OT Metric gearing
Close
Login to Your Account
Results 1 to 19 of 19
  1. #1
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default OT Metric gearing

    I have an old A....s lathe manual. No I don't have the lathe and not interested in details about the lathe but the manual covers everything. Great book for a newbie! (No I'm not a Newbie) I cut a lot of metric threads, have a Hendey set-up to cut metric threads and gears to convert other lathes to metric if I have too. Manuals for the lathes have the typical 127/120 compound gear referenced and I know how to compute the desired threads if a chart is not available. However this At..s manual claims 44 and 52 tooth gears combine to give a ratio 44/52 or .846154 which is an almost exact function of 25.4 the English to metric ratio. I'm Lost!!! can someone explain that and how I can use that to calculate gearing needed to produce metric threads with Inch lead screw.

  2. #2
    Join Date
    Aug 2014
    Location
    bainbridge island
    Posts
    1,002
    Post Thanks / Like
    Likes (Given)
    184
    Likes (Received)
    242

    Default

    i did some math and came up with a ratio of 1.2508 using that 44/52 combination which i'd not seen before. then i forgot how i did it.

    anyhow,

    take all the gears you have and plug them into here and see what you can do.
    Change Gears for Threading - LittleMachineShop.com

    the issue you've presented is a bit complicated because you're asking what's the closest approximation you could get from a given fixed ratio of change gears.

  3. #3
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default

    I understand the complexity being different machines have different gear ratios and lead screw TPI. To be honest I've never thought much about it before and tried to come up with ratios to product metric threads by buying gears available on ebay. Most lathes have 20 degree gears and most companies have the size I need in 14-1/2. I did find a 120 tooth gear in 20 degree have it compounded with a 127 tooth 14-1/2 degree. I mounted a 14-1/2 degree on the head stock stud and the 20 degree mates with the existing 20 degree in the quick change box. I have no problem computing the needed gears. My Hendey, P&W and L&S manuals have information on calculating the required gears plus the Machinist Handbook has it too and explains that 127 is the lowest number 25.4 will divide into. All the lathe manuals listed above use the 127/120 compound set to produce metric threads with a lathe with inch lead screw. The At..s manual is quite interesting, attachments were available to do just about everything including tool and cutter grinding! They even had a lead screw reverse attachment so metric threads could be cut without reversing the motor.
    My question is the sentence in my original post about 44 and 52 tooth gears are a direct quote from the At..s manual. What do they mean by "almost exact function of 25.4"?
    On my #1 Hendey I can get all the metric threads by simply changing the "stud" gear and the thread selection. I haven't figured why but noticed the gears given 45, 54, 63 and 72 tooth are all divisible by 9. I'm using a 36 tooth for most of my threads.
    Also now that I'm thinking about it there has to be some kind of similarity in the gear rations simply because there must be an exact amount of carriage movement per rotation of the spindle to give TPI. Using 127tooth gear I understand but I lost trying to figure why 44/52 or 846154 is an almost exact function of 25.4

  4. #4
    Join Date
    Dec 2012
    Country
    AUSTRALIA
    Posts
    1,200
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    376

    Default

    All the old machining textbooks give a conversion using a 63 tooth gear,and always state the 63 isnt used as a approximation of 127/2,but as a .....well the rest I forget.

  5. #5
    Join Date
    Jul 2004
    Location
    Houston, TX USA
    Posts
    28,619
    Post Thanks / Like

    Default

    Seems rubbishy to me

    44/52 = .84615
    52/44 = 1.1818
    No where near what is needed

    Here are two dead on combinations

    127/100 = 1.27
    127/50 = 2.54

    And their reciprocals

    .7874 and .3937

    ON EDIT:

    You can take away the rubbishy aspect by providing the missing info - like the other two gears.

    If I have a 30 stud gear and a 45 screw gear that is 1.5:1

    Those in combination with the 44/52 compound make a fairly useful ratio

    .84615 X 1.5 = 1.2692 : 1

    Not very far off from the gold standard of 1.27 : 1

  6. Likes Hudson, muckalee, Paolo_MD liked this post
  7. #6
    Join Date
    Oct 2012
    Location
    NE Ohio USA
    Posts
    586
    Post Thanks / Like
    Likes (Given)
    108
    Likes (Received)
    209

    Default

    My gut feeling is that the ratio is a misprint. However it would be helpful to know the leadscrew pitch to determine what ratio would have been correct. Please let us know the leadscrew pitch for the lathe that must not be named. Thanks.

    Best Regards,
    Bob

  8. #7
    Join Date
    Oct 2012
    Location
    NE Ohio USA
    Posts
    586
    Post Thanks / Like
    Likes (Given)
    108
    Likes (Received)
    209

    Default

    On further investigation the reciprocal ratio (52/44) does produce some close approximations depending on thread pitch of the lead-screw:

    8-TPI produces 3.75mm < 3 micron error
    10-TPI produces 3.0mm < 2 micron error
    12-TPI produces 2.5mm < 2 micron error

    formula is 52/44 * 25.4 * (1 / LeadScrew Pitch)

    The confusion was stating that the ratio is a factor of 25.4. It is a factor of the ultimate metric pitch calculation though. Also there are better combinations that produce less error, as others have mentioned. So 'almost exact' is open to interpretation.

