Free 173 Best Homemade Tools eBook:

# Thread: Improving lathe turning accuracy.

1. ## Improving lathe turning accuracy.

The compound rest holds a secret to improving the resolution of infeed setting on a lathe. This is useful for both turning with a regular cutting tool and when using a tool post grinder.
I think that most lathe uses in Europe and the UK and its old colonies tend to set the compound rest at 0 degs. except when using it for turning tapers. On the other hand I have noticed that many US machinists leave their compound at the angle used for thread cutting, which is just a touch under half the included thread angle. Typically 29 or 29.5 deg for 60 deg thread. Particularly as most hobby machinists only cut threads once in the blue moon it seems rather illogical to leave the compound thread ready.
Is there a better general purpose angle to set the compound rest to most of the time. I think there is and I think that it's 5.7392 deg, that might be a touch fine to set easily so let's say five and three quarters. Basically I set it as close to that as is reasonable by eye. For the non maths phobes that number is arcsin(0.1).

The sketch below shows why I use this angle;

We can see that with that angle the inward feed of the toolpost is 1/10 of the travel of the compound slide. We have therefore increased our setting accuracy by a factor of 10 compared to adjusting the tool post position with the cross slide dial. !0 thou on the compound slide is 1 thou infeed. Likewise 1 thou on the compound is a tenth infeed.

Here is an real example.

I mentioned in another post that I always like to do an error analysis when working with numbers, so how critical is the calculated angle. Roughly the error in infeed is approximately 4 to 5% for 1/4 deg error in setting. That sounds a lot but let's put it in perspective in comparison to using the cross slide.
Say we set the compound at 5.5 deg instead of the target 5.74 deg then if we move the compound by 10 thou aiming for 0.001" at the work piece we will only get 0.000958" or less than 1/2 a tenth error. Try that with the dial on the cross slide. In any case if you can't set the compound angle any better than that you probably couldn't set the infeed any better than 1/4 thou on the cross slide.

I have been careful to talk about tool post movement or infeed and not actual metal removal. When using a lathe tool, advancing the infeed by 1/10 thou with usually not remove 1/10 off the radius, often it will just rub on the surface. Tool sharpness and material properties have important influences on this and is beyond the scope of this post. However, if you are using a tool post grinder then a 1/10 infeed might actually take 1/10 off, when using a fresh wheel.

A side advantage, though often an important one, is that a small angle on the compound rest gives more clearance when you are working close to the tail stock. Many lathes with the compound at 0 deg. restrict the infeed position of the cross slide.

Two views of the offset compound rest close to 5.74 deg.

Thanks for watching.

2. ## The Following 24 Users Say Thank You to tonyfoale For This Useful Post:

aphilipmarcou (02-17-2017), C-Bag (01-28-2017), Cascao (02-01-2017), Christophe Mineau (01-27-2017), jjr2001 (01-30-2017), jwitkowski (01-11-2019), Moby Duck (01-30-2017), mudnducs (01-30-2017), mwmkravchenko (01-25-2018), Okapi (01-31-2017), olderdan (01-31-2017), Paul Jones (01-28-2017), saintrain (07-01-2018), Steved53 (01-09-2019), Stevohdee (06-13-2018), that_other_guy (01-14-2019), threesixesinarow (09-28-2018), Toolmaker51 (01-25-2018), tooly (05-27-2018), Tule (05-28-2018), UncleBob (05-27-2018), Wmrra13 (01-28-2018), Woodgeezr (01-30-2017), zarembak (05-27-2018)

3. Really interresting post, thanks Tony !
For sure it's true that threading is not my favorite occupation on the lathe

One comment would just be that this kind of adjustment would not be measurable for the one who possibly has a DRO on the cross slide. (of course, simple reflection, I'm just thinking about the pertinence of DROs on my lathe by the time being)

thanks again

4. ## The Following 2 Users Say Thank You to Christophe Mineau For This Useful Post:

jjr2001 (01-30-2017), tonyfoale (01-27-2017)

