Pendulum weight
  • I've finished my Septimus, but it's not behaving well, and I hope you can offer some more hints.

    My first mistake was painting the workings. I tried to keep the (acrylic) paint off the touching faces, but a little spilled over, and it cost me quite a few hours removing little bits of paint off the faces and edges of all the teeth. There's no paint on any touching surfaces now, I believe.

    I've gone through the various tutorials and it all seems to run quite easily. It's easy to get runout causing parts to touch that shouldn't, and I've removed all of those.

    As a result, my Septimus runs - erratically. The anchor does not plunge into the escapement wheel properly and so as suggested in the build manual, I've gone through it several times looking for and hopefully reducing friction.

    It does all move when I blow on the escapement. I've done the "let it run without the anchor engaged" trick a few times, and the weight will run happily to the floor in a couple of minutes with all the cogs spinning. The pendulum runs for about 40-45 seconds with a normal weight on. If I add to the driving weight a little, the escapement just moves more quickly and hits the impulse face of the anchor more loudly!

    I've now found that if I increase the pendulum weight, it goes rather better. I've now got no less than 115g on the lower end (was 20g originally!) and it works much better - the anchor does go into the escapement wheel usually, although it does sometimes reduce its movement until it's back on the impulse faces, from which it doesn't revert to full swings. If I let the pendulum swing freely now, it runs for several minutes (I got bored watching it after 2 minutes!)

    Any suggestions please?

  • Let me start off by explaining what should be happening with your pendulum, as with my Septimus, and I stress it is built from one of my kits, 1kg drive weight and pendulum exactly as specified. When I stop the pendulum I can only get it to rest on the very tip of the impulse face, and the merest flick of the pendulum will start it swinging and within a couple of swings it is already plunging into the escapement wheel.

    Provided everything is assembled as it should be, there are 2 reasons why your pendulum does not swing as it should.

    Firstly, and probably the least likely, you have "over-sanded" the anchor and have changed the impulse faces, I say least likely as you would have to do some major sanding to have an affect.

    Secondly is friction, this could either be in your gear train or the pendulum arbor. To test this gently assist the turn of the intermediate wheel counterclockwise as shown below.
    Use only moderate pressure, as too much will cause the anchor locking faces to lock on the escapement wheel, if you can simulate the correct anchor motion by doing this, then your problem is friction.

    600 x 450 - 45K
  • My Septimus behaves just as Dave describes with the specified weights, but it took me several times of taking it apart and polishing the arbors, easing the holes and lightly sanding the gears until I got it running consistently even though I thought each time that I had got it as good as I could. I was beginning to tear my hair out as it would run for a while but then stop and at one stage would only run with the addition of a pair of pliers and a screwdriver (the nearest heavy things handy!) sprouting out of my bottle. It now runs fine, so persevere and you should win!

    I tracked down friction on another clock I made by using a plastic milk bottle as the weight and, starting with just the drive gear on its own, adding just enough water to keep the gear turning once I had given it a gentle nudge to overcome initial friction. I checked that the gear ran steadily like that. Then I added the next gear in the train, added just enough extra weight and did the same. And then I added the other gears one by one, each time increasing the weight just enough to keep the gears turning. I found this made it easier to trace any friction, either by noticing when the addition of a wheel needed a lot more weight (I had a 'sticky' arbor) or finding that only a bit more weight was needed but that the gears would occasionally stop ('sticky' teeth). It's a bit fiddly but I was able to trace friction that spinning the gears by hand or blowing on them didn't find - I guess because the force applied was more consistent.

    Good luck!

  • Thanks for both your comments. I did try pushing gently on the intermediate wheel as you described, that's why I thought I'd nothing to gain from trying to further reduce friction. Your Septimus, Dave, sounds like it's got a fair excess of energy available from the drive train to drive the pendulum, which is great, I'll aspire to that. I see yours was friction, petec, but as mine didn't respond to Dave's suggestion, I'm thinking that isn't the major problem with mine.

    I have sanded the impulse faces as part of my paint removal work, as I'd painted them thinking that would be OK. I'm wondering if I have thus altered the impulse faces a little. The energy from the escapement wheel has to be transferred efficiently to the pendulum. Can you please tell me the correct angle of the impulse faces? Do they have to be flat with sharp corners (seen from the front)? (I might have rounded mine). I'd rather know than experiment, else I might remove too much and have to build up the impulse faces again.

    Here's my clock on its test bed (aka the garage wall!)

    1526 x 4592 - 2M
  • Here's what mine looks like when the anchor is resting against the upper tooth. Hope that helps you compare angles.

    515 x 546 - 66K
  • To check your anchor, you could remove it from the clock and lay it over the full scale front view on the last page of the instructions. The impulse faces should be flat and sharp.

    I still tend towards friction and I think Pete´s experience is typical, after all how can you quantify friction, when I write "with as little friction as possible" I realise myself that it is a bit of a larry farry statement. One thing I sorely miss is a friction-o-meter, which pointed at a clock would tell me where the friction is, if anybody could invent one, well put me down for 10.

    Also check your pendulum arbor, can it move back and forth, and if, does it do so without any resistance or stickiness.

    By the way, I do like the look of your Septimus, the black frame and brass weight work well together.

  • Success! Well, several steps forward, anyway. It's been running for about half an hour now, plunging just enough to keep a consistent tick. Not quite right yet, as I've got a bit of extra weight on the drive weight and on the pendulum.

    The thing I did was to notice that the escapement was coming round "too quickly", and the point of the next tooth was so far round that it couldn't do anything but strike the impulse face. Hence adding to the driving weight didn't help, as the "door" was shut by the time the anchor tried to enter. Reason? I must have removed some wood from the impulse faces so that effectively the gap between the anchor tips and the escapement teeth was too great. Looking at petec's picture confirmed that the anchor teeth were a little short. So I glued on a small piece of 2mm thick ply onto one impulse face and reset the anchor, and it worked better. Adding the extra weights got it going.