    Best Regards,
    Bob
    Last edited by rjs44032; 03-13-2019 at 10:41 AM. Reason: clarity and grammar

  9. #8
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default

    I don't have the lathe so I had to search the manual for the lead screw TPI, It's 8TPI. I don't think the 44/52 is a misprint it is .846154 so it would have to be a double mistake. Furthermore the chart showing the gearing needed to cut metric threads all have both 44 and 52 listed. I don't know when 25.4 was determined as the inch to metric conversion but this manual uses 25.4. It's very old I think but it's loose leaf bound with a plastic multiple ring binder but the pages are crumbly. I know this forum does not allow discussion of the At..s lathe because it's a hobbyist machine but the photos in the book shows industrial applications. I'm not sure how I got it but I think it was with a pile of manuals I got from Curtis-Wright auction. I had at one time manuals for just about every machine made. Still have quite a few. I do remember seeing in another part of the building that was not being auctioned quite a few At..s lathes. I just sent a large flat rate box of manuals to OzarkWoodWorkers, in that pile was Walker Turner manuals and catalogs. I do have a pile of Bullard manuals if anyone's interested.
    I never thought I would have interest in an At..s lathe though my father had one made for Craftsman. During WW2 he made parts for the war effort in the basement of my grandfathers home. It was a simple basic machine with a complete set of change gears though I can't remember if it had those for metric. Looking at all the attachments available and I had more money than I knew what to do with I would like to have the complete outfit just to look at!
    Anyway what interests me is the gear ratio they are using, all other references to metric conversion included 127 tooth gear and usually compounded with a 120 but any gear will work and being near the size of the 127 the 120 will allow for other gear sizes.

  10. #9
    Join Date
    Oct 2012
    Location
    NE Ohio USA
    Posts
    586
    Post Thanks / Like
    Likes (Given)
    108
    Likes (Received)
    209

    Default

    Yes I agree. It's not a misprint. Please read my last follow-up post. If it is 8TPI lead-screw, then 52/44 does evaluate to a 3.75mm pitch < 3 micron error (formula in my last post). Hope this helps.

    Edit: Also 25.4 is just the MMs in an inch. 127mm is exactly 5" that's why 127 is used for perfect metric conversions.

    Best Regards,
    Bob

  11. #10
    Join Date
    Feb 2006
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    12,481
    Post Thanks / Like
    Likes (Given)
    568
    Likes (Received)
    3503

    Default

    In 1982, I decided to make some watch lathe screw cutting attachments. Some watch lathe slide rests have metric feed screws and others have inch feed screws. Some watch threads are metric and some are inch, so I had to make the attachments capable of cutting either type of thread from either type of slide rest. I had no way to cut a 127 tooth gear, so I spent some time searching (with a pocket calculator; this was before Internet) for gears I could cut that would be close enough. I found that 40 and 63 is accurate to one part in 8000 or .012498% and could find no better combination.

    The 37 and 47 combination is not nearly as good, with an error of .021281%.

    Larry

  12. #11
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default

    Yes I know what 25.4 is and the reasoning for 127 as I mentioned in my previous post. However I assumed the ratio given 44/52 was "almost exact" Doing the math for the gear train listed for 1.25mm it came up 1.25075757575mm, so the error is almost .00003". Not too bad!
    Thanks guys, you made me think! But I'm still impressed on what I see in the photos of attachments. It's more than a Hobbyist lathe but I'm only a guest here so will not argue the decision made. Plus I don't have one and prefer the full size lathes.

  13. #12
    Join Date
    Mar 2005
    Location
    Germany
    Posts
    484
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    45

    Default

    44/52 x 3 gives you 25.385, which is about as close as you'll get to 25.4 unless using a 120/127 transposer. It's an approximation.

  14. #13
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default

    Probably because the lathe is so small putting a 127 tooth gear inside the cover will be difficult and .00003" error on a lathe is not an issue and so not necessary to get it better. Illustrations of the feed gear chain so it's capable of having 3 sets of compound gears so I'm assuming it's possible to lower that error amount but why on a lathe that could not preform to that degree of perfection! In my calculation for 1.25mm thread the screw and the nut will have to be 125mm long and the interference will only be .0757575mm.
    If I get some time I will post another topic on my P&W feed gears. I completely removed the quick change gear box and the gears that drive it. When replacing the drive gears I noticed the needed idler and mating gears were there but someone compounded the Idler (easy to do because it's mounted on a shaft with a spacer, spacer removed and another gear added). Then added another gear on the quick change shaft to mate with the compounded gear. Size of the added gears are much larger so that the other set don't mesh plus the pitch is larger. Stock DP was 16 and added gears 10 or 12. Plus the gear teeth numbers are and odd number so that the ratio is fractional number but when doing the math when compounding one with the original idler the ration came out the same as if the two stock gears were used, I'm wondering why that was done!
    Last edited by Froneck; 03-14-2019 at 07:36 AM. Reason: spelling

  15. #14
    Join Date
    Oct 2012
    Location
    NE Ohio USA
    Posts
    586
    Post Thanks / Like
    Likes (Given)
    108
    Likes (Received)
    209

    Default

    Just a quick note. As I stated before, the 52/44 does evaluate to a 3.75mm pitch < 3 micron error or 0.2273%. There is an interesting thing that I have found with approximate metric transposing gears. Dividing them in the gear train decreases the error. So the 1.25mm has error of 0.0758% and 0.75mm pitch will have 0.0455% error. Multiplying has the opposite effect. So as you go larger than 3.75mm the error increases. 7.5mm pitch will have 0.4545% error.

