I have a 3/4 steel rod in a 3 jaw chuck, sticking out
about 2 inches. When I turn it down, the part farthest
away from the chuck is about 5 thou bigger. At first
I thought it was just the part deflecting. I added
a live center to it in the tailstock, but the results
were the same.
Then I decided to make a check. I have the compound moving
in the same direction as the carriage. I put an indicator
holder on the tool post holder. With the indicator on the
cross slide I move the compound the full length of travel.
I see 5 thou difference in height. I then put the indicator
on the way and found 10 thou difference in height.
I interpret this to be that the compound is not parallel
to the cross slide and not parallel to the ways.
So my question is is this normal? I was moving the compound
to turn down the stock. Am I doing this wrong and should
move the carriage instead?
I must say your method of turning a piece of stock straight is a bit
unusual. Yes, one normally moves the carriage, not the compound.
For the most part, most engine lathes are left with the compound set at 29
degrees, just like you're going to chase a thread, and cuts are generated
with the cross slide and/or the carriage. The compound is normally used
for generating angles. You can use your compound for straight turns,
but it usually will be in the way if you're using a center, and it's clearly
a bitch to get dead straight.
I'm not exactly clear where you were using your indicator, so the readings
you mentioned don't make a lot of sense. If you were tracking the top of
the compound as it was traversed, unless it was finish machined, the reading
may not mean anything. You have to be certain to check against the right
surfaces before the reading will have any value. If you feel you have a
problem, please describe it in better detail, and post a pic if possible.
If your compound is just 0.5º out of parallel with the axis of the spindle
(and carriage), you would be out .017" (radius) over 2" of cut.
I sometimes use the compound to advance the tool for smaller parts. When I
do this, I indicate the side of the compound by moving it along the z-axis
with the carriage. Assuming the lathe manufacturer did his/her homework,
this should yield fairly good results over shorter distances. 2" is
excessive and yes, you should be using the carriage feed.
"Harold & Susan Vordos" wrote in
I'll have to agree with Harold on this one...using the compound for
actual straight turning is a bitch. Are you sure you don't have a slight
angle in the compound producing the 5 thou?
Set your indicator on the machine bed, and against the machined surface
on the side of the compound, move the carriage back and forth..sounds
like the compound is on a slight angle.
I was under the impression that the compound was generally set to 30
degrees (hence forming a 1, 2, sqrt(3) triangle) for the purpose of
making very fine adjustments to cut depth. At least that's the way
that I've used mine.
It's not recommended for chasing 60 degree included angle threads. The
whole idea behind feeding at a slightly reduced angle is for the threading
tool to take the majority of the cut from the side of the thread that loads
the cutting tool against the lead screw. That means you have to change
sides with the compound depending on the nature of the thread you're
machining. Opposite sides for left versus right hand threads, and
opposite sides for internal versus external threads.
There are various well founded reasons to set your compound properly. It's
not carved in stone, so most guys do what they feel is correct for them.
I've spent almost my entire life in commercial machine shops and consider
myself a rather gifted lathe man. I've always set my compound at 29
degrees. Interestingly, I worked with one man, the finest lathe man I
ever encountered, heads and shoulders above the average guy, who insisted
that his compound be set at 30 degrees, and the quick change tool block set
exactly square with the spindle. He always did fine work, but you leave
yourself open to one little problem when you thread at 30 degrees.
The markings on machines are generally pretty good, but rarely dead on, what
ever that may mean. If, by chance, the marks on a given machine are off
as little as a few minutes of angle, and it's the wrong way, if you set your
compound at 30 degrees, you may be set beyond the half angle of the thread.
What that does for you is to leave a stair step on the back side of the
thread, and eliminates the possibility of cleaning up the thread unless you
plunge cut a few passes. By setting the compound at 29 degrees, you're
more or less guaranteed that the thread will continue to wipe clean with
There's one big fly in the ointment where chasing threads to the formula is
concerned. The formula addresses a sharp V thread, but in practice they
are rarely, if ever, machined. Based on that, if you grind a proper
threading tool, you have no idea how the tool tip relates to the pitch
diameter of the thread. For chasing a thread that you want to fit to a
nut, that might be fine, but if you're trying to work to proper dimensions,
you have no clue where you are with the pitch diameter. The only way to
know is to measure with wires, or to use proper thread gages, either rings
or snap gages.
This argument is endless. There seems to be two factions, one that thinks
that threading what I like to call "properly" is not necessary, especially
for the home shop guy. I maintain that it is no harder to do things
properly than it is to do them in a less than acceptable manner, so I
always encourage folks to learn the "right" way. If you do, when the need
arises, you know how to deal with the problem at hand and it's no challenge.
Hope this let a little light in for you, Harry. If you're interested,
here's a link where there was a very heated discussion about this very
subject. You might get several viewpoints to either support or debunk
your concepts. Hard to say.
It's a long thread, over 100 entries. Some VERY good.
Some other reasons to set the compound along the ways (and I've done
this myself on occasion) is either because of a lack of a travel
indicator on the carriage (compound dial then shows infeed), or
when drilling - with the drill chuck mounted on the toolpost.
