A question of tooth design with regard to suitability and manufacturing methods

Discussion in 'Clock Construction' started by Zahnrad Kopf, Mar 2, 2014.

  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  1. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Warning - Newbie posting

    Greetings. This is my first post here and it comes as a question of methods and suitability.

    I am a newbie to the world of watch making and clock making. I have it in my head to manufacture my own pocket watch and was directed here by others more familiar with the processes after asking for some direction.

    I've been reading and researching, but have the same question that most newbies seem to have -

    Why are Cycloidals so much more preferable to Involutes?

    If this was 50 years ago ( and earlier ) and I a bit more ignorant I would understand and not bother with the thought. But it isn't and I'm just slightly better than completely ignorant about toothed wheels and their creation and use ...

    The two common reasons I have found for mandating Cycloidal tooth forms is friction and strength. This got me to considering the issue, the applications of the pinions and wheels, the way they are manufactured, as well as the materials they are manufactured from.

    Given today's available methods and materials, just exactly why can't Involutes be used in the train?

    The issue of friction seems to come from two points - The first is that clocks and watches are able to supply relatively very little power and the Cycloidal form leaves ( no pun intended ) less engagement ( most notably during the approach action ), thus reducing friction. The second deals specifically with the considerations of the specific method of manufacture and that "leaves" are oft left needing polishing to remove the marks and imperfections of manufacture. I have to admit to being left shaking my head at the concept that this is so blithely accepted without questioning...

    The issue of strength seems to typically refer to that of small pinions and the fact that in such small number of tooth count pinions there is noticeable undercutting, leading to loss of strength and possible breakage.

    So, in considering these issues I have to admit to not being completely convinced. As regards friction, we have ( in general ) much more sophisticated methods of manufacture available to us these days and these give much better results with regard to surface finish ( and accuracy ). This doesn't even need to touch upon the topic of being able to diverge from the mathematical "standards" of gear design and utilize modified tooth profiles to achieve very low frictions between pinions and wheels.

    Where it's a question strength one has many more choices of available suitable materials to take advantage of and thus render the consideration moot in my humble opinion.

    So, to the point - One has available to them all the tools and experience to easily make any tooth form they should have the whimsy to desire. And with today's methods and materials, one can easily surpass the strengths and end results desired of decades ago. ( and one might even argue - the power available to drive said trains increases with these same considerations )

    EXAMPLE - It would be a trivial matter for me to create a pinion or wheel of Stainless Steel, .015" ( .381 mm ) thick, with any tooth form desired, and having a surface finish comparable to that of something polished with a stone. And contrary to days gone by, this could be done easily enough from drawing to finished product within the span of a few hours.

    So...

    Are Cycloidals really still so superior to Involutes with regard to watch and clock manufacture these days?

    Obviously, I have it in my head that Involutes might be used exclusively in the creation of a watch or clock movement, but as an obvious and admitted newbie I want to hear from others more knowledgeable and with more experience than I.

    Thanks.
     
  2. David S

    David S Registered User
    NAWCC Member

    Dec 18, 2011
    7,201
    246
    63
    Male
    Professional Engineer - Retired
    Brockville, On Canada
    Country Flag:
    Region Flag:
    Very interesting thread. I am a clock guy and will be interested in how this goes. One thing I have been told is that the gear trains in clocks and watches operate "backwards" from typical gearing applications in that with time pieces we have big gears driving smaller pinions. Most of my experience with involutes has been with power tools where the application is gear reduction. So I wonder if cycloidals are more efficient when larger gears drive smaller ones?
     
  3. Jim DuBois

    Jim DuBois Registered User
    Gibbs Literary Award NAWCC Member Sponsor

    Jun 14, 2008
    2,934
    635
    113
    Male
    Magnolia, TX
    Country Flag:
    Region Flag:
    #3 Jim DuBois, Mar 2, 2014
    Last edited: Mar 2, 2014
    You might review the attached papers for some differing points of view. IMO there have been many millions of clocks successfully built and ran a lifetime with Cycloidal gearing, a few have run successfully with involute. Involute works, but it usually requires more power, tighter tolerances, is less tolerant of dirt and dust,.....on and on.....if you review other threads in this segment of the forum you may find some others who have built using involute and you may notice some of their issues...also, in the attached paper regarding the advocacy of involute the author cites multiple failures, but someone else is always to blame, not the involute gearing itself...and ultimately he does state the failures are due to not meeting tolerance requirements.....if the best the government can buy doesn't execute successfully then that alone give me pause...
     

    Attached Files:

  4. cmnewcomer

    cmnewcomer Registered User

    Mar 24, 2009
    48
    1
    8
    I cannot speak to the theory of it all and there are many threads I have read over the years here that suggest the involute profile is not suited for clock making but looking at first hand empirical evidence, I have found that the involute profile works very well for clock making. Confused by the posts I read many years ago on this subject, I decided to make two identical clocks with the exception of the tooth profile.

    The first clock I made used 0.8 module epicycloidal gear cutters I purchased from P.P. Thornton. The second clock I made used 0.8 module involute gear cutters from Russell, Holbrook & Henderson. I cut both the gears and pinions. Both clocks are setup with 12lbs of weight on the time and strike train and keep relatively identical time. I say relatively identical since I would need scientific measuring equipment to identify the difference.

    This was of interest to me also since my Grandfather made 104 clocks in his lifetime. After repairing clocks and watches for many years, he decided to make clocks and used the involute profile for all of his clocks. I have Clock #20 built in 1968 and it's running strong with no visual signs of wear on the pivots, gears, or pinions. I know of over a dozen clocks still in the family that are also still running strong including Clock #1 which was built in 1964. While 34 years is relatively young for a clock, I anticipate this clocks will run strong beyond my lifetime.

    Also, it is my understanding that the epicycloidal profile is actually weaker than the involute profile since the epicycloidal profile has straight sides near the root and the bottoms are square in comparison to the involute profile.

    Best Regards.

    Carl
     
  5. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    From Jim Dubois' attached references -
    Thoen writes, "Nowadays, although experimental evidence shows involute gearing to be more efficient than cycloidal gearing, conventional theory still indicates just the opposite."


