View Full Version : Beginner Lathe Help Please

11-23-2004, 04:46 PM
I am taking a class at the local community college right now and we started on lathes. Well my problem is zero'ing out the micrometer collar on the cross-slide feed. Our instructor showed us how to spin up the metal and then slowly move the cutting tool into the turning metal till it just makes a mark and then zero out the collur from that point. I think I am doing this right but I do not find it to be very accurate. Is there a better way to do this?

-Troubled New Machinest

Spin Doctor
11-23-2004, 05:08 PM
About the only way I know to do it. One thing you could do though is mark up the surface with layout fluid or a dykem paint stick (sort of like a magic marker) and ease up to the work with the tool. You will take off the marking media before you touch the work

11-23-2004, 05:30 PM
By the time you actually see your cut, you're already a few ten thousandths into it, times 2 for reduction in the diameter.

Is this what you mean by not accurate or are you referring to repeatability?

In actual practice, I usually take a light cut, measure it and then zero the micrometer dial. This will probably get you a failing grade if it's not what the teacher wants http://bbs.homeshopmachinist.net//wink.gif


11-23-2004, 05:34 PM
I like to set the dial so the zero is very close to a real size, like 0 will be turning .250,.500,.750,1.000 you get the idea.

for radius set dials 0= .500, 1.000,1.500,,,

so if you touch and turn a diam to .980 set the dial to .020 (on a diameter type dial or .010 on a radius dial)

[This message has been edited by tattoomike68 (edited 11-23-2004).]

11-23-2004, 05:42 PM
nheng, I think you hit it on the head and I need to think about that more. It just seems like a squirrley way to do it.

11-24-2004, 04:48 PM

Without stacks of practice even getting the work piece in a 4 jaw you aren't going to get the job centred to much better than .001-.0005 so you will need to take a light cut first so you can guarantee a round job to measure!!

The other thing you can do is to hold a dial gauge against the compound slide so that as you advance until it touches the work note the dial guage position then bring the tool back half or a thou (accounting for backlash) and you'll be pretty close to zero, then take your first light cut.

It is a waste of your and your employers time to chase .0001 on the first cut, unless the job warrants it. It is nice to learn to do it, so you can though!!

I have also seen people use a piece of paper between the tool and the job, move the paper and wind the cross slide in until resistance is felt in the paper - measure the thickness of the paper (80gsm copy paper is .004) and then advance the cross slide to compensate.

11-24-2004, 06:44 PM
The only comment I would add is that most stock is not round or square for that matter, unless ordered that way. I see no problem doing as your teacher sugested. Don't waste time on that first cut. Just get the tool aligned to round the stock and when it cuts the full perimeter you can begin with measurments to determine what has to be removed. Just remember X2.

11-24-2004, 08:13 PM
you can align the tool with the tailstock live/dead center? This way you dont cut any material?

11-24-2004, 09:21 PM
Scotch tape.When I train a new hand I have them stick a piece of scotch tape to the part.Then spin it up and advance the tool in until you see tape shred.It works because the tape is only .0015-.002" thick.

11-24-2004, 09:31 PM
<font face="Verdana, Arial" size="2">Originally posted by Elninio:
you can align the tool with the tailstock live/dead center? This way you dont cut any material?</font>

very good idea, it may be eye balling but would get you in the &lt;.005-.010 range.

I like to "time" all my tools and know where they are at.

I ran a lathe for 6 years that you could not move the dial(i did a set screw fix on it so it would not free spin after a fast backout while threading, it came loose all the time)

I would get a 1" shaft running true run the tool holder close with the tool loose(a stopped machine) and pull the tool to the work piece and tighten the screws to hold it there.

9 times out of ten I would turn a part at the zero and it would be 1.002"

I did the same with threading tools and could make the same part over and over and it worked great.

[This message has been edited by tattoomike68 (edited 11-24-2004).]

[This message has been edited by tattoomike68 (edited 11-24-2004).]

11-24-2004, 10:15 PM
As you can see, there are several ways to "skin a cat" noted on this board. All of these methods are correct, and use varies with the personal choice.

I will address seting the dial, and centering the tool.

I do the spin it up and touch method, but also ask the student to take a .005 or so thin cut to get the correct/actual diameter reading on the micrometer measuring tool. The trick is not so much the micrometer dial, but in the end, knowing what the total stock removal from that point is - doing the math. You can "zero" either on touch off and thin cut, or after taking the thin cut, but once again, it is the stock removal and final cut ammount to size.

