinking in two dimensions but we have three dimensions to play with and the mechanism doesn't have to lie in one plane. I have in mind a new design that will work with the striking mechanism mainly to the right as I had originally intended. The lever with the rack on the end will have a peg which will drop onto the large snail cam, so the lever/probe clears the cam by being in front or behind it. This diagram shows the idea with the peg represented by small circles on the lines that represent the positions of the rack lever. The large snail cam no longer needs a curved face between 12 o'clock and 1 o'clock. The line on the right represents the edge of the frame.

I've found another couple of problems with this design. The curve on the large snail cam between the 12 level and the 1 level fits a design where the probe enters from the top with the pivot on the right (or from the bottom if the pivot is on the left). Plus the whole thing is too big to fit withing the clock frame. These two problems are not simple to rectify. One solution would have been to reverse the rack lever and put the pivot on the end and the probe part way along but this would mean having the striking mechanism on the left hand side of the clock but my design so far has the clock gear chain that side. It isn't possible to do a mirror image of the striking mechanism as the large snail cam has to rotate clockwise (unless I used a pair of gears to reverse it - which is daft). Two possibilities occur to me - swap the whole clock round with the main gear train on the right and strike on the left (which is actually how most clocks are arranged I have since found out) or redesign the striking mechanism so that the probe part and large snail cam actually work together. My way of thinking is that I work from left to right as with writing and to me the main clock gear train comes before the striking mechanism. That is why I designed the clock that way round. Clocks usually dont't show the works so it doesn't matter esthetically but mine will show off all the workings as a main design principle.

Here we go... First diagram shows shows the rack plus its lever and probe with the small snail cam in resting and starting positions. Resting and stopping positions being the same. The second shows the stopped position of the rack part as well.

Hold on a minute - this isn't right!! That's not 10 o'clock - it's 3 o'clock! Thought that rack position looked a bit odd... OK... Reworking it...

This new diagram shows one method of stopping the triking action. After the 10th strike the small snail cam turns until it hits the end part of the rack.

Since the large snail cam (on the hour shaft) is in front of the front clock plate, it would seem to make sense to have striking start cam (on minutes shaft) in front of the clock plate. This means the hour drive gear can be behind the front clock plate. It all depends on how the striking mechanism works out.

Here is a diagram showing how the striking is controlled to give the correct number of strikes according to the hour. The 10 o'clock position is shown with the striking just starting. I haven't shown the parts that control the starting and stopping. The large snail cam on the left is on the hour shaft/tube and the small snail on the right is on the strike main shaft and driven by the striking gear train. The bell is struck with each revolution of the small snail and the mechanism is stopped when the ratchet reaches its highest position. Both cams rotate clockwise. A latch pawl (not shown) holds the rack as the small snail clear it. This pawl is released just before the hour and drops the rack, raising the probe until it reached the large snail cam (as shown).

It isn't possible to finish the main clock mechanism without looking at the strike mechanism and the strike needs a cam on both minute and hour shafts. Another cam is needed on the hour shaft (tube) to drive the moon globe dial. This means the hour tube needs a gear (to drive it) and two cams. The minute shaft will have a coil spring (to press the great wheel against the clutch plate), great wheel, clutch plate and cam that starts the strike mechanism plus gear to drive the hours. The clutch plate can be combined with either the strike start cam or the gear for the hours drive. With the present design, there isn't room for both behind the front acrylic plate. I shall now look at the strike mechanism to see which arrangement would be best.

As a little digression, I shall look at the calculation of the drive system for the strike mechanism allowing for the possibility of no winding for 12 hours as a starting point. I shall deal with auto-winding later. The amount of motion over the twelve hour period may be calculated from :- Total number of strikes for 12 hours will be 1+2+3+4+...+10+11+12 = 78. Using the same chain and weight system as the main clock drive gives 3m as the range of chain motion over the drive sprocket. I have an 8 tooth sprocket with diameter approximately 50mm giving a circumference of 50xπ = 157mm Number of revolutions of drive sprocket before weight hits the floor = 3000/157 = 19 Each strike can be allocated 19/78 = 0.24 approx. or just under a quarter of a revolution of the drive sprocket. To allow some tolerance we will allocate 5 strikes per revolution of the drive sprocket, giving a 5:1 gear ratio between sprocket and main strike shaft. To prevent the strike mechanism from running too fast I shall use the standard fly-fan governor with a further gear train from strike shaft to fly-fan shaft. Some experimentation will be required to find the gear ratio required as a calculation would involve the amount of air resistance and the torque driving the fly-fan (which is beyond the capability of my poor old brain!). A larger fan will run more slowly for a given torque. I can use the fly-fan I used in the Mk1 clock for the perpetual calendar mechanism.

Further progress - more gears printed plus drilled holes and added bearings.

Great stuff James

I've been burning up rubbish in my garden taking advantage of the nice weather and very light breeze.