Gear Train Calculator

[evan-e-cent]

Does writing software count as a "Project"? Well if it does I would like to tell you about a program I wrote for lathe users. RideTheGearTrain.com

It was designed to tell you how to set up the gear train to cut any thread. When it runs it usually gives you a long list of different ways you can set up your gear train. Beside each setup there is a button labeled DRAW. Click that and it produces a scale drawing of the gear train so that you can spot any problems that might occur.




The data input is in 5 steps and then you run the program.
(1) First choose your lathe.
(2) Enter the set of gears that you own.
(3) Choose what type of thread you need: metric, imperial or feed rate
(4) Enter the % error you are willing to accept in the final thread. It may be zero or perhaps 2%.
(5) Enter the pitch of the thread you require. Actually there are 3 ways you can do this:
(5A) Simply enter the pitch or TPI
(5B) Enter a range of values to produce tables
(5C) Look up a reference table eg metric or UNC etc.
(6) Finally click RUN.

If you choose option 5C the reference tables provide a lot of extra information which allows the program to display "Thread Specifications" after the list of gear trains. This includes two tables of possible metric or imperial drill sizes that could be used for the thread you chose. Not just a single drill size that most tables give you. I have been adding calculation of the stress on the bolt and thread when tightened to a given torque (under construction).

It will work for any lathe without a gearbox, including mini-lathes. For lathes with a gearbox it needs to know the gear ratios. There is a list of about 25 lathes with gearbox data included. It was originally written for Boxford and other SouthBend 9A clones so if you have a clone choose a "Norton" gearbox. Myford is included and so are a few other models like Colchester, Chester, Nile, Hendley, Precision Mathews etc. If you want your lathe added please contact me or look up "Custom Gearbox" at the end of the list of lathes.

There are extensive help files with help button beside each input box.

There are 3 Youtube tutorials listed below the menu.
My email address is listed at the top of the info, above the menu.

Keep in mind that I am not an engineer - my only training was 4 years of engineering in High School with my father as the teacher.
Let me know if you have any questions or discover any problems or flaws.

PS it does not cost you a cent and there are no hidden costs.
I hope you find this useful. Enjoy!
Evan

 

[Weldingrod1]

An excellent project!

Sent from my SM-G892A using Tapatalk

 

[RJSakowski]

Very much a project! I have dabbled a bit in finding possible gear variations for the Grizzly G0602 lathe using a spreadsheet and a spreadsheet for calculating depth of cut using a 29º compound feed for common Imperial and metric pitches but haven't put them together in a single package.

I haven't done much in the past two years as I have added an electronic lead screw to my lathe, eliminating the need for any gears.

I haven't delved deeply into your program but one thing that I noticed offhand is that when I entered the G0602 into type of lathe, it enters 8 tpi as the lead screw pitch. The G0602 has a 12 tpi lead screw. I will review your program more deeply after I've had my coffee.

I'm sure you will find plenty of beta testers on this forum.

 

[cwilliamrose]

I started to enter the data for my lathe but found it was going to be pretty involved and I wasn't sure of all the terms being used. I did look at the tutorial and I now have a better idea of how to proceed. When I get the time I may try again to get mine entered and check the results against the spreadsheet (specific to my lathe) I created years ago. The spreadsheet is more of a manual process which could take some time to go through for a given pitch. I originally needed a 3.3333 TPI solution which is why I did the spreadsheet. Turned out there were several solutions that were very close and after a bit more trial and error I found a couple of exact solutions. Having the computer do the work would be a very nice option.

 

[evan-e-cent]

Hey, THANK YOU both for the feedback. I have been through many iterations of input design trying to make it easy to use. Let me know if there are any areas that are particularly confusing. Long chunks of text explaining things should help, but tend to make it look more complicated and 'forbidding'.

RJS, this is the kind of information I need to expand the usefulness of the program for people who have a gearbox. Each case has to be entered separately, although, if your lathe is otherwise identical to another Grizzly lathe you will have noticed that you can change the leadscrew pitch setting.

The easiest way to proceed would be to ask you to send a photo of the gearbox label showing the threads it can cut. I also need to know the gears that are used with the gearbox but I think these will be included in the thread tables.