    David, thanks for posting. My own experiences with toothed wheels have been primarily in the area of Involutes, with a smattering of other, less traditional tooth forms along the years. The only difference in the transmission of torque with regard to the direction of amplification/reduction would be in regard to pinion strength when very low tooth tooth number gears are driven by large numbered gears in that the torques are increased/decreased by the same ratio as the tooth counts along individual teeth. Given the levels encountered in such mechanisms as we are discussing, and compounded with the techniques and materials available to us today, I have a very hard time blindly accepting that it even comes into play enough to actually matter.


    Jim, thank you for posting those materials in the thread. While they are admittedly in agreement with my own thoughts, I would like to see actual evidence of both points of view presented here in effort for us to have a meaningful discussion of the subject.

    A few points, if I may -
    This goes to the heart of my quest, I think. I have yet seen nothing that addresses this in a logical, scientific manner and still states as much. In fact, quite the opposite. The few papers or studies that I have come across seem to indicate quite the opposite.

    Here is just one that I happen to have handy - http://proceedings.asmedigitalcollection.asme.org/proceeding.aspx?articleid=1572229. Of which, the abstract reads - "In the previous study, we have developed two non-involute spur gears that have involute-cycloid composite tooth profile and modified cycloid tooth profile, and proved that these gears have greater bending strength and surface durability than the involute spur gear. In this study, we measured the friction losses of involute-cycloid composite tooth profile gear, modified cycloid tooth profile gear and involute gear using a power circulating-type gear testing machine, by the oil immersion formula, whose method is based on the oil temperature rise due to the power loss in the gear drives. It was found that the involute-cycloid composite tooth profile gear has the least friction loss in the three gears. Furthermore, we calculated the gear friction losses of these gears. The calculated value agreed with the experimental data."

    ( I've seen others, as well, but do not have them at had at the moment to cite from. )

    As well, my own experiences in Gear Making over the years leads me to be very skeptical of statements that Involutes require more power and tighter tolerances. You'll pardon me if I liken it to the FUD that Microsoft used to spread about Unix/Linux years ago, establishing their arguments and counters by sowing Fear, Uncertainty, and Doubt, but never really giving solid examples or reasoning.

    Too, at a certain point even the question of tolerance becomes questionable. How much tolerance could the Watch & Clock Makers of yore adhere to on demandable, repeatable basis? As so many point out so often - the practice most commonly adhered to is that of optical comparison. These days it is not uncommon in typical manufacturing settings to adhere to much more stringent standards and tolerances, with a few tenththousandths of an inch ( .005 mm ) not being an uncommon tolerance for other parts to be made to.

    To address the last ( that being the tolerance of dirt and dust ) I simply do not see it. The common point is that the Cylcoidal form offers a place for these materials to fall into.

    But... ... ... just how does this dust and dirt know which way is the correct pathway to the root... ... ? See what I mean? What about a watch that is always in motion about one's person? Now, the dust and dirt must not only know which direction the root is in, but be able to keep track of such a moving target that is constantly changing its orientation...

    And please pardon the digression, but what about the case of a watch ( no pun intended ) that is waterproof? We are to expect that water cannot penetrate, but that dust and dirt are to be a concern?

    Please, make mo mistake - I am completely open and willing to use whatever form is to be the better one for my endeavor. It literally makes zero difference to me. I can make any of them. But I want to know that I am making whatever I am because it is the best I can use/achieve for the given mechanism I desire to make. And I wish to do so based in solid factual comparison and not just a time honored superstition. :)

    Carl, Can you tell me if you have ever measured the time that each would take to require winding? What about long term accuracy?

    To be fair, I think strength is of trivial concern here, in my own endeavor. I just cannot see how modern materials and processes could possibly result in something that would be of any more concern than those produced decades ( or scores of them ) ago. I think the flanks of a modern gear tooth ( no matter what the form ) with resulting finish similar to that of an older example with hand polished flanks, and made to tolerances easily adhered to in the single thousandths of an inch ( .0254 mm ) or better will be just fine over the long run.

    So ( if this is to be accepted and believed ) that leaves the question to be one of friction and power required.

    Thank you all for answering. I am trying to learn as much as possible along my journey toward making my own time piece and appreciate anything I might gain from discussing it with you. Even if it conflicts with what I might think I might know. :)
     
  6. cmnewcomer

    cmnewcomer Registered User

    Mar 24, 2009
    48
    1
    8
    Not sure I understand the question in regards to winding the clock. They are both 8 day longcase clocks with the same gear tooth count and the winding barrels are the same diameter so they both run the same length of time between windings.

    Clock #20 that I own runs more accurately than my Rolex watch. While I have never taken the time to try and see how accurate I can regulate it, it has been consistent over the years not needing adjustments and when I checked many years ago would be with 15 to 30 seconds a week but I would need to check again for a better answer.
     
  7. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Please pardon my ignorance on asking with the correct vernacular as I'm still very early into my education about these things. What I'm getting at is does one or the other seem to use more power than the other?
     
  8. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    317
    2
    18
    Cape Town, South Africa
    Country Flag:
    I am making a 1 second pendulum wall clock. It has 4 pairs of gears in the drive train. Originally with involute gears but I wanted to reduce the drive weight and fall so I made cycloidal replacements for the last two gear sets. Same number of teeth, module etc. I then did extensive testing to see which used the least power. I was able to even run the involute wheels against cycloidal pinion or vice versa. I also had lantern rings and leaf pinion s for the cycloidal. There really was little difference between them but the cycloidals were marginally better. I decided to stick with involute for pairs 1 & 2 and use cycloidal for 3 & 4 just because it makes the clock a bit more interesting.

