Clock Design 101- Why is it? and what if's...
  • While assembling the Sextus I'm repeatedly twisting the winding barrel for the purpose of motivating the gears while testing and looking for any elusive "tick stopper" gear unions. At this drive arbor, it takes over 6# to power the system. Be it grip or dead weight.

    My 101 design question is this:
    Why not configure the escapement arbor to share the drive function also? Since the escapement only requires mere grams to get the entire system to motivate, wouldn't it then be possible to run the clock with considerably less weight?

    I'm sure I may be missing the forest for the trees, and the answer will leave me all the more humbled, but I gotta ask.
    Would it be the speed of that arbor rotation is the issue? Creating a need to rewind multiple times in a day rather than the one time every 2.5 days?

    Thanks for your comments!

    - and do add any new subject additions to this discussion thread.

  • Ok, so a shot at the answer...
    I believe that the drive weight IS further down the train for the exact reason you describe, multiple rewinds. Since the Weight is generally further down the train, through gear ratios the amount of drop is limited in time through gear reduction. That allows for a longer run time between winding. If the final gear ratio is 12:1, its drop is slowed proportionally to that same rate. Imagine the barrel of the weight pulley having a circumference of 3 inches. That would equate to 3 inches of drop x the gearing its placed upon based on its rotation. That said, if placed in the hour hand gearing, it would equal 3 inches of drop/hr. Move it further down the train to the escapement. Utilizing the exact same barrel (3'' in Dia.) you now have 3 " of drop A MINUTE. So just in theory, a 40 hr clock ( 120" of drop utilizing a pulley or 2) now runs on the escape for around 40 minutes ( if my math is correct, and it may not be, its late :).As with anything, there is "Ying and Yang" to it all. This reduction and the massing of gears between the weight and the escape (with the addition of our dreaded enemy, "Friction") reduces the amount of force on the escapement to the mere grams as described. There are other ways to drive up the run time in a clock if the weight were near/on the escapement ( longer impulse times on the pendula to say 2 seconds, alteration of the gearing, additional pulleys to lengthen the amount of time till weight hits the deck) but they all have there challenges, and none of them afford the run times weight further down the train do.

    So ends my diatribe, which benefits all :)
  • Hey nice job Chris! Well, that all sounds good to me! : )

    So not so much which arbor that the weight is assigned to in the system, but more so the LOW gear ratio and diameter that is required.
    You know, winding every 2.5 days is pretty frequent attention! Heck, I've had plants that require much less attention that have perished on my "watch", so my challenges will not be complete even after a prolonged and proven "smooth run".

    My novice experience in this hobby leads me to thoughts of pulleys for that cure. What are some of the downfalls to implementing pulleys, other than making them ascetically pleasing and virtually invisible somehow? Four days would be more manageable with my current schedules.

    I'm just not savvy enough (yet) to ponder the practical details for your "longer impulse times on the pendula" option.

    Perhaps studying JamesP ( Andrew Cocker's) spring drive is next in the queue!?
  • Well, a ton to discuss then. We will leave out the plants. Think it's a different thread than here, but many of em on my side have been sacrificed due to "time" on my end too :)........
    Again, I'm fairly new to the field of horology and clock design, and Dave may have greater insight to add here, but here's a go as to the drawback to more pulleys (I may be incorrect in this assumption ). More pulleys reduce the driving force delivered to the escapement in the same manner as gears do in the train. I BELIEVE this reduction in drive force is less due to the pulleys VS gears, but it's still a reduction. That said, it would mean an increase in drive weight in proportion to the loss of drive force to the escapement. It's my understanding ( I could be wrong ) that the goal is to run a clock with the minimum amount of weight to reduce strain and wear on the movement. That said, more pulleys =more weight = more strain /wearing on the train. With all that outta the way, there are other methods for extended run times on clocks. 8 day movements are readily available for many types of clocks, and I do believe one of Daves designs ( ?the nonus?) is of this variety. They are not a bolt - on for our clock, but looking them up may show you how they work, and you could ad lib from there. Lastly, and unfortunately for the purpose of this discussion, an 8 day clock movement is usually incorporated in the original design via changes in the impulse to the escapement, gearing ECT, so this isn't something usually done "post design ". That doesn't mean it can't be done, it just means it requires some good old ingenuity and outta the box thinking ;)
  • I'm so far from ad libbing on most anything that's nuts n' bolts-clocky Chris, it's silly to even insinuate that I might!! LOL!