I just ran the program and see what you mean. I have split the entry for Grizzly lathes into two models so that 0602 is listed separately with it's 12 TPI leadscrew (DONE). But then I need to check that it is giving correct results and for that I need the info from the gearbox labels.

Could you send them using the email address listed at the top, above the menu when you run the program. You might like to check it yourself now that the G0602 is listed with a 12 TPI leadscrew. I expect it will give incorrect answers until I change the "primary ratio" of the gearbox. This is a factor that I multiply all the gear ratios by.

I wonder how many different Chinese lathes there are with gearboxes! Perhaps I should ask on the Grizzly page. I am afraid to ask. Many of these complex gearboxes take me a day or two to decipher! I hope the G0602 is similar to the G0752.

Thanks

 

[RJSakowski]

As far as I know, the G0602 and G0752 only differ in the spindle drive. The lead screw and gear box should be the same. The on line user's manual actually covers both lathes.
Here is the chart that I made for the 602 gearbox.

Gear Box Ratio (Out/In)
III C	II C	I C	III A	II A	I A	III B	II B	I B
0.5833	0.416667	0.5	1.1667	0.833333	1	2.3333	1.666667	2

And here is a threading table listing most possible Imperial combinations (not mine, from the internet, unknown source). I say most because the the 104, 120, and 127 tooth gears aren't brought into the mix.

I hope this helps.

 

[evan-e-cent]

Thanks for this. So I guess both of these lathes have a leadscrew with 12 TPI.
When you say the only difference is in the spindle I assume you are referring to spindle speeds.
I assume the number of teeth on the spindle gear are the same - 40T.
I have put them both into the program now as separate lathes.

I wasn't absolutely sure what the gear ratios you listed meant so I derived them myself and got the same answers!
I looked up 12 TPI on this table and calculated the gear train ratio and it was 1.0
The overall gear ratio has to be 1 to cut the same thread as the leadscrew.
Therefore the gearbox setting A1 must also have a gear ratio of 1.0
From that I can calculate all the gearbox settings for the program.
From the table:

1B = 6TPI
1A = 12 TPI
1C = 24 TPI

So relative to A1=1 the ratios (in terms of TPI, not gear ratios) are
1B=0.5
1A=1
1C=2

Lets look at the number lever - With the same gear train a=60 b=40

Actually the online table shows the gear ratios in the grey panels on the left and your message contained the same.
1A = 12 TPI Ratio =1.0000
2A = 14.4 TPI Ratio 0.83333 = 12/14.4
3A = 10.29 Ratio 1.1667 = 12/10.29

However my program uses "F-factors" for the gearbox and these are proportional to TPI which are the inverse of gear ratios!
That way the table looks like the TiP table on the gearbox in most cases.

So I have to invert these ratios as follows:
1A = 12 TPI Ratio =1.0000
2A = 14.4 TPI Ratio 0.83333 = 1.2
3A = 10.29 Ratio 1.1667 = 0.8571428

This gives me everything except a calibration factor that i call Primary Ratio of the gearbox. In my program I use the equation:

TPI = primaryRatio x GearboxRatio x Leadscrew / GTR

Re-arranging to get the Primary Ratio

primaryRatio = (TPI x GTR) / (GearboxRatio x Leadscrew)

We know the data for several threads but I like to use the one that is the same as the leadscrew, so they cancel:

primaryRatio = (12 x 1) / (1 x 12) = 1

primaryRatio =1, so in fact no correction is required in this case.

I have entered these specs in my program.

I had already split G0602 from G0752 so I will leave them as separate entities, but now the specs are identical.

Could you try it out and see if it looks correct. I have included the extra change gears you suggested. Of course the 127/120 combination gives you metric conversions so the program should give you metric threads.

It is interesting that the G0750G is totally different and very complicated.

 

[Just for fun]

Very cool stuff. Thanks for sharing!