    But I want to state very clearly that my gears are pretty poor. They were made on a good cnc machining center by profile milling along the tooth outline on brass flat plate using a 1mm tungsten carbide slotting end mill. The big mistake I made was that I did not clamp the wheel once it had been crossed out and then cut the profile with a 0.2mm offset followed by a very light 0.2mm finishing cut. I cut then in shallow downsteps but to size so tool deflection (on a 1mm tool) was significant. I can adjust the depthing on the finished clock so that is not the problem. When the clock is running the rate is not consistent & I attribute that to inconsistency in the force at the escape wheel because of the poor gear train. I will be making Luke Mester’s pendulum timer to see what is really going on https://mb.nawcc.org/showthread.php?101788-A-home-built-clock-timing-machine-What-are-the-most-needed-features

    I am very interested in the post by cmnewcomer. The ‘identical’ clocks. It would be very interesting if you were to replace the drive weight on both with jerry cans of water and reduce the weight of both until the clocks stopped ticking. It would really tell us what the difference in friction losses are between the two gear systems.
     
  9. Jim DuBois

    Jim DuBois Registered User
    Gibbs Literary Award NAWCC Member Sponsor

    Jun 14, 2008
    2,934
    635
    113
    Male
    Magnolia, TX
    Country Flag:
    Region Flag:
    Running a clock mechanism on 12 pounds, be it involute or cycloidal, is not a grand accomplishment. Do the following with involute gearing on an 8 day mechanism and maybe it will convince some of us more traditional clockmakers... and this experiment has been duplicated by others and has more recently been superseded at least one party using ball bearings in the entire train....and it was not done with involute gearing...
     

    Attached Files:

  10. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Thank you for the input. I appreciate it and would very much love to hear from as many others as possible regarding this subject. Most especially if they have done any side by side testing like cmnewcomer has. So much so, that I would be willing to assist with the manufacture of the wheels if needed and they are coincident to my shop's capabilities. I have a lot to learn, and would rather not have to make my own article twice I should think. :) I would really like to be convinced that Cycloidals are superior to Involutes in Clocks and Watches, though. So, I'm hoping someone will chime in ( so many accidental puns, so little time ) < ha ha! > with some empirical knowledge or reference to set the record straight.
     
  11. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Jim,

    Thank you. I was writing while you were replying it seems. I'm still very green, here. Can you explain in lay terms what that means? I would be happy to assist in such an experiment if someone wanted to make one for comparison.
     
  12. Jim DuBois

    Jim DuBois Registered User
    Gibbs Literary Award NAWCC Member Sponsor

    Jun 14, 2008
    2,934
    635
    113
    Male
    Magnolia, TX
    Country Flag:
    Region Flag:
    Basically I am suggesting someone building an involute gear train clock and achieving a the mechanical efficiency already demonstrated by this old experiment...namely run an involute mechanism using a 5 oz drive weight while using a 19 pound pendulum...for 8 days...I have not done this myself but I have run a clock with a 30 pound pendulum for 8 days using a slightly under 16 oz weight....Personally I would not go to the effort of trying this with involute gearing, but I have only cut a few thousand gears and made a few clocks and all but a few gears were cycloidal , so what could I possibly know about gearing? I have repaired a number of involute geared clocks made by others, generally the intentions were good, the execution not so good.....there are a lot of things about clock making that is not covered in mechanical engineering, and those needed lessons can be painful.... since there are least 2 parties here that have already suggested I don't know anything on the subject I am inclinded to stand back and watch the effort go forward and hope that I am actually wrong....about the efficiency of involute gearing for clocks
     
  13. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Jim,

    Thanks for the clarity. I hope that you did not take any of my own comments as one of those that "suggested" you "don't know anything about the subject". If so, please accept my apology, for that is not even remotely my impression, nor my intention. In fact, the opposite is true, and I'm hoping to learn from that knowledge and experience.

    Too, I have a little experience with gears. I have quite literally made hundreds of thousands of them and designed a fair number of complex trains from nothing but requirements given. ( the vast majority of them being non-standard, as well. ) I have even made some Cycloidal gears, though admittedly for others and so have no idea of how they performed ultimately.

    As I've written earlier - I have no concern over which one is better suited. Rather, I desire to make the most suitable one for my own project based in fact and not theory, supposition, or superstition. It will be trivial for me to do either, given the equipment and experience. A part is a part, and knows not what it will do or where it will go. :)

    Basically, I would like to see some well made Involutes run in comparison to some well made Cycloidals and let the chips fall where they may.
     
  14. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Jim,

    While we are at it, can you tell me ( from the position of a Clock maker ) what typical accuracy one might expect with regularity in regard to both the center distances of wheels as well as tooth Profile achieved when making one's gears? I've not found much that addresses the latter, yet. My reading and researching is leaving me with the impression that several thousandths of an inch ( .0254 mm +/- several ) is not uncommon/unheard of. If so, this might very well explain some things. As well, it seems like a great and vast majority of wheels are being cut by form cutters and then hand finished to suit. Even the Involutes seem to have been done this way where I can find reference to them. That would speak volumes, in itself... What I am proposing to do ( regardless of what Tooth Profile is used ) is to actually make these in a manner of precision and accuracy normally encountered in other disciplines. Your observations and thoughts would be welcome.
     
  15. Jim DuBois

    Jim DuBois Registered User
    Gibbs Literary Award NAWCC Member Sponsor

    Jun 14, 2008
    2,934
    635
    113
    Male
    Magnolia, TX
    Country Flag:
    Region Flag:
    I generally use Thornton's cutters for both wheels and pinions, but also have a set of Bergeron wheel cutters. I also have a lot of involute profile cutters. As to the math and the like using cycloidal I tend to use the information that Thornton supplies...I did the attachment sometime back and it is shared here in other threads. Perhaps you will find it if assistance, perhaps not. Regards finishing wheels and pinions after form cutting them, generally the wheels are ready tooth wise to go right off the form cutters. Pinions, not so much, they may require polishing and hardening and the like depending upon what you want to do with them. And to that point many many antique clocks have run for hundreds of years with dead soft pinions and pivots. In some cases like fine French clocks, or English regulators one may find the finish on pinions to be mirror like and the steel to be dead hard, but much work is not.....depends on how much time you have and how nice everything needs to look. The are a number of arguments about polishing the teeth in wheels..... and there are several good threads here on how to polish pinions...regards the accuracy of wheel to pinion placement and the like in the world of clockmaking it is my approach to calculate the center distances, then verify and mark the proper depthing using depthing tools. Frequently the calculated depth and the most efficient depth are not the same. .002"-.005" will frequently be the difference between to optimum distance and the calculated distance, so close matters.