I have a practice I teach of doing stock removal to within .010 or .020 of the finish diameter, measuring often during the cutting process to assure your "comfort" that you are not going undersize, AND adjusting your cuts WITHOUT changing the micrometer dial often. Many lathes have friction mic dials, and when you do adjust them, there is always the possibility of moving ever so slightly in on the cut, or out on the cut (messing up backlash). Once you do that first sdjustment, you try like crazy NOT to have to adjust the dial, but rather work NOT so much wih the numbers on the dial, but the "number of lines" on the dial.

An exmple is this. You touch on .750 stock and wish to get a .400 diameter. Touch off tool on the spinning part, like you noted. Set zero. back off, move in .005 and cut. Your first diameter reads say .736 (radius dial), or .743 (diameter dial). You then do the calculaions, and you have .336 stock removal. On radius dial you divide this by 2, so your dial removal (remember, you will not re-set the dial to PREVENT but not necessarily eliminate, this is machining you know any potential additional human error) is .168. On a radius dial, I would have the student move in .158 with depth cuts (not all at once), checking now and then - say after .100 stock removal. On a diameter dial, move in to .318 (nose radius determines this, but this comes later with experience, metal, such on and in adnauseum, lets assume a pretty sharp nose radius HSS ground tool). I might go closer in some cases, but lets work with this.

You measure your part for the "final cut". For both examples, you get a diameter of .421. Opps, a little wear or something there....... This is where you can make your final stock removal determinations - BUT do not move the dial setting. On a radius dial, move in .0105, or ten and 1/2 LINES. You may want to mark your starting point in pencil, just for grins, giggles, and self assurance. On the diameter dial, move in .021 LINES, same pencil mark idea for your peace of mind.

Many may remove stock closer or futher, this was just an example. My idea is get the actual starting diameter, set zero when touching off or after the very first thin cut, Do stock removal calculations from the actual diameter of first cut, but DO NOT MOVE THE DIAL AGAIN - to PREVENT possible human error additions to the situation.

Probably clear as mud, and there are other methods out there, all are also very good, for machining is an ART, and art is individual. I learnt this method through my mean old german apprentice master.

Centering off a live center can be very good, but wear on older ones makes this harder. I have made little "posts" for my lathes out of soft steel that sit on a flat way and you set tool to flush with the posts. This is one way. the live center gets you very close in the ball park, but another method (if using carbide, be very careful) is the "scale' method. Put a scale between your part (if a cylinder) and the tool, touching the tool lightly to keep the scale between the part and tool tip. If the top of the scale is leaning to you, the tool is low. If leaning away from you, the tool tip is high. Do adjustments until the scale is dead perpendicular. This is also an "eyeball" method, much like the live center though, and can have error.

Another method AFTER eyeball centering. Face the part. If there is a square "nipple" on the part, the tool is low. Use that nipple point, and adjust the tool to center of the nipple - kind of like using the center of the nipple as a "bullseye" on a darts target. If there is a conical "nipple", you are high on center. You use the tool marks on the center of the cone to center to. If the pat is already faced, center to the tool marks (like a bullseye)to th part edge. This works quite well, and is my "experience" method that uses my experience from CNC work.

Jut wrote a book, looking forward to many more responses, will print them out, and will share with my students, for i love to show them the meny methods my 'friends" here mention to add these to their skills.

[This message has been edited by spope14 (edited 11-24-2004).]

Dave Opincarne
11-24-2004, 11:18 PM
nheng's method is mine too. It accounts for tool and cariage deflection. and you do need to clean up to a radius true to your spindle axis. Even the best job of centering a 4 jaw will show you cutting one side of the blueing before the other. It's the difference between measuring results instead of intent. That may be simplistic but it's a very important concept in producing accurate work.


11-24-2004, 11:42 PM
thanks spope14, you are just like the professor in college..

that man's a pro.

[This message has been edited by tattoomike68 (edited 11-24-2004).]

11-25-2004, 10:10 AM
WOW, thank you guys I am finding learning all this stuff to be amazing and fun. Thank you for all the help. Hopefully I will get good enough to get out of IT.


