Grizzly G4003 Lathe-modifications to RideTheGearTrain

(free online to calculate the change gears needed to cut a thread.

I had quite a busy time in the days leading up to Christmas. I had contacts from a couple of people who needed their lathe gearboxes added to the program (RideTheGearTrain).  One lathe had crazy features I had not heard of before. It was a Chinese-made Grizzly G4003 (2005) model sold by the American Grizzly company.

 

The spindle shaft is 16mm and the leadscrew shaft is 19mm, so they have two sets of change gears. This meant that I had to modify the program to handle it.  It would have been a huge job to re-write the whole program, but not long ago I added a filter system. The user enters ALL their change gears initially, but they also click the button, currently labeled "Exclude Large gears", and here the user can enter a separate list of gears permitted for each position (or stud) in the gear train. If the program produces results with gears that are NOT permitted in any position, those results are deleted (filtered out) and only the reduced list of results is displayed.  

 

I added this filtering system a few months ago to prevent the display of results where a gear is too large to fit on the machine. A user had this problem.  So the user enters a size limit (by number of teeth) for any position they want restricted.  For example on some lathes there is a casting near the stud gear that prevents large gears from being fitted. If the size limit for the stud gear is 60 teeth, then the user enters 60 as the maximum size for the stud gear. Results using larger gears are filtered out.

 

An advancement on this idea was to actually measure the distance between studs on the banjo, and from the DP (Diametral Pitch of the gears) the size of gears can be calculated.  When there are gears on two 'studs' (or shafts) on the banjo, the radius of each gear is proportional to the number of teeth on that gear. The space required is the sum of the two radii. We can use the sum of the teeth on the two gears and from that calculate whether the two gears will fit between the two studs. If they don't fit, the filter program will remove the results from the list of results to be displayed.

 

This method is rather complicated and I doubt whether anyone uses it. I haven't had any feedback about it. I did not have my Boxford in Tennessee when I wrote this. Now I am in New Zealand I could measure my lathe and see if it works. It probably isn't foolproof because fitting one pair of large gears on the banjo will limit how much room is available for the next pair. We could add up the teeth of ALL the gears on the gear train, but that would only work if the gears are arranged in s straight line, which they are not. Some more thought may be required.

 

Going back to the weird Grizzly lathe, it has another quirk.  It is fitted with a 91/86 compound pair for imperial to metric conversion.  It is designed in such a way that this gear cannot be flipped over. RideTheGearTrain automatically tries flipping over compound pairs and it is quite surprising to me that it produces nearly as many results with the conversion gear flipped.  A conversion gear with a ratio approximating 1.27 is fairly close to 1.25 which happens to be quite handy!  So, how can we prevent RideTheGearTrain from displaying results with this gear flipped?  Easy. Just use the filter system to tell the program to filter out any results where the gear in position Compound 2 (comp2) is larger than 87. 

 

Now with this filter system we can accommodate this peculiar lathe.  Now I understand better why it is done this way. The gears designed to go on the leadscrew shaft have a long extended boss on one side - like a built-in spacer. If you put it with the spaced to the right, the gear engages with the metric conversion gear. If you flip the leadscrew gear over, so that the spacer is on the left, it now meshes with the gear on the compound pair which acts as an idler. This is used for cutting imperial threads. These leadscrew gears with the extensions cannot be used as stud gears.  So, this lathe comes with two of these leadscrew gears and 6 change gears for the stud gear position.

 

 

Now comes the question, why do they use a 91/86 compound pair.  The gear ratio is 1.058139534883721   The answer can be found if you click the button labelled 'Metric-Imperial Conversion' (which I described in a previous post but now we have a real world application).  There you will find a table produced from an equation I derived.  

 

I used PhotoShop to make a copy of this table as a JPEG file (.jpg) but groups.io says "Dropped file type is not supported"!  What!  So I tried JPEG2000, and finally loaded .PNG but get the same message. Whether you get it is a matter of luck!) I wonder if the message is too long.

 

 

R= 25.4 / (Li * Lm)

 

 

The Grizzly lathe has an 8 TPI imperial leadscrew, so look down the column marked 8TPI. You will find a ratio that is very similar: 1.0583333. The difference between these numbers is incredibly small at 0.0183%.  Now look across this row where it says 3mm.  What does this mean?  Well, it means that this gear combination will convert an 8TPI leadscrew to behave like a good French metric leadscrew with a pitch of 3mm (just like the metric Boxford).  In the above equation Li is the TPI of the imperial leadscrew, and Lm is the pitch of the metric leadscrew.  What if they had used the standard conversion gear 127/100 with its ratio of 1.27.  You can see from this table that it makes an 8TPI leadscrew behave like a 2.5mm metric leadscrew.  

 

If you want to use RideTheGearTrain to find a suitable compound pair for any application, you can enter the ratio (scroll down to the bottom of the page). The program then produces a list of gear sizes that will approximate this ratio. If we enter 1.0583333 as in the above example, it displays a long list of compound pairs that you could use. Alongside each pair you see the amount of error produced. The smallest errors are displayed in red. Now scroll down this list to find the smallest error. Guess what! The smallest error is for the conversion gear 91/86 with -0.0183% error!  That is the best choice.

 

Merry Christmas and happy New Year

 

Evan Lewis

Lathe: 1955 Boxford Model A with screw cutting gearbox, power feed with several accessories, hand tools and a pillar drill press.
Try my Free Online Gear train Software:https://RideTheGearTrain.com
You enter a thread pitch or TPI and it shows you a range of gear trains and gearbox setting to use and even a scale drawing of the gear train.
It also includes calculations for taper turning by the tailstock offset method, and cutting speeds.
It includes the specifications for many thread types eg metric, UNC, BSW, and BA.
Displays drill sizes for tapping threads at any percent thread depth (with full explanations).
My YouTube Channel and Playlist about using an engineers lathe:  https://www.Youtube.com/evan-e-cent
Project to build a Greek Hero steam engine and measure its power output: https://HeroSteamEngine.com