    Like you I have no vested interest in which tooth profile is best for our purposes, I am only passing on my experience. I have tried both, but I did not follow scientific methods with my various experiments. I can speak to those involute clocks that have come in here for remedy/repair/consulting.....they did not give me any cause to try one myself.....
     

    Attached Files:

  16. cmnewcomer

    cmnewcomer Registered User

    Mar 24, 2009
    48
    1
    8
    I usually don't respond to the threads here due to the tension it sometimes creates as appears to be occurring here. I have a lot of respect for the opinions of everyone here even if I don't understand or agree with them. Jim and others have always been very helpful and clearly they have years of experience that we are fortunate to receive here on the forum.

    The only reason for my original post was to reflect that while not popular, the involute profile works perfectly fine depending on the requirements. My requirements were simple, a reliable clock that will last a lifetime and I suspect that was the goal of my grandfather.

    In any event, thanks again to all for sharing.

    Best Regards.

    Carl
     
  17. David S

    David S Registered User
    NAWCC Member

    Dec 18, 2011
    7,201
    246
    63
    Male
    Professional Engineer - Retired
    Brockville, On Canada
    Country Flag:
    Region Flag:
    Carl I hope you change your mind regarding responding. Yes for sure there is some tension. Some of our members have a deep passion for what they believe in, either due to theory or experience. I just suck it all in and go with the flow. If someone doesn't like what I post.. so be it. Please keep sharing and don't get upset with the feedback.
     
  18. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Jim, Thank you for including your book. I have saved it so that I can re and re-read it in effort to continue toward my goal.

    Ahhhh.... This explains a lot to me. Thank you, Jim. I understand that this is normal ( and expected ) of the trade and its Tradesmen. My plan is a little more updated and possibly regarded as heretic by some, but none the less an inarguably more accurate one than that of form cutters. Truth be told, I don't doubt that some more traditional craftsmen might find it downright offensive. My plan is to use what is already available to me, though. Too, this explains why so many see slightly lacking results of Involutes, when created this way. ( form cutters being known to be very much less than perfect or optimal for them )

    What I, personally, would like ( whether Involute or Cycloidal ) is for the watch I create to keep good time, last my lifetime, and look reasonably attractive to me. :) My plan is to take advantage of modern tools, tooling, equipment, and materials to do so. Nothing new, I know, and I'm sure you guys see this all the time. ( some pompous twit walks in and proclaims their intent to build the be-all, end-all, thinking they know it all )

    I know that I don't know anything yet, likely won't know enough at any given time, and know that this is going to take a bit of a long time to wind up. ( there's those pesky puns again ) I've got a little experience with making things from scratch and a few days worth of experience doing so... ( I make small, precision parts and assemblies, Surgical/Medical prototypes, Implant prototypes, ... ... ... and ... ... gears for a living. )


    Right now, my thinking is to make these with my Wire EDM machine. This will allow me to do so without fear or consideration of overall size, with reaonably good accuracy down to .0001" ( .00254 mm ) repeatably and with very thin materials of my choosing, and in conditions of hardness of my choosing ahead of time so that the finished result will be hard enough to last, accurate beyond common examples, with surface finish superior to many.

    Other than that, my interest is drawn most to Tourbillons, and I've an affinity for self winders. And a pocket watch, to boot...

    You may snicker, now. :)

    Thanks, Jim. I appreciate your taking the time.
     
  19. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Carl, I too hope you continue to post in this thread, and the board. I can use all the help I can get! :) If you ever get the urge, I might be interested in making a set of new gears for your two identical clocks in an effort to compare and contrast the two.
     
  20. topspy

    topspy Registered User
    NAWCC Member

    Nov 17, 2013
    652
    35
    28
    Male
    Video engineer
    Oregon
    Country Flag:
    Region Flag:
    Can you elaborate on this? What disciplines, and what methods do you contemplate using? What type of tolerance improvement / finish improvement do you propose to acheive?

    Edit: Never mind, I see you answered above. I'd love to play with wire EDM...
     
  21. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Truth be told, it's a wonderful tool. And it can get expensive very quickly. But it's also brilliant for stuff like this. And as long as I have to pay for it, I might as well have a little fun with it! :)
     
  22. jhe.1973

    jhe.1973 Registered User
    NAWCC Business

    Feb 12, 2011
    496
    9
    18
    Arizona, USA
    Country Flag:
    Region Flag:
    Hello Zahnrad,

    I recognize you from your activity at the Home Shop Machinist forum. I am happy to hear of your interest in making your own watch. I also want to welcome you here as this forum has a great group of knowledgeable and helpful people.

    It looks like we come to horology from paths with several similarities not the least of which is that I grew up in Milwaukee.

    I note with a smile on my face that when you signed in here you didn’t refer to Milwaukee as tropical. Could that have anything to do w/this winter?:D

    Because you intend to use wire EDM for your machining I feel that my experience and areas of concern may not be of value to you, but here goes anyway.

    I long ago looked into the differences in the gear designs. I received the most of my information from the book, “Gears For Small Mechanisms”, by W.O Davis. I found this to be of particular interest in that he discusses both types of tooth forms.

    I don’t remember if he is the writer to mention it but my memory has long been that, historically, gearing followed the cycloidal path for centuries because of the cutting methods. When hobbing was invented, involute was adopted because it was much easier to hob an involute tooth. I.E. the involute shape was developed more for manufacturing ease than for ideal motion transfer.

    It seems to me that the contact thrust of an involute tooth is on a slight angle in relation to the pitch diameter. If so, this angle will have a slight cam effect trying to drive the arbors apart which could increase friction.

    This may not have any effect for 98% of horological trains (and may not exist at all) but I have not had the time to investigate it further, so it is just a gut feeling at this point. This is the only point that I think would have any relevance, but because of the small diameter pivots of a watch, it could be important.