(here the carriage runs to a stop on the bed, and the drill feed
is done with the compound. This allows rapid backing out to clear
chips, and rapid feed back in to the bottom of the hole, via the
carriage traverse wheel)
Reading the original post, the *real* problem he was having with
the lathe was that it was turning taper when using the compound
to turn the diameter. The trouble was simply that he had the
compound angle incorrectly set. One cannot rely on the divisions
on compounds for 'zero.' I'm suprised he got as close as five
thou in a few inches. All of the other issues about should a
dial gage read zeros when swept (via the compound) across the
top of a part in the chuck, or across the bed, etc. are moot.
In such a case the indicator *should* read zeros, but will only
do so in a brand new machine with no wear. In principle the
axis of travel of compound should be in a plane parallel to the
bed travel of the carriage. Wear over time will destroy that
Which is another good reason to not use the compound to turn
please reply to:
JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com
Harry didn't say: "A 1, 2, 3 Triangle is NOT a 30 degree triangle". You
Harry said "(hence forming a 1, 2, sqrt(3) triangle)" meaning a triangle
with an opposite side of 1 unit, an hypotenuse of 2 units that form an
adjacent side or sqrt(2*2 - 1*1= 3) = 1.732 units which he just called
the sqrt(3) side.
Mount your indicator in a suitable adapter (if necessary) on the tool
post and use the longitudinal feed to move it along a round bar
observing the change in indicator reading. NOTE! This is done with the
lathe NOT running! Do this several times turning the chuck between
trials. Significant variation would indicate that either the test bar
is not straight and round or the chuck jaws are not true.
Now reverse the bar end for end and repeat. If the results are
different, the test bar may be tapered.
I did all this on my Smithy when I got it with the result that it will
turn a few inches with well under one thou taper. Beyond a few inches,
you have to consider deflection.
No one has mentioned the possibility that his tailstock was offset. If his
was a new lathe, it would be highly unusual for the tailstock to be
correctly aligned. I am more surprized, though about the compound not
being parallel to the ways. I'd call that a serious defect in the lathe.
However, if you look at the geometry of the situation, a 10 thou difference
in height over two inches would not result in more than a thou or difference
in diameter. But that still looks like a hefty slope for the compound. A
small piece of swarf on the compound's mating surface could do it, though.
I'd check the tailstock allignment and the dirt before I started agonizing
over a faulty lathe.
Although, in the original post, he said that the effect
did occur when *not* using the tailstock, to start.
please reply to:
JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com
Odds are that the lathe hasn't been "leveled" so that it will cut straight.
Leveling consists of making the two ends of the bed "level" so that there is
no twist in the bed. Note that I put level in quote marks because the
simplest way to insure that there is no twist in the bed is to make the ways
level all the way across the width of the bed. In truth, you don't need the
ways to be level but rather just untwisted. .005" in 2" is a bit radical
but it just may be twisted that much.
I'll also note that if you want cuts square with the spindle, you don't move
the compound bur rather the entire carriage as the compound may not be setup
properly to the ways but the carriage always is.
Losing weight is easy! If you ever want to lose weight, eat and drink less.
Works every time it is tried!
Ok, now I feel like an idiot. It is obvious that
the compound had a slight angle. If it had been
off more, I probably would have figured it out.
I guess being 5 thou off only eyeballing it isn't
too bad. I tried cutting using the carriage travel
and that came out ok.
I bought the lathe new 2 weeks ago. It was kind of a
mess, but I got most of it cleaned up. I did level
it (at least somewhat). It isn't levelled left to
right yet, but I did work on front to back. Using
a Starrett 98-12 level it looks good placing it on the
ways on both ends. If I put the level on the cross
slide, the bubble ends up on different marks. When
I move the carriage from one end to the other, the
bubble moves 1 division.
I have 2 more weeks to try the lathe. If I don't
like it I can return it. I still have issues that
have to be sorted out with the place I bought it
from. But before I deal with those I wanted to make
sure that the lathe itself is ok.
So now that I know it cuts reasonably straight, is
there anything else I should look at?
It sounds like sweeping from the toolpost holder to
the ways may not be important. I guess this just
changes the center height a bit when moving the
I would rather make this lathe work for me
than return it. However I don't want it to be
a big project either.
I have used a mill, but not a lathe. I supposed
I used the compound because it felt more familiar.
I wasn't quite ready to kick in the power feed.
Up until now I've been doing all my lathing
in the mill.
Lew, I didn't post 1, 2, 3. I posted 1, 2, sqrt(3), the sides of a 30,
60, 90 degree triangle. This is why the sin of 30 degrees is 0.5, and
that of 60 degrees is sqrt(3)/2. [Note: Sqrt(3) = 1.732...].
Bill, exactly. My real point in that post was that in making linear
cuts, 30 degress permits a vernier (and easily calculated) adjustment
to the depth of cut that is not easily achieved using the cross slide.
While I'm an engineer/physicist and only an amateur machinist, but
while a Co-Op student I was taught this simple trick by a very
experienced tool and die maker.
Cutting and chasing threads is another issue.
The 98 is a good level, but it is not the master precision level that you need
for leveling the lathe. The master precision level is ten times as sensitive.
Maybe you can get the seller to set it up correctly.