Tim

 

[B2]

@evan-e-cent

Does writing software count as a "Project"? Well if it does I w

I just found your thread. Pretty nice website, but I like to see the code. I too wrote a program, but in Excel, to help folks figure out the gear arrangements for their lathe. Lots of lathes have gear plates on the lathe which are very incomplete. So they do not really know what can be made. This was the case for my PM1440GT I purchased from PRECISION MATTHEWS - Quality Machine Tools . So I started out to figure out what I could make and then started teaching myself more about the Microsoft macros. So, I posted it a while back and just recently up dated it to included a number of features. So you can download it and see the gear ratios I needed for my lathe and for a few others that I incorporated. I live in Pittsburgh and this is the location for "PRECISION MATTHEWS - Quality Machine Tools". These lathes and many others seem to all be made in Taiwan or China and are basically the same as many other imported lathes, but with different model numbers. Anyway, the gear boxes, even though they are described in some of the manuals are not described completely nor correctly. I also find that it is common that the Feed rates maybe wrong and it is common that the x-feed rates are incorrect. If you have studied this I am curious if you have found the same.

My Excel program has individual sheets for each lathe model. This sheet lets you input the gear box ratios and the external gears. It is a flexible layout to allow for a variety of lathe models, all in one work book. It is laided out with pull down menus so that you can manually select gears or gear box levers to see what TIP/feeds you get. Macros that I wrote also generates a large table listing of all of the possible gear arrangements. For my 1440GT, which has a gear box and comes with 10 external gears plus the 120/127 metric conversion, the table generated has 32,000 possible combinations when you allow that the feed bar can also be used cut threads. However, some of these are redundant. Removing the redundant ones, which the macro does automatically, reduces the list to 21840. I then wrote macros to manipulate this table. It will search to find Imperial, metric threads as well as any other value such as feed rates. It does this with wild cards so that either exact values or approximate values can be found. Many times even though there is no exact thread available there are close approximations, better than the accuracy of the operator. They sort the orders etc., hide columns, etc. I also wrote a macro which calls up a users list of threads he would like to make and searches for each and then generates a table... so one can make a table for all of the standard thread values.

I did not hide my code so you can see how it works. I have written lots of code, but never in VBa so this was a learning process. The code is hardly optimized or cleaned up. I saw no point in posting it as a web site.... as I do not try to make money via code. I am an engineer and physicists and work in the physical sciences. Code is a means not an end for me. So there is no slick front for it, but there could be. Anyway, Excel is not the best programing system for this sort of thing, but everyone has a copy so the universality to the users was important.

Anyway, if you would like to can take a look at my posting here:

Improved Version: TPI, Feeds, X-Feeds: Generalized Lathe TPI Excel file: "TPI_ManyLathesRev1

Dave L.

 

[evan-e-cent]

B2​

Hi B2, thank you for your detailed description. It has been a while since I visited this site, so sorry for the delay. The method you described is quite similar to my approach, but I have done it using PHP programming on a server so that it is easily available to anyone, and can be used in a workshop on an iPad etc.
So you generate all possible gear combinations and the thread pitches they can produce and once it has been produced in XL you can search it for any thread you want. Sounds efficient.
My program asks for the users list of gears and lathe type and then generates all the possible combinations. While they are being calculated they are compared with the thread that the user requested. If they match to with a specified acceptable error, the gear train is recorded in an array. When it is finished the array is sorted into order, usually sorted with the smallest errors first.
As in your program the user can enter a list of threads to solve eg all the pitches in M-series metric threads. It then produces a table.
Alternatively they can look up tables of standard threads to get the pitch or TPI.
The user can decide whether his gear train is going to be simple with just a stud gear and lead-screw gear, or whether it will contain one or two compound gears (pairs of gears).

You asked about the equations used. These are shown in an information box after the results when you run the program. The HELP files give more detailed explanations.

In trying to include every possible variation in lathe design I tried to make it simple to use, but still people tend to find it daunting to have 5 to 8 steps available to set it up. Originally all these steps were on a single page, but someone complained about too much scrolling, so I split it up. That has the advantage that you can go back and alter one parameter such as the % acceptable error, and run again without going through all the steps.

I have been studing the problem of finding possible metric-imperial conversion gears and will post that separately.