Escapement friction
  • Hi All,

    I have built my Septimus being very aware of the friction issues, however, now that I have completed it, it just won't "go".
    I have read a lot of the previous threads on here and tried the various suggestions so here is a bit more detail:
    If I remove the escapement anchor and place a 1kg weight on the system, it runs through really well - no problems at all it seems. With the escapement anchor in place, it only strikes once or twice before coming to a complete stop, although the pendulum swings quite easily when the anchor is moved out of alignment with the escapement wheel. On one of the previous threads some one suggested trying to manually move the intermediate wheel (I think I have the right terminology), and I found this to be very clunky and just caused the escapement anchor to "catch" the teeth of the escapement wheel. Does anyone have any suggestions?

    Another point I should mention is the fact that when I dismantled the escapement assembly just recently (looking for friction) I noticed that one of the bearing wheels might be damaged - if you continue to rotate it all the way round, every now and again it catches and needs quite a push to get it going round again. This was probably caused when I mounted it into the front of the Septimus 7 as it was tight and so I had to tap it into position. I am not sure that this would actually cause an issue as this arbor only sweeps a small arc and doesn't continually rotate around like all the other arbors. If this is a problem, does anyone know where I can source some replacement bearings?

  • There are basically only 2 reasons why a mechanical clock should not work, the first is the drive force not being sufficient enough to power the escapement and second the escapement itself.
    The whole point of pushing the intermediate cog is to check whether the escapement works as it should. You mention that when you push the escapement catches, that is actually a good sign but shows that you are using too much force, the intermediate cog should be assisted with a minimum of force, just enough to get the kind of anchor swing shown in the Septimus video .
    As you can see on the video the anchor dips so far into the escapement wheel that the teeth hit the locking faces of the anchor, the anchor slides off the tooth and then connects with the impulse face pushing the escapement.

    The connecting phase can be split up into four parts.

    1) At the start of the cycle, the anchor locking face engages with the escapement wheel. This action stops the gear train, hence locking face.

    2) The momentum of the pendulum pushes the anchor deeper into the escapement wheel, once it has reached its limit it will shortly stop.

    3) Now the weight of the pendulum will pull the anchor out of the escapement wheel.

    4) Once the anchor has swung far enough the escapement tooth will move from the locking face to the impulse face, this is the stage where the gear train starts moving again and there must be enough power transmitted to the impulse face for the other side of the anchor to engage in the same way as described, which of course repeats the cycle.

    So as you can see too much force applied to the intermediate cog will just cause the escapement wheel to rub against the anchor locking face with enough force to stop it.
    And also you can see that too little transmitted force to the impulse face will prevent your clock from ticking.

    Testing your clock by letting it free run is just going to show you that you are moving in the right direction, but as it is freewheeling the momentum of the gear train will just compensate for any friction points. The Septimus stops totally once every second and then has to accelerate from nothing to enough speed/force to impulse the escapement, this is where too much friction will be noticed.

    There are of course a few "standard" culprits than can cause your clock not to work as it should, and those I have covered in my various tutorials, what I can not predict are the mistakes that can happen when the plans and or instructions are, for whatever reason, not followed.

    As an example I have had situations where the anchor has been mounted back to front, obviously with no chance of working. One of the best was a Septimus that's escapement wheel was directly connected to the minute cog instead of the intermediate cog, I was not even aware that that was possible, the result was an escapement wheel that turned in the wrong direction.

    384 x 483 - 83K
  • Would bearings from here be a suitable replacement:
  • Hi Dave,
    Thank you for the very detailed response - it has given me a much better understanding of how the mechanism works now. It also gives me loads to investigate and I will have another look tonight. Immediately, I am thinking that I may not have sanded the escapement teeth and anchor faces enough (I didn't want to over sand and change the geometry) and so there is probably too much friction there.
  • Hi Dave,
    I have gone back and sanded all of the teeth on all the wheels but especially the escapement wheel and the anchor faces. This definitely seems to have improved things but I have also put a bit more weight on and now the whole system ticks for a good 4-5 ticks. Now I think there is too much friction in the actual pendulum mechanism, possibly the bearings but also closing the front plate on seems to be rather tight, and so I am going to take a closer look at that area and check the overall length of the pendulum arbor.
    I put a bit more weight on the drive and it seems to go for longer - I assume this is just still indicating too much friction in the system.
  • I have made a short video showing some of the things to check, I hope it helps further.