    I too have no axe to grind as to which form is better. My own clock has something that marries both tooth forms and has fair efficiency – but still has given me surprisingly good results from an accuracy standpoint.

    I feel that one of the major advantages to hobbing a tooth (over form milling each tooth) is that several teeth are generated at one time AND both sides of each tooth is being finished/supported in the process.

    A drawback for extreme precision in form cutting teeth one at a time is that the last tooth is not as well supported as all of the others. This means that the last tooth is going to flex away from the cutter slightly because there is nothing to back it up any longer.

    Any machinist should be familiar with this cutter/part flex situation.

    If the teeth are first gashed all the way around and then form cut in a second operation, this flexing would be more uniform. However, the last cut will still be on thinner teeth than the first cut.

    In the book, “The Modern Clock”, by Ward Goodrich, he shows a wheel cutting mill from the Seth Thomas factory that had 3 spindles to allow the teeth to be formed with successively wider cutters. This will still have the same problem as I mentioned above, but at least it minimizes it to a large degree.

    A drawback to polishing pinion teeth is that the polishing always degrades the shape somewhat. How much degrading depends on the method used.

    If the polishing is done by hand, one tooth at a time, it is not possible to assure that each tooth gets the same amount of metal removed. If it is done under power, with a somewhat soft lap that rotates the pinion, the concentricity is more accurately maintained, but the consistency of the tooth shape suffers. This is primarily due to the lap’s contact spending more time at the top of the teeth then it does at the bottom.

    This unequal polishing can be compensated for a bit by leaving more material where you know the lap will spend more time.

    And then, of course, the lap wears ……. changing everything!

    So I think that it becomes obvious that this is more a matter of what gives you the degree of accuracy you are after.

    I have doubts that anyone will be able to set the record straight until someone takes the time to do a side by side comparison. Ideally I feel that this should be with the same plates & bearing surfaces and only changing the gearing w/a set that has the same component material and weight. An accurate, consistent method of measuring any difference is also vital.

    Yeah, I sure wish I had one! Maybe if I ask Santa and am a good boy .......... oh, too late!:whistle:
     
  23. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Well hello Jim! Thank you for the welcome. Very good to "see" you here. :) Ha ha! Yes, this winter has certainly been one of discontent... No, no other reason than it slipped my mind. There. I have corrected my location for you. :) You'll have to let me know about your time here some time. I love this city and state.

    Truth be told, I've been wanting to make my own watch for a very long time. I'm just finally at a place in life where I think it's the right time. I make some really neat things, every day. Some big. Some small. Some tiny. Why not a watch?


    No, please Jim - I wish to hear of all experiences and learn from them. Please - let me decide their relevance as I am positive that I can learn from everyone here. Your suspicions about Involutes is not completely wrong, but in truth there is both a "rolling" and "sliding" action being achieved and so that camming effect is so minute so as to be of much less concern than most other factors.

    If one considers this and also takes note of the fact ( possibility ) that the Cycloidals being used are in most reality not correct in their finished form ( as you touch upon, below ), then it seems to me that with attention paid to all the lessons learned over the last 100 years in regard to Involute gearing that it would very likely for Involutes of modern manufacture to perform just as well as proper Cycloidals. ( more on this below )

    Material hardness and surface finish also play factors.


    As I am sure you and others here are already aware, one ( of many ) ways around this is to back up the gear being cut with a solid ( or much thicker ) additional piece of material and cut through and past the gear, letting the mandrel of sacrificial back up material to provide support for the individual teeth/leaves.

    As I propose to WEDM the teeth or leaves, there is no cutting pressure being exerted upon them during creation.


    This is a very important and key point in my estimation and opinion! Once one starts hand polishing anything, one absolutely loses any sense of strict accuracy with regard to many of the important things about ANY gear, regardless of the type. Concentricity, tooth spacing, tooth form, tooth to tooth accuracy, and on and on...

    I plan to negate the need for any hand polishing by utilizing the fine finishing capabilities of the WEDM machine and be left with what amounts to a BETTER surface finish than one that is polished. As well, my current thinking is to use both thinner and harder materials, reducing both weight and friction. I see that many wheels are made of .031" thick brass... One can easily make a VERY skeletonized wheel, of full-hard Stainless Steel, only .015" thick, and having fully finished/polished teeth. I think this could be a real boon to my efforts. Your thoughts? ( Of course, one can also use Brass or any other material as long as it's conductive )


    Well, I'm willing to assist in that endeavor if possible. Maybe you would be interested in doing something like that? :) I would be happy to help create the gears for something of this concept in effort to be able to finally answer the question.

    Over 400 views in this thread. I was hoping more people would lend their insights by now.

    Thanks Jim!
     
  24. jhe.1973

    jhe.1973 Registered User
    NAWCC Business

    Feb 12, 2011
    496
    9
    18
    Arizona, USA
    Country Flag:
    Region Flag:
    I know I am at the same place - just trying to force my life to slow down so I can concentrate on the clocks I want to build/finish.:cyclops: Glad to see the tropical description. :D The most tropical thing about much of the Midwest is the warmth of the people & that I do miss.

    I feel the same about all the aspects of machining in regards to horology. A lot of ideas/products do not receive sufficient/complete development because some new technology comes along and gives a cheaper way to make something. Using the new technology to improve a current product isn't usually considered unless reduced cost can be shown. This happened with the introduction of electricity to keep a balance or pendulum vibrating.

    I wouldn't be surprised to find many improvements possible now to mechanical timekeeping just because of improved machining/measuring over the last 100 years.

    I should have been more specific in my description. I was picturing the sideways flex away from the cutter that would slightly change the last tooth flanks being shaped by the final cut.

    Are you going to attend the NAWCC national in downtown Milw. this June? My wife and I will be there and I feel it would be enjoyable to meet then.
     
  25. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    I've actually been giving that subject some thought, and I think it should be fun to create a timepiece of very high precisions and materials. I'm very much looking forward to giving time tested details the benefit of modern methods, materials, and treatments. Example - I've been wondering how NiTinol might work for springs... I use a bit of it for many of the Surgical/Medical devices and implants we make. I have a fair amount of Stainless and Titanium about, as well.