    So anyone have "the jewels enough" to state a position here on the lesser of two (or more) evils in the best design of extending a clocks' run time between winding's?
    (Short of adding an electrical cord, etc, etc.!) Ha ha!
  • The best solution: leave the clock alone and, like Huckleberry Finn, get someone to wind it for you.
  • The easiest way to double the run time is with a pulley, see below schematic, it really is that simple.


    image

    It adds no further strain to the winding barrel as the load is equally distributed between both barrel and hanging point,but with the Sextus do not add the hanging point to the frame as it is not designed to take nearly 6 Kgs.
    You could use a hook on the wall or hang from the ceiling, with a skinny string it will hardly be noticeable.

    Although I do like the Huckleberry Finn variation.
    pulleysystem.jpg
    363 x 394 - 33K
  • GREAT replies guys.
    Purposefull and most entertaining!
    Sourcing a pull this week. (To exibit my confidence and determination to own a running Sextus.)
  • So to conclude, if you incorporate a pulley design you must multiply the weight by two (for a single pulley) totaling 13 lbs. for Sextus!
    OMG, that's an awesome amount of dead weight.

    I'm still in pursuit of a brass/bronze foundry that will make me the "climbing squirrel" weight and may have found one! Since the statue will be hollow, I'll be able to tweak the weight up as needed with lead shot.
  • Each of Sextus pendula currently will free swing independently for over 55 seconds.
    https://youtu.be/yyysYDU9grA
    https://youtu.be/q0uyH5eUODU

    Together they free swing for over a minute.
    https://youtu.be/MAXsm6C4Ikc

    The blow test confirms there is very little friction effects as a unit.

    So far the longest duration run with 3033 gr. weight so far has been for 75 minutes.
    4+ video minutes of a weighted run:
    https://youtu.be/4xOjivJ5sSo

    My next question is one I'm certain is so predictably typical of this "hobby" ;
    Why does it run 75 minutes only to stop ???

    Why would it run for this extended time only to eventually yield to the dark forces of some truly minute friction and or Murphy's law? Every bearing and arbor moves with pure fluidity. All gear teeth have been honed and manicured like one's visa depended on it! Why does it stop when so much is right?

    Why 75 minutes and not just two, or six?

    The closest of scrutiny of the pallets' to gears delicate ballet of connecting and releasing in the nick of time to gain that needed nudge appears spot on. And yet the pallets escape that union before confounding the gain the escapement's energy has to offer. Why stop? Maybe the left pallet's sliding into the cavity of the escapement fear is a probable cause? ( @ 3:09 )

    Why do I get 75 minutes? When I first started this "tinkering" process the Why's were at first easy to sort out. But now I'm finding myself at a loss of where to go next in this goal of easing the friction (?) or unknown yet to be discovered.

    Your suggestions / observations encouraged.
  • I wish I could help mark, but as stated in the international shipping thread I just received my kit, and have no experience with a grasshopper escape as of yet. Just spit balling an idea though.....in the 75 minutes it did run, how well did it keep time? Was it slow? Was it fast? Reason I even mention it is according to what I've read, the sextus timing is fully based on its drive weight and pendulum adjustment. If it ran very slow, it could mean a upward adjustment in drive weight or........

    Our evil enemy. ....

    Friction.

    Bear in mind how delicate the "balance" is when dealing with these movements. Grams.....just grams....on an escape wheel. That's it, and all it takes to not function. Keeping that in mind, and the fact that it looks to spin freely on a bench doesn't necessarily corilate to vertical hanging. Has there been some frame sag? Is the wind cord bound on itself? These could be potential intermittent stoppage causes to a clock also.

    One last thing to mention that someone once told me that applies to anything mechanical that fails to function per it's design. It helps me everytime in this troubleshooting phase.....

    " The catastrophic failure of anything mechanical is rarely one single event. Is is most often attributed to a series of circumstances that align perfectly that lead to its demise."

    All that means is that it's usually something that one would consider anomalous for its one function alone aligning with another causing the item to fail. Break that down to its smallest simplistic component and the reason will become so blatantly apparent to you you will wonder how ya didn't catch it in the first place :)


    And seriously, after a 75 minute run, she is so close to having a permanent heartbeat, the problem is minor, therefore finding it will definitely be more challenging, bit in the end, more rewarding :)
  • Finding out what stops a clock from running can be quite a challenge. For what it's worth, here are a few things I do to try and find the problems, and bear in mind that I do have lots of problems when I cut on my scroll saw as I'm not all that accurate!

    I firstly check that all arbors are running freely in the frame. The easiest place to start!