5. Originally Posted by Christophe Mineau
Really interresting post, thanks Tony !...........
One comment would just be that this kind of adjustment would not be measurable for the one who possibly has a DRO on the cross slide. (of course, simple reflection, I'm just thinking about the pertinence of DROs on my lathe by the time being)

thanks again
Christophe,

I was wondering who would be first with that comment. Thanks.
In fact I do have a DRO on my cross slide and it is very useful, it is hiding under the white cover in the photos.
Many DROs have resolutions of 0.01mm (4/10 thou). Of course the best way to use the angled compound when you are chasing accuracy is to have a DRO on the compound rest as well. Cheap digital calipers can easily be fitted but when I need to be very concerned about accuracy I simply use a dial gauge on a magnetic base on the compound rest.
I use the cross slide DRO to get close to size and then use the compound rest to finish off.

6. ## The Following 3 Users Say Thank You to tonyfoale For This Useful Post:

jjr2001 (01-30-2017), olderdan (01-31-2017), Paul Jones (01-28-2017)

7. Thanks tonyfoale! We've added your Lathe Accuracy Improvement to our Machining category,

tags: lathe

8. ## The Following User Says Thank You to DIYer For This Useful Post:

jjr2001 (01-30-2017)

9. If you use this idea, be careful when using a carriage stop to turn up to a shoulder. Cos(arcsin(0.1)) = 0.995 so the tool will move closer to the shoulder by an amount only a tiny bit smaller than whatever you dial in on the compound.

10. ## The Following 4 Users Say Thank You to mklotz For This Useful Post:

jjr2001 (01-30-2017), Paul Jones (01-26-2018), PJs (01-30-2017), tonyfoale (01-30-2017)

11. Nice job Mr Foale! Currently helping a friend build an LSR bike....you're being tracked! <G>

12. Originally Posted by mklotz
If you use this idea, be careful when using a carriage stop to turn up to a shoulder. Arccos(arcsin(0.1)) = 0.995 so the tool will move closer to the shoulder by an amount only a tiny bit smaller than whatever you dial in on the compound.
Marv,

You are right and wrong, arccos returns an angle not a relative length. arcsin also returns an angle which we usually express in deg. or radians. If in deg. then it 5.74 but that is an invalid argument for arccos. If in rads then it's 0.1 and arccos = 84.3 deg or 1.47 rads, neither equals 0.995.
However, cos returns a relative length and Cos(arcsin(0.1)) does = 0.995.

Of course you are right about turning to a stop, you must allow for the self evident fact that winding the compound rest in with a shallow angle will invalidate the carriage stop setting.

13. ## The Following 2 Users Say Thank You to tonyfoale For This Useful Post:

olderdan (01-31-2017), Paul Jones (01-26-2018)

14. Originally Posted by mudnducs
Nice job Mr Foale! Currently helping a friend build an LSR bike....you're being tracked! <G>
what does "being tracked" mean?

15. Originally Posted by tonyfoale
Marv,

You are right and wrong, arccos returns an angle not a relative length. arcsin also returns an angle which we usually express in deg. or radians. If in deg. then it 5.74 but that is an invalid argument for arccos. If in rads then it's 0.1 and arccos = 84.3 deg or 1.47 rads, neither equals 0.995.
However, cos returns a relative length and Cos(arcsin(0.1)) does = 0.995.

Of course you are right about turning to a stop, you must allow for the self evident fact that winding the compound rest in with a shallow angle will invalidate the carriage stop setting.
Yes, brain fault on my part. It should have been "cos". I've corrected it in the original post.

16. ## The Following User Says Thank You to mklotz For This Useful Post:

Paul Jones (01-26-2018)

17. Originally Posted by tonyfoale
what does "being tracked" mean?

Page 1 of 2 1 2 Last

There are currently 1 users browsing this thread. (0 members and 1 guests)