    Whoops... my error. I misunderstood. Thanks for clearing it up. And yes, you are correct. :)

    I was not aware. Count on it, then. It will be my pleasure to meet you in your old home town in a setting of my new interest.

    Thanks very much, Jim. I appreciate your time and input.
     
  26. jhe.1973

    jhe.1973 Registered User
    NAWCC Business

    Feb 12, 2011
    496
    9
    18
    Arizona, USA
    Country Flag:
    Region Flag:
    Hi Zahnrad,

    Just sent you an e-mail through this forum. Haven't done that yet, so let me know if you receive it.

    The form said that you should have my e-mail address now.:thumb:
     
  27. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    317
    2
    18
    Cape Town, South Africa
    Country Flag:
    I would like to repeat my plea to Cmnewcomer to measure the minimum required drive to keep his two ‘identical’ clocks going. In this thread wh have the following extracts from numbered posts.

    #4 cmnewcomer. The first clock I made used 0.8 module epicycloidal gear cutters I purchased from P.P. Thornton. The second clock I made used 0.8 module involute gear cutters from Russell, Holbrook & Henderson. I cut both the gears and pinions. Both clocks are setup with 12lbs of weight on the time and strike train and keep relatively identical time. I say relatively identical since I would need scientific measuring equipment to identify the difference.

    #6 cmnewcomer They are both 8 day longcase clocks with the same gear tooth count and the winding barrels are the same diameter so they both run the same length of time between windings.

    #8 Tok-tokkie I am very interested in the post by cmnewcomer. The ‘identical’ clocks. It would be very interesting if you were to replace the drive weight on both with jerry cans of water and reduce the weight of both until the clocks stopped ticking. It would really tell us what the difference in friction losses are between the two gear systems.

    #19 Zahnrad Kopf If you ever get the urge, I might be interested in making a set of new gears for your two identical clocks in an effort to compare and contrast the two.

    #22 JFE 1973 I have doubts that anyone will be able to set the record straight until someone takes the time to do a side by side comparison. Ideally I feel that this should be with the same plates & bearing surfaces and only changing the gearing w/a set that has the same component material and weight. An accurate, consistent method of measuring any difference is also vital.

    I would really like to appeal to cmnewcomer to run a comparison test between his two ‘identical’ clock.

    As long as they have the same escape and pendulum. The pendulum will need to swing exactly the same angle on both for the unlocking to take place.

    The amplitude of the swing is controlled by the impulse the escape delivers to the pendulum. The less the impulse the smaller is the arc of the swing .

    It takes hours for the pendulum to stabilise after it has been restarted, after 12 hours the period and amplitude should have stabilised but 24 is a more certain time period.

    If the drive weight is progressively reduced the amplitude of the pendulum swing will be reduced because the impulse is reduced when the drive weight is reduced.

    When the impulse is too small the amplitude is so small that the escape no longer unlocks and the clock stops. (I am not really familiar with graham escape – would that damage the escape?)

    By using a gallon plastic jerry can as the drive weight for each clock the quantity of water can be progressively reduced until the clocks stop ticking. Currently they run on 12 lbs. A US gallon of water is 8.35 lbs. I expect the clock should still tick on that drive weight otherwise a slightly larger jerry can will be needed.

    This test will really show the difference between the losses in involute and cyclodal gear trains in clocks. If the minimum weight to keep the clock going is much different then that shows the difference in the losses in the two drive trains because the force (and power) at the escape is the same to keep identical pendulums swinging the same arc sufficient to unlock identical escapes. So the identical force is arriving at the escape in each case & we know the original force at the start of the train. So the difference in the drive weights is the also the difference in the losses of the two drive trains.

    If it is possible to then swap the escape and pendulum between the two clocks and repeat the test it will confirm if it really is the gear trains that are causing the difference if the results remain the same. If the results are different then we will know how much of the difference was caused by the escape and pendulum.
     
  28. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:
    Jim, got it and replied. Thanks much. :) :thumb:
     
  29. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    70
    0
    Calif. USA
    I've not read every thing on this post yet but my understanding
    is that involute gears have to have specific cutters for each
    ratio, for both pinion and wheel. If not, they will not have minimum
    friction.
    Cycloidal have the advantage that they are more tolerant of cutting.
    One can have a couple cutters based on the leaf count for the pinions and
    one cutter for all the wheels.
    One doesn't have to worry about matching based on ratio.
    Tinker Dwight
     
  30. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:

    Tinker Dwight, that is not wholly correct. Very generally speaking, two of the more common ways to make gears* are with Hobs and Gear Tooth Mills ( form cutters ). With Hobbing, one has a single tool that can generate any number of teeth within the Diametral Pitch ( or Module ) of that Hob. In other words, the same Hob is used to generate the teeth for a 9T gear as is used to generate the teeth for an 85T gear.

    Gear Tooth Form mills are form cutters much like the form cutters for Cycloidals. Here, the cutters are available in "sets" ( or individuals ) that cover a range of tooth counts for each cutter, for a specific Diametral Pitch ( or Module ). So, a #8 cutter will cut 12 and 13 tooth gears of whatever specific Pitch one chooses. The # 7 cutter will cut 14, 15, & 16 teeth gears. A #6 cutter will cut 17 - 20 teeth, and on and on... These resulting teeth are - above all else - approximations of the correct tooth profile. This is where many people run into issues...

    * - there is also Gear Shaping, Gleason Processing, Wire EDM, Ram EDM, and a few other processes available.

    To be clear, Involutes do not need any ratio matching either. Everything is based solely on Pitch, Pressure Angle, and Tooth count. Further, they are very tolerant of center distance variations as well. My education of Cycloidals has left me with the impression that Cycloidals are very INTOLERANT of center distance variations.