    Then I let the clock run until it stops and lightly mark with pencil all the teeth that are engaged with each other when it stops. Then I wind the clock back a few minutes (by disengaging the escapement if I can) and run it again and see if it stops in the same place. If it keeps running, I let it carry on until it stops again and check the marks. If I'm lucky, one pair of teeth that I previously marked may be coinciding again which points to them causing the problem. I also mark up any new teeth that are engaged (my gears can get covered in marks!). Eventually I can have loads of places where the clock stops. Any time it stops where it has stopped before, I check how those teeth are engaging with each other. It can take a surprisingly small amount of resistance to stop the clock.

    Other things I try are to remove whatever prevents the escapement wheel from turning (the anchor(??) or 'fingers' of a grasshopper, depending on the clock) so that all the gears can turn freely. I use a small milk bottle as the drive weight and add just enough water to get the gears turning. Then I let it run until the gear holding the weight has made at least one complete turn and watch a listen carefully for any places that slow down or stop.

    The reason that I let it run for at least one turn of the weight wheel is to make sure that all teeth that are going to engage have engaged. Depending on the gearing, you may have to run it for longer (if I remember correctly, my Septimus drive gear has 60 teeth and the gear it meshes with has 18, so every turn of the drive wheel gives 3 and 1/3 turns of the smaller gear and not until three turns of the drive gear are things back to where you started) . That could be the reason you can get 75 minutes but no more. To begin with it seems illogical that it should run for so long and then stop, but until all the gears have been though all their rotations you haven't been through all the possible teeth meshings.

    Another thing I do is to put each meshing pair of gears into the clock frame on their own so that it's just those two gears engaging. Then I slowly turn by hand the gear that will be driving in the direction it will be turning whilst using my other hand to put resistance on the gear that is being driven. I find it surprising how many little lumpy engagements that I can feel through my fingers that don't show up any other way.

    Oh, and I make sure that I mark my gears so that I always put them back in meshing in the same places! If I'm spending hours fettling away so that things mesh smoothly, I don't want to start again each time by putting them together differently! It seems obvious now, but took me quite a time to realise that it was helpful.

    I hope some of that might be of some help. I expect you'll get there eventually. It just takes lots of patience. Good luck!
  • Ok these are very helpful suggestions! THANKS! I'll change my weight set up to the 6# dumbbell and a canister for adding sand ballast to fine tweak the balance of weight.

    Marking the gears is also an idea I'll work out somehow since they are painted. I agree eventually a pattern should reveal the problem connection(s).
    Last couple runs (2x) have been for 1 hr 38 mins. I went ahead and glued the frame joints to assist in preventing any joint related sagging. BTW, the clock is mounted on a wall for all these previous trials.
  • DeterminING if the runs are slow or fast will require constant review as the stopping point is well mostly unpredictable.
  • Timing is kind of irrelevant until a consistent run can be achieved ( just a statement of fact, no I'll will intended ), so I would focus on the reasons it stops before focusing on its timing. One small step for the sextus, one giant leap towards the "heartbeat ". :) one quick tip I learned from a master clock maker-once she runs without fiddling, keep a journal of its timing nearby and attempt to check it at the same time every day, and wind it then. Note whether it is running fast or slow, and if you wish, the weather that day. Tracking this over the course of a couple of weeks will show you the tendency of its action and how humidity, temp and air pressure affect it's action. This will give you a direction to take to keep her closer in time. That said, keep in mind, if you want perfect time, a digital clock will do. If you want to marvel at your creation, this hobby will be addictive.



    Good luck and march on! ;)
  • After marking the gear connections @ the last stop, I proceed to reverse the clock back to the 12:00 starting point by slowly reversing the drive gear.

    I noticed the hour gears seemed sticky so I removed that assembly from the clock at 9:15pm. This am @ 5:45 it was still running (minute hand only). This is the longest run time to date.

    Can't wait to fiddle with the hour drive in hopes this may be the last obstacle to achieving the 24/2.5 run!? Whoo hoo!
  • Then it will be on to the pully addition with a customized bronze climbing squirrel 13+ lb. weight statue.
  • the sextus has been the least frustrating clock I made. it only stopped once when it it was first up and running and needed the pallet stops adjusted since then it has run. though with mine adding weight didn't to speed up the clock enough and just caused the pendulum to hit the frame. I reduced the weight to 6# and dropped the pendulum weights till it ran at speed. Here is how I do the weight peg. just glue a 2" screw that fits the wall faster into a wooden peg and screw it in. it lets you adjust the placement of the string so you can control how far away from the wall it is. plus using a walnut dowel looks cool.
    IMG_1466.JPG
    4032 x 3024 - 2M