    Septimus Help Video
  • hello all!

    I'm in the final stages of my Septimus assembly and the elusive tick still remains, well, elusive. Following all of the advice on these forums, I think I've isolated the issue to either be the pendulum or friction in the escapement wheel. The drive train runs very smoothly and accelerates fast when not connected to the pendulum. The intermediate arbor has a ~1.5mm wobble to it.When I try to drive the clock using my thumb (like in the above help video), I can get it to run, but the clock is very sensitive to pressure.

    Regarding the pendulum, it swings very freely but with very little force. Pulling the pendulum all the way to the frame, it swings roughly 30 seconds before stopping. In the Septimus help video, Dave mentions his ran for 60 seconds but that times may vary. Is my run time OK? If it's not ok, any advice on how to improve this? So far, I've added a bevel to the arbor, polished the ends, and made sure the arbor has remove to move while connected to the ball bearings.

    Given that the pendulum swing seems weak (I think), the connection to the escapement wheel doesn't really work. Using Dave's above drawing as a reference, my clock only hits positions 3 and 4. The anchor barely moves up the escapement wheel (position 2). I maybe get 4-5 ticks before it grinds to a halt. Typically, the clock stops with the anchor on the very tip of the escapement wheel (i.e. the moment before progressing from #4 to #1 on the other side). I've tried adding more weight to the drive (0.75kg -> 1kg -> 1.5kg -> 2kg) to no avail.

    I compared the anchor to the full scale drawing in the directions and the bottom point is slightly rounded off. I don't think it's enough to impact performance, but I'm not sure.

    To top it all off, one night, the clock ran for 8 hours straight. Any help would be appreciated.

    1224 x 1632 - 620K
  • If your clock has run continuously for 8 hours then you can start ruling out a lot of factors straight away.

    Firstly your escapement obviously works as it should.
    secondly all parts are in the correct place and orientation.
    thirdly there are no forgotten tabs on a cog tooth.
    forth the drive weight seems to be sufficient.

    So what do you have left ? The answer of course is friction.
    Never forget a clock stops and goes once every second and never develops momentum, as opposed to just letting the gear train run freely.

    When a deadbeat escapement stops because of a lack of drive force it will always be on or close to the impulse face tip, so do not worry absolutely typical and absolutely solvable.

    As mentioned previously I have handled friction reduction in so many tutorials that I think I do not need to roll it out here, but as you mentioned that you think your pendulum might be the problem take a look here BALL BEARINGS

    It refers to the Sextus but the ball bearing principles are the same.
  • Dave, The ball bearings video did the trick! After a quick sanding / polishing of the escapement arbor, the clock ticks!

    Now on to the fun part of trying to get the clock to keep the correct time.

  • hello, do you think is working well the escapement? Unfortunately, after about 20 seconds it stops
  • My clock has been running for almost a day. It made me happy for a while. I thought it was successful, but it runs fast almost doubled. After looking at the suggestion of Dave, it was found that the operation of the escapement wheel was totally wrong, and anchor could not be pushed into the escapement wheel. It only ran 3 steps and 4 steps in 4 steps. I increase the swing of the pendulum, so that it can run all 4 steps, but it will soon decrease, and can not be pushed into the escapement wheel or stoped when pushed into the escapement wheel. I have been adjusting for a long time, and I have no success. I polished the wheel, and no matter how smooth it was, there was always some friction between the wood. I really have no way, dose anyone can help me?