Work Holding

[JimDawson]

What I would do is build a plug that fits the center of your RT, and also fits the bore in your part. The RT does not have to be exactly centered under the spindle because the part center will be the RT center. Now you can cut the OD concentric to the ID.

Once you have done the OD, then you can easily find the center with your edge finder to lay out the hole pattern.\
.
.

 

[aeroHAWK]

king,

Yes, that is an edge finder, not a center finder. It is very useful - don't throw it out.

To find the center you need one of these:

And one of these:

Use it like this:

 

[RJSakowski]

RJSakowski I only need to find center to drill my hole pattern so I was talking about in the mill.

pineyfolks I'd thought about profiling it worth a band saw but really wasn't looking forward to that task. Just cutting the end off however might not be as bad. I think roughing the conners off in the mill would be easier.

aeroHAWK I don't have a indicator that could sweep that tight of a inside radius. At least I can't figure out a way to mount it so it would.

The tool I use zero my machine is probably actually called a edge finder. It works well on the outside edge but can't move as much running on the inside. It looks like this.
View attachment 104905

Sorry. I didn't realize that you had a hole pattern.

I used an edge finder similar to the one pictured all the time on my Tormach and mill/drill. I use the 1/2" diameter rather than the .200"; it seems to give a better trip point. I find that I can repeatably locate the center of an inside or outside diameter to +/-.0001". Inside diameters are more tricky but still doable. I recently posted a description of the technique that I use: (http://www.hobby-machinist.com/threads/finding-center.35687/#post-303075). It works best with a DRO but can also be used with dials if you know what your x and y axis backlash is and account for it. The key on using the edge finder is to advance slowly.

 

[kingmt01]

king,

Yes, that is an edge finder, not a center finder. It is very useful - don't throw it out.

To find the center you need one of these:
View attachment 104911
And one of these:
View attachment 104912
Use it like this:
View attachment 104914

I don't have the piece to clamp my indicators to the spindle.

ETA: Actually while I don't have that I could take my magnetic base apart & put it in a collet. It would be huge on my little mill but should work.

Sorry. I didn't realize that you had a hole pattern.

I used an edge finder similar to the one pictured all the time on my Tormach and mill/drill. I use the 1/2" diameter rather than the .200"; it seems to give a better trip point. I find that I can repeatably locate the center of an inside or outside diameter to +/-.0001". Inside diameters are more tricky but still doable. I recently posted a description of the technique that I use: (http://www.hobby-machinist.com/threads/finding-center.35687/#post-303075). It works best with a DRO but can also be used with dials if you know what your x and y axis backlash is and account for it. The key on using the edge finder is to advance slowly.

Funny you mention that. The one picture isn't the one I have it is just the same type. Using it today I just realized it has three different sizes on one shaft. I've been using the tiny end all along.

I do have a DRO but my scales aren't as accurate as your. Mine are only good to .0002" +/-.0001". Now if I get them flickering back & forth then I can guess at the other .0001" lol. They are actually only rated to the .001" +/-.0005" but I'm not using their head unit so I get better then that. Repeatability is only rated at +/- .001 per 10" I think but mine are always right on.

ETA: I like the rotatory table idea. That way I can do two operations in one setup.

 

[aeroHAWK]

king,

Just because you can read a DRO to four decimal places, it doesn't mean you have that accuracy. Don't count on any better than +/- .001 (if you are really lucky) with the machine you have. There are so many variables that go into a machine's accuracy, the most important being that you said it was from Harbor Freight. Having a DRO goes a long way toward making a silk purse out of a sow's ear, so at least it is a usable machine. But I wouldn't get too hung up on that fourth decimal place... it is meaningless.

For almost anything that you would be making (or most of us for that matter), +/- .003 is doing great!

 

[kingmt01]

Thanks. I realize this & have checked it with different indicators. These cheap scales are really accurate. I can only measure them up to three inches at .001" & to .120 at .0001" with the indicators I have. I don't waste a lot of time trying to get to within a 10ths but I do like keeping things close. Seems to make things easier on the next setup & especially going to another machine.

This particular piece is going to be a 100 hole dividing plate. It's first job will be to help me index a gear I'm making.

I don't know how accurate I have to be but figure being too accurate is batter then not accurate enough.

 

[RJSakowski]

king,

Just because you can read a DRO to four decimal places, it doesn't mean you have that accuracy. Don't count on any better than +/- .001 (if you are really lucky) with the machine you have. There are so many variables that go into a machine's accuracy, the most important being that you said it was from Harbor Freight. Having a DRO goes a long way toward making a silk purse out of a sow's ear, so at least it is a usable machine. But I wouldn't get too hung up on that fourth decimal place... it is meaningless.

For almost anything that you would be making (or most of us for that matter), +/- .003 is doing great!

You are right about accuracy. However, repeatability is a different issue. I drilled and reamed an array of dowel pin holes in my CNC table and machined a cylinder to fit. When I find the + and - extremes of the cylinder in both x and y axes, their positions a repeatable to within .0001".

This however says nothing about the linearity of the lead screw, backlash, flexing of the frame, temperature effects, etc. I know, for instance, that I have about a thousandth lost motion in all three axis, as verified by comparing a ten thousandths reading digital indicator to the CNC DRO.

While I can realistically machine to a thousandth or two at best, I still locate my positions to a ten thousandth. I do so because the final error is the sum of the individual errors and anything I can do to reduce that, I will. That is not to say that if I am just punching a hole in a piece of plate, that I will go to all that trouble. I will "eyeball indicate when appropriate.

 

[aeroHAWK]

Thanks. I realize this & have checked it with different indicators. These cheap scales are really accurate. I can only measure them up to three inches at .001" & to .120 at .0001" with the indicators I have. I don't waste a lot of time trying to get to within a 10ths but I do like keeping things close. Seems to make things easier on the next setup & especially going to another machine.

This particular piece is going to be a 100 hole dividing plate. It's first job will be to help me index a gear I'm making.

I don't know how accurate I have to be but figure being too accurate is batter then not accurate enough.

It's great to hear you are aware... it sounds like you have the right attitude. By working as close as you can within the limitations of your tools will still make a difference.

RJSakowski,

I'm sorry, but NO you CANNOT, as you say... "still locate my positions to a ten thousandth". Not with these machines (as a matter of fact, only with very, very few machines). But you can certainly create the "illusion" that you are.

I'm not sure what you are trying to be right about. You already state that you know the difference between accuracy and repeatability. But you then speak as though repeatability is accuracy.

Basically you can put a hole in a a place you're not too sure where it is, go and make other holes and then come back to within .0001" of the hole you're not really sure where it is.... So what?

 

[Highsider]

Your idea for a stepped shaft should work if you thread the end for a 3/4" nut. You shouldn't need a press fit. A slip, tight slip or light press, depending on your concentricity requirements should work fine. Use as large a shoulder as practical to maintain your perpendicularity. The torque from cutting will keep the nut tight. As Tom says, slow, steady, and lube.

In addition to this, for extra security, use a tablet paper washer between the shoulder of the holding fixture and the work. If you're using water based coolant, you may want to seal the paper area after clamping everything down.