    I could go on, but I don't think the actual study of Involute Gears is of interest to most here. :)
     
  31. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    70
    0
    Calif. USA
    Interesting that my opinions are mostly opposite.
    Tinker Dwight
     
  32. Zahnrad Kopf

    Zahnrad Kopf Registered User

    Mar 2, 2014
    15
    0
    0
    Toolmaker / Gearmaker
    Tropical Milwaukee, Wisconsin
    Country Flag:
    Region Flag:

    How so Tinker Dwight?
     
  33. cmnewcomer

    cmnewcomer Registered User

    Mar 24, 2009
    48
    1
    8
    tok-tokkie,

    I don’t mind doing the test but give me some time to get back to you as I have a number of competing priorities right now.

    It should also be noted for the record that I am not a clockmaker. I’m a hobbyist that happens to have grown up around a clockmaker but was too young at the time to take advantage of the knowledge he had to offer. My curiosity with this particular subject isn’t to prove myself correct or someone else wrong but to better understand in quantifiable terms the long term problem if any. I don’t believe that anyone disagrees that the epicycloidal form has less friction. But the nagging question from my perspective is to what avail?

    For example, assuming the average clock runs without repair for X years I would like to know that a clock made with the involute tooth form will require repair in X * Y years where Y is assumed to be a fraction from all the feedback we are hearing and possibly even as low as 0.5! While I realize this is difficult to quantify, another gauge might be a clock made with the involute tooth form will show accelerated wear on pinions and/or pivots in as early as X years. In other words, if this difference requires scientific equipment not available to the average person to measure, I would conclude that it is not an important design concern. Here again, a hobbyist making non-scientific conclusions agreed!!

    As stated before, I’m collecting non scientific feedback on a number of clocks made by my Grandfather that are over 40 years old to see if there are any problems emerging. The pinions show no visual signs of wear and the pivots show no signs of wear. So, what conclusion can be made from this non scientific information about the longevity of these clocks?

    It should also be noted that my Grandfather repaired clocks and watches for many years before deciding to design and build his own clocks. One of the major design decisions he made was to make gears that were much wider than the average clock. Most of his clocks had gears that were 0.140” wide to reduce pinion wear. In addition, he did not harden the pinions or the pivots as far as I’m aware. I’m fairly confident with this statement as he used pinion wire and I don’t believe the dies used to extrude the pinion would work with high carbon metals. As a matter of fact, I tried to get Quality Rod and Pinion to make pinion wire out of 1144 and I was told it was too hard on the dies and machines. 1144 can’t be hardened but does have a relatively high carbon content from what I understand.

    More to follow.

    Best Regards.

    Carl
     
  34. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    317
    2
    18
    Cape Town, South Africa
    Country Flag:
    Thanks cm.

    I too am not taking sides in this debate. The clock I have made uses both involute and cycloidal. I am a mechanical engineer so 'naturally' favoured involute since, to me, that is what a proper gear looks like. When my clock was requiring more drive weight than I would have liked I decided to try cycloidal wheels with lantern ring pinions where the trundles could rotate in the quest for minimal friction. The trundles ran in PTFE cages in the quest for minimum friction. There was little difference in the weight required to drive the clock. So I then made proper leaf cycloidal pinions. Again little difference. The small differences were lesser drive weight in both cases.

    But, having been exposed to clock posts on the internet, I started to associate the cycloidal tooth form with better looking clocks. So that decided me to use the cycloidal wheels and pinions in my clock - they appeal to me more aesthetically. They look like proper clock wheels!. My clock has 4 wheel/pinion pairs. It was just the last 2 that I had made the cycloidal versions. So I have left the first 2 as involute. I calculated the change in torque & hence tooth force and pressure and found that after the first gear set they become quite small and after the second set they were insignificant. The first wheel is 4.8mm wide (your grandfathers were 3.56mm wide), the second is 4mm wide the third is 1.6mm wide and the last is 1.2mm wide which is about as thin as I can conveniently cut or handle.

    The forces at the teeth and pivots is what causes the wear. If you can reduce the force then you will reduce the wear correspondingly. Of course the materials used, cleanliness of the interface and lubrication also affect the wear rate. I judge the effectiveness of the clock mechanism by the amount of power required to drive the clock. It was this that made me look at cycloidal gears and rotating trundle lantern rings in the first place. With the passage of time I was able to have my clock running reliably on 16 µW (micro Watts). I wanted to see how that compared to other clocks so started a thread asking for the weight, fall and time of fall for other clocks in this thread: https://mb.nawcc.org/showthread.php?108078-What-power-needed-to-drive-various-clocks&p=822282#post822282 I did not get many replies but found my clock is pretty good so the tooth loads must be pretty small in comparison to many clocks.

    But I am very interested in the difference in power transmission efficiency of the two gearing systems. Repeatedly we read statements about the superiority of the one against the other - I would like to see what the real difference actually is. Your clocks can give us the answer.

    That will only answer part of the question as it does not tell us what the radial loads are = the force separating the wheel and pinion = the load onto the pivot. I suspect that when the same torque is being transmitted the separating force of involute gears will be greater than cycloidal. The pressure angle of involutes is usually 20° which gives rise to the separating force.
     
  35. Jim DuBois

    Jim DuBois Registered User
    Gibbs Literary Award NAWCC Member Sponsor

    Jun 14, 2008
    2,934
    635
    113
    Male
    Magnolia, TX
    Country Flag:
    Region Flag:
    #35 Jim DuBois, Mar 9, 2014
    Last edited: Mar 9, 2014
    I don't think it necessary to build 2 clock movements, one with cycloidal and one with involute, to add at least some degree of clarity to this discussion. While I like the methods used in toc tockkies other threads of computing power consumed over time it is not clear to me that it tells us the actual power available from the train to the pendulum. There could be several times the power necessary to impulse the pendulum, or just enough to keep it in motion. Hard to measure with any degree of certainty unless the weight at the other end of the train is reduced gradually until the mechanism no longer escapes, and with precision clocks that can be very hard to discern also.

    I suggest building two trains, using wheel and pinion tooth counts similar to those we use in our subject clocks, making them as identical as possible, save the tooth forms. With the strain gauges available today I would measure the torque required to start each train, to run each train at a fixed RPM, time to coast down from a fixed rpm to stopped, etc., and do my calculations from the statistics accumulated from those tests (and others).