    So approximate transposing gears that target larger pitch would be preferred in my opinion. The 3.75mm is a good target number for the most popular metric threads. Ok I'll leave it there now.

    Best Regards,
    Bob
    Last edited by rjs44032; 03-14-2019 at 12:08 PM. Reason: grammar

  16. #15
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default

    I get more info on the P&W 16" lathe gears and post another topic but it's not set-up for metric. For what ever reason I'm wondering why a set of gears were installed that have a larger pitch and the originals are still in the machine, one of which is compounded to the larger tooth gear. So it hasn't been crashed. After I post the tooth numbers maybe it was done to add ability to do metrics. I did convert to metric by using a 127/120 compound I made from some Boston change gears I had. But they are only 1/2" wide cast iron. Later when taking heavy cut with high feed rate I stripped the 127 tooth gear. The metric job was completed previously so I simply went back to the inch set-up and continued cutting with out problems.

  17. #16
    Join Date
    Oct 2012
    Location
    NE Ohio USA
    Posts
    586
    Post Thanks / Like
    Likes (Given)
    108
    Likes (Received)
    209

    Default

    Wow, sorry to hear that. I did see a set of transposing gears on Ebay for ridiculous low price. But they are 14 DP. So I have no use for them. Perhaps they could be a replacement for yours. They are certainly beefier than 16DP.

    LeBlonde lathe Metric double gear for 15"swing Regal servo-shift 120 & 127 teeth | eBay

    Best Regards,
    Bob

  18. #17
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default

    Quote Originally Posted by rjs44032 View Post
    Wow, sorry to hear that. I did see a set of transposing gears on Ebay for ridiculous low price. But they are 14 DP. So I have no use for them. Perhaps they could be a replacement for yours. They are certainly beefier than 16DP.

    LeBlonde lathe Metric double gear for 15"swing Regal servo-shift 120 & 127 teeth | eBay

    Best Regards,
    Bob
    I checked ebay, they are a total of 1.375 wide so not much wider that the 1/2" I had since thee has to be a space between the gears plus being 14DP makes it harder to get the other gears, 14DP is not very popular. No mention what the pressure angle is so to find 14DP and 20PA that Leblonde probably used will make finding other gears really hard. Gears on my lathe are about 1 or 7/8" wide. I looked on ebay for 16DP 20PA cutters but only found Canadian #8 cutter. McMaster only has 14-1/2PA and MSC I think don't show any at all! I did find Ideal Gear or something like that has what I need, I just didn't call them yet being I have metric gearing for 2 other lathes.

  19. #18
    Join Date
    Aug 2014
    Location
    bainbridge island
    Posts
    1,002
    Post Thanks / Like
    Likes (Given)
    184
    Likes (Received)
    242

    Default

    20 degree pressure angles look more like the teeth in car transmissions than the square blocky teeth on most lathes which are 14.5 degrees.
    a friend of mine can print a 127T 16DP gear, no larger. but if i really needed one i would cut one out on a band saw from a template. yes, really.

    how you stripped the teeth on yours i have no idea.. the leadscrew should have sheered off.

  20. #19
    Join Date
    Dec 2010
    Location
    Mifflintown, PA 17059
    Posts
    1,391
    Post Thanks / Like
    Likes (Given)
    14
    Likes (Received)
    124

    Default

    The square blocky teeth are 20 degree. The involute pattern is more noticeable on 14-1/2 degree gears. Plus having owned over 30 lathes made by the top makers all stated they had 20 degree gears in the parts manual. Granted I had the lathes made in the USA made for industry not hobbyist. Possibly the light duty toys they are now making overseas might use 14-1/2 degree gears. Furthermore the lathes I mentioned do say they have 20 degree gears. So I'm thinking the manufacture's manual is right plus if trying to use 14-1/2 Degree gears in the machines mentioned they will not mess correctly!
    Plus I did mention I broke the teeth using heavy Feed, I was not threading, on the P&W both feed and lead screw don't run at the same time, one or the other. Plus it is not possible to engage the feed with the half-nut engaged. I have seen crashes done while threading, none broke the lead screw on the lathes made in the USA for industry.
    Being I broke the gears in feed that is much slower than threading in that the carriage gears slow the speed down and should put less strain on the gears. The 1/2" wide cast ioron gears couldn't take the load! Removing it and using the "Inch" gears since I was not threading worked fine even with heavier cuts! I have made quite a few gears and would never use a gear I cut on a band saw unless it was done to rough out the blank. The attempt here is to get the gears running quieter and anything cut on a band saw will really make a racket.


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
2