3 tries and 3 fails to make a 1/4 20 thread

[RJSakowski]

If you really want a challenge, my Grizzly G0602 Had a thread dial where the marks weren't concentric with the shaft so while some marks lined up exactly, others were off by a half a thread. A new dial cured the problem.

Threading On A Grizzly G0602

On March 19th, I replied to a thread begun by Cody Killgore entitled "Threading, Something Went Wrong Here" (post #19) where I show evidence of an problem with the threading dial on my Grizzly G0602. I had determined that the dial markings were not concentric with the gear shaft, creating an...

www.hobby-machinist.com

 

[WobblyHand]

@RJSakowski Wow, not a challenge I'd like to experience! How did you instrument that error? Don't want to thread jack, but the methodology of the measurement intrigues me.

 

[RJSakowski]

@RJSakowski Wow, not a challenge I'd like to experience! How did you instrument that error? Don't want to thread jack, but the methodology of the measurement intrigues me.

It was actually fairly simple. I just recorded the DRO reading for the z axis position at the points where the dial markings lined up with the index mark and plotted the difference in the readings between adjacent marks. A dial indicator could be used as well or a digital caliper with the depth rod connected to the carriage and the body clamped to the bed would work.

 

[ErichKeane]

If you have an 8TPI lead screw, every other engagement is usable for 20 TPI. I'm guessing you are off by 1 on at least 1 of your cuts. The result would be a two-start thread, each of 20 TPI

So the problem it seems is you just need to practice engaging the same way each time.

 

[pacifica]

Well, I've given up for the day on cutting my first threads on my South Bend 10K. My objective was a 1/4 20 thread in a piece of brass stock. I took my OD down to 1/4 inch and set my lead screw speed.


View attachment 357509

I put my drive into back gears and set the RPM to about 150.

View attachment 357510

My compound was at 29 1/2 degrees. My tool is ground to 60 degrees. My first pass looked OK and measured 20 threads per inch as expected. But after several passes of about 5 thousandths, I kept coming up with a mess that looked like this.

View attachment 357511

The threads were rough and read at 40 TPI, not 20. I tried to engage my half nut on any of the 8 lines on my thread gauge since it was an even number of threads.

View attachment 357512
I did notice that my gauge does not allow me to engage exactly when the lines are on the mark. I can go a little before or a little after. Maybe that is where where I am making my mistake. The next time I can try always going a little before the mark. Does anyone know if I can adjust the threading gauge to engage closer to the mark?

Anyway, any advice is appreciated.

When I was first learning single point threading I had rpms at less than 50, as low as 30 until I gained knowledge of the lathe and confidence. Now 200 rpm is no problem. Also I thread straight in (90 deg.) to remove a variable. Until you get more experience can you leave lead screw engaged and reverse to the start?

 

[twooldvolvos]

When I was first learning single point threading I had rpms at less than 50, as low as 30 until I gained knowledge of the lathe and confidence. Now 200 rpm is no problem. Also I thread straight in (90 deg.) to remove a variable. Until you get more experience can you leave lead screw engaged and reverse to the start?

Thanks for the tips. I could reverse to the start but from what I am reading, disengaging the half nut and cranking the carriage to the start is a preferred method, so I thought I would try to get into a good habit.

 

[BladesIIB]

Based on everything I read here and just checking the thread chasing dial on my lathe, here is what I recommend you test quickly. My thread dial had 8 lines, 4 numbered and 4 not. However, it will engage in 16 total places around the dial so it will engage between each Mark. Your earlier post was that you were not consistent in engaging just before or just after the Mark. Mine is also off slightly and I always engage just after the Mark on the dial. If you did one just before you would have split the thread being in one of the 8 non marked spots. Quickly check your engagement and test for this. You don’t even have to run the lathe. Just move the carriage slowly by hand and engage the nuts in the next position and I bet you will see there is an extra spot between marks. Good luck. And let us know when you get it sorted.

Edit: as others have said a little pressure before helps it drop in. I am fine with mine off a bit I start to push on the line and it drops in just past. Easy to keep up with.

 

[derf]

One thing nobody mentioned...make sure that the thread dial gear is fully engaged with the leadscrew. If there is excessive clearance, the dial will not be spot on, and the backlash will be evident. Matter of fact, while your at it, remove the whole assembly and clean it. The gears have a habit of accumulating chips and debris preventing full engagement.

 

[BGHansen]

Hopefully you are close, but make sure you don't touch the feed reverse lever. That synchs your spindle to the gear box. If you disengage the feed reverse lever and rotate the spindle, you'll very likely be off your mark.

You may also try hitting a mark on the thread dial and sticking with that number (mentioned many times above). At the end of your pass, shut off the lathe, back off the cross feed and move the carriage back past the starting point. Keep going until your mark lines up again. Engage the half-nuts with the lathe shut off and chuck the carriage back and forth to confirm you are right on your mark. Advance the cross feed and compound and start the lathe for the next pass with the half-nuts already engaged.

Bruce

 

[Forty Niner]

Couldn't see your compound, so I ask the question about the 29 1/2 degree angle.
29 1/2 degrees from the X axis or the Z axis?
Some compounds are marked with zero one way and some another way.. On some compounds, setting the angle at 29 1/2 degrees to thread will indeed make a mess.