11-26-2004, 06:05 AM
I'm certainly not going to dispute any of the above methods as I use several of them also. The one concept that was most important (to me anyway) was actual results versus intent. There are several variables involved here. Setting the dial to zero is only a rough approximation for starting point reference. Carriage and way deflection are almost always present in every machine to a greater or lesser degree. A .150 cut made several times will produce a repeatable result as all of the slack in the machine has been taken up. Now try taking a .003 cut with the same tool a few times and you may find that the results aren't as repeatable. Tool wear or a geometry that worked well for roughing might not be good for finishing. This is not to set up points for discussion but to point out that the measuring tool should have a greater priority over the machine reference dial. Many times I have dialed in a cut only to find that I'm not getting what I dialed in. Is the tool cutting properly? Is there play in the ways? Has the piece simply cooled down and is now a different size? Measuring instruments are what are to be relied upon. Your instructor was probably just imparting a rough guide for starting the machining process. Stay safe.

Spin Doctor
11-26-2004, 10:07 AM
One thing I do I have enough stock on the material is to simply take a touch and set the coumpound short of zero and take a cut, check size and do the math for how much I have to go. Now comes the tricky part,keepning track of how much you have taken off so far as you work down towards size. Have seen guys get carried away with ripping the stock off they shot right past

11-26-2004, 10:45 AM
I've been machining with my South Bend for over twenty years. The dial isn't settable. I just do the math.

[This message has been edited by Evan (edited 11-26-2004).]

11-26-2004, 08:45 PM
Cigarette papers are .001" thick. Pinch it with the tool against the work, then back off until it just slides. Eyeball it and then come forward (dial into the work) to the same slidy point and set the dial one tick short of zero. Cig papers handy for eyeballing the 60* threading angle too. a small magnet or just a dab of oil will hold it to the bottom of your tool holder, and gives great reflection for getting that threading angle set up. What everybody else said too, this is just real handy. Not being a smoker, one pack of Zigzags has lasted me over 10 years.

11-26-2004, 09:52 PM
Amazing how great minds come up with similar conclusions and actions. http://bbs.homeshopmachinist.net//biggrin.gif

After reading this post I discovered I came up with some of the same conclusions and actions as many of you did, not that I an any kind of "Great Mind".

I had the same problems with dials as tattoomike68 but I ended up using nylon screws so I could reset, lock and not scrunch anything.

I took me a while to decipher the terminology when I began this processd, but finally understood what stiction (stickshun), deflection, temp changes, tool sharpness, slop in the screws, etc., was doing to make things go south.

I remember my dad talking about zigzag paper and watched him using it a few times when I was little boy. In this modern day, asking or buying Z's will get you a hard look. Now I use tape or whatever sheets of waste paper, envelopes, invoices, etc., I have laying around. Enco invoices are a nice shade of yellow and measure 0.004" thick and work great when setting up my mill.

But for the most part now, after finally figuring out that what the gunsmith said is exactly true, and learning some of the foibles of my machines, I usually just take a light cut first, measure, then continue.

I experimented with using a dial indicator, a dead stop and a micrometer stop, and moving the compound outward so all the slop was taken out of the screw, but then the tool would get suck in unless I had the compound locked down tight. Seemed like Murphy was alway messing with me. I use dead stops on both sides of the cross slide AND lock it down when I want to be certain. Talk about suspenders and a belt.

For the other beginners; You might make a tool height gauge out of scrap as a project, to make it easier to set your tool bits at the correct height. Mine is nothing more than a magnet base, a piece of 3/8" all thread, a couple of nuts and a piece of 5/8" keystock that I milled, drilled, tapped and turned to fit, then epoxied everything together. It has a prominent place stuck to the side of the headstock ready for use.

I also have several tools, basically permanently mounted, in toolholders that have the adjustments epoxied or locktited to the correct height so I can quickly swap around for whatever action I want to accomplish.

Many of these projects started as learning processes for me, which is what newbies need to help sort out the intricacies of machining. They are also reminders of my progress or in some cases, lack there off.

Time and practice are the best teachers, along with having forums to help when the darkness envelopes you.

Maybe these thoughts will help you.

11-26-2004, 10:37 PM
Someone mentioned cutting to "tenths", or .0001 increments. Actually very easy. This, however, works best with a tool with a smaller nose radius though. Set your compound rest to 2 degrees off of parallel of part centerline, or 88 degrees off perpendicular. each .001 on the compound rest = .0001 off of diameter. try it sometime. I use this regularly.

I think that the point of "actual results" and doing the math came up, and this was my point throughout. I too often come across machines where the dials are unsettable, or the readings from the dials are not exactly correct, as is the case on one of my machines. However, the final .020 are correct.

My best tool is a red or blak grease pencil, or my little "pentel" pencil used to mark my dials starting points, and finishing points (yes, i get interrupted quite a bit, so this is an absolute necessity).

Gauging, and the age old addage "measure twice, cut once". ESPECIALLY near the actual finish dimension.