    So now I think I'm into friction reduction again so that it does it without the extra weight.

    One of my other hobbies (I have lots) is model aircraft, and when we used to make rubber powered indoor models, we used to make tiny torque meters from bending small gauge piano wire. I reckon we could possibly devise a friction-o-meter by measuring the torque on the arbors. Or maybe just the force on the teeth of one wheel on each arbor - that might be easier. The range of force/torque would be quite large from the beginning to the end of the train, though, so it would be a challenge. And of course we'd need to know what it should be in order to detect a "fault".

    Thanks for your comment, Dave. I was trying to emphasise the different working parts and the 7. The brass weight was discovered when routing through my metal stocks (another hobby), and I thought it went well with the brass spacers and screw heads, too.

    Great! Quintus next!! It will take a bit longer though as this time I've got the wooden parts to make! Great learning experience, I now take my hat off even more for people who make watches! (As well as for those who design wooden clocks!)
  • Hello Stuart, I have been following your progress with interest. Well done! I am not sure i understand your friction - o - meter, but i don't need to. The information for 'what it should be' could be gained from a working version?
  • Hi Ian,
    Yes, indeed, we'd need to persuade Dave to loan us his ;-)
    It can go on my projects list, but alas it's quite long.
  • Well, it's been going for about a day now. I did lower the friction by several methods. I examined the teeth with a magnifier to ensure all paint was off, and then polished them all with a Dremel buffing wheel that I'd rounded on sandpaper. The arbors were polished (again!). I found my pendulum was not hanging vertically, so I packed the baseboard out from the wall a few mm at the bottom so that the pendulum is now parallel to the weight cord. I've still got a little extra weight on the pendulum, but the driving weight is the 1kg brass weight in the picture. I put a little vaseline into the arbor holes. I reduced the lengths of the brass spacing tubes about 1.5mm so that all my arbors now have only about 1mm end float, whereas when I built it they varied due to my inaccuracy.

    I'm new to clocks, so this might be obvious, but a lesson for me is to ensure the escapement wheel impacts the anchor on its end faces, not its impulse faces! I had to build my impulse faces up again (after carelessly removing paint) to get this to happen.

    I've just moved it into the house and set the correct time - we'll see how its timekeeping is!

    I've just ordered some 4mm 6-ply aircraft grade Birch ply from Dudley Pattison (Swindon Aircraft Supply Co) so await this before starting my Quintus. Nothing I had in my workshop was the same quality as Dave supplies, but I think this should be suitable. Alas Duds only sells it in 50" square sheets! That should make a few.....
  • At first it was losing about 40 minutes a day. However after a week of adjustments it's now keeping time to well under a minute a day. I've attached a couple of pictures of it.

    Interesting points - to make it go faster, the lower weight has to be moved UP! The opposite of what I did initially.

    Sensitivity of mine is a movement of 1mm on the weight causes a speed change of about 2 minutes a day.

    It's now got got a push like Dave's - very strong, and a mere slight push gets it started and up to its full swing in a few cycles. (It's now on 1kg weight and "normal" lower pendulum weight)

    The tick is so loud we can hear it in the bedroom upstairs! I might reduce the weight a bit.

    I'm now experimenting with a jig to make gears for my Quintus (I bought a hardware kit only) using a simple router. Alas NC routers are a tad expensive for hobbyists! I'll post some information if people are interested and I manage to make it work.

  • Like my stick? It's fabulous. It was made by a friend who wins competitions all over the UK for such sticks. It's keeping guard of my Septimus!
    Septimus and stick 2 small.JPG
    600 x 800 - 232K
    Septimus close small.JPG
    600 x 800 - 284K
    Septimus and stick small.JPG
    768 x 1024 - 389K
  • Nice stick - and the clock's not bad either! Glad you got it sorted. I'll be interested to hear how you get on with the Quintus.
  • Building a clock for the first time is a bit like learning to ride a bike, you can watch people riding, they can tell you how it should be done, you could even read about how to ride a bike, but until you put your bum on that saddle and push off it is all just theory, and after a few falls all of a sudden it comes together, without really knowing what you did differently.
    Thanks Stuart for sharing your experience and thank you also for not giving up until you had a ticking Septimus, the next one will be a lot easier, trust me:-)

  • Dear All

    In my place is difficulty to find the carbon rod. So...can i changed the carbon rod to another allumunium or the other material?

    Need suggestion☺

  • Let me start by saying why I went for carbon rod, it has a few advantages as a pendulum material, for a start is it very rigid, which is important as a flexible pendulum will end up absorbing too much energy. It is very stable allowing it to cope with changes in humidity as well as temperature, without hardly changing its dimensions. And lastly it has a very small surface area which means it is less influenced by pressure change/air movement.

    Basically any half way rigid material will do the job but the more it differs from carbon the less accurate the clock will become, also an increase/decrease in material weight will effect the pendulum length (the heavier the shorter, the lighter the longer)

    So to finally answer your question, you could take wood dowel, aluminium, brass, steel etc. Just be prepared to experiment as far as the pendulum length is concerned in order to get your clock accurate, best start off longer and to speed the clock up remove some length top and bottom.

  • Hi Dave

    Thank's a lot for your comment,
    I will tried soon,

  • Stuart, the plywood from Dudley Pattison (Swindon Aircraft Supply Co) is top quality and he will supply less than a full 50" square sheet but he does charge extra for that plus a few quid to cut into smaller pieces. I think, of the half sheets I ordered and received promptly, I'll get four Septimus clocks, which is good value. Will be starting on my clock in the next few days.