    There are some things that affect gear train efficiency that may be clear to some, or everybody, that need be considered. Some design formulas suggest that involute is most efficient at ratios of 4:1 or less. Usually in clocks we tend to use ratios of about 8:1, in some cases 12:1, and cyclodial seems to better tolerate those ratios more efficiently, a premise that would be nice to prove, or disprove.

    E Howard did a series of tests on their regulators 100 years ago testing 8 leaf pinions versus 10/12/16 leaf pinions in regulator trains, but they were testing how it did or did not affect timekeeping. They found, contrary to other published opinions, that the higher count pinions did not improve timekeeping. Not that provides us with answers to our questions, I just found it interesting......

    Some design formulas suggest that involute pinions should not be used with less than 8 teeth. Others suggest optimum efficiency cannot be obtained by using less than 16 teeth, a number that would require some rather large or fine toothed wheels if used...also something that would be nice to test.....if I were to attempt these tests I would do so using ball bearings in the train, just to minimize other sources of friction other than gear tooth efficiency.

    Just food for thought.....
     
  36. jhe.1973

    jhe.1973 Registered User
    NAWCC Business

    Feb 12, 2011
    496
    9
    18
    Arizona, USA
    Country Flag:
    Region Flag:
    Not trying to disagree, because I like these suggestions, just to add a bit more food.

    :)

    I feel that, because of the sensitivity to friction that exists at the escapement end of the train, it seems likely that just changing the escape arbor pinion and it's driving wheel to a different tooth form would yield enough information to evaluate which form has the least friction.

    Besides saving a bunch of work, this would assure that the train up to this stage is a constant for each evaluation.

    I find this interesting too. It seems to go along with what I have found in many writings on this subject.

    Many writers equate friction & accuracy as though reduced friction is the only way to increased accuracy.

    I have found consistency in power transmission to be far more important than reduced friction.

    However, tiny driving weights impress people and sell clocks.........................:whistle:
     
  37. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    317
    2
    18
    Cape Town, South Africa
    Country Flag:
    It is not necessary to build 2 clocks. They already exist - cmnewcomer has two clocks with identical train counts; one involute & the other cycloidal.
     
  38. cmnewcomer

    cmnewcomer Registered User

    Mar 24, 2009
    48
    1
    8
    This is turning out to be more of a PITA than I thought. I bought a counterpoise weight set with hanger and the larger 4 and 2 pound weights interfered with the pendulum. So I milled off a quarter inch or so and was able to begin the test.

    Starting on the conservative side by beginning with the original weight and then reducing it each day was a bad approach. It took me over two weeks to realize I could get down to about 5.8lbs before the clock stopped. This is where things got real finicky. Moving up to about 7.2 before things seemed to work well again. After allowing the epicycloidal to run for about a week on 7.2lbs, I removed the test weights and filled one of my brass shells up but only had about 6.8lbs when I closed it up so decided to try that. Well, it worked for about a week and then stopped. I know have about 7.4lbs and will report back as I get more information.

    In an attempt to save some time, I loaded up another brass shell with 6.8lbs several days ago and started the involute with that weight. It's still running but I suspect it too will stop and need more weight.

    It seems that anytime I adjust the time quickly with the friction pinion things get a little unsettled so I'm not sure how informative or reliable this test will be. More to come.

    Best Regards.

    Carl
     
  39. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    317
    2
    18
    Cape Town, South Africa
    Country Flag:
    I have also adjusted the drive weights on my clock, reducing until the clock stops. My clock has a center sweep seconds hand on the same arbor as the escape. What I learned is how vital it is that it is perfectly balanced - at the limit I could see clearly when the heavy side was rising or falling by the action of the escape & movement of the second hand. The strange thing is I thought I had balanced it pretty well only to find it was significantly out when I checked it again. The minute hand is also balanced.

    As you say, it takes days to run the tests. I keep an Excel spreadsheet adding a new row at the bottom each day to record the drive weight & other stuff with comments about what I notice. That has been going since 2010 and has about 2500 lines & I am very pleased to have that record. It has been a slow process!

    I am very interested in the results of your tests.
     
  40. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    70
    0
    Calif. USA
    With a 20 degree thrust angle, the additional load on the
    pivots is only about 6%. I'm not sure it will be easily seen.
    On a well maintained clock, it might make the difference
    of a few years in 100 difference of life.
    It might be a problem as it wears as the trust angle increase.
    It might be more useful to purposely make two clocks with
    about the same amount of wear and see which continues
    to run with minimum weight.
    Which would be worse I could only guess.
    Tinker Dwight
     
  41. cmnewcomer

    cmnewcomer Registered User

    Mar 24, 2009
    48
    1
    8
    This should be a final report back on this test which further demonstrates the fickle nature of these movements. I got Clock 1 with the epicycloidal gears to run on 6.6 lbs for a couple of weeks. Anything below that and it sometimes stopped mid week.

    Clock 2 with the involute gears has been running with 7.0 lbs for a couple of months now reliably. It did stop once unexpectedly but I was able to set it back it motion.

    I now wonder what a reasonable amount of additional weight is needed when a clock will not be run in a controlled environment. That just made me laugh since my shop is by no means an ideal place for a clock. It would appear that the 12 lbs I was using previously was excessive but what would be a good compromise?

    While testing the clocks, I did not have the dials installed which may require more weight due to the moon wheel advance but hopefully that is very small if at all.

    I'm going to clean Clock 1 and install it in a new case I had made several months ago so I can finally get this up in the main living space and enjoy it more.

    Best Regards.

    Carl
     
  42. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    317
    2
    18
    Cape Town, South Africa
    Country Flag:
    Many thanks for that Carl. I agree with you in wanting to leave it there. No significant difference (<10% on a statistically insignificant sample).

    As for extra weight for reliable running - I would suggest 20% - say 8.5 lbs. It will be a significant reduction on the original 12 lbs (30% less) which will be kind on the mechanism.
     

Share This Page