PDA

View Full Version : Eyeball accuracy



Boucher
01-31-2010, 04:03 PM
As a kind of take off from the thread on visual accuracy I thought I would check out my set up. I was going to drill and tap some holes for setscrews. The part was too large to go into a 5C collet. So I put it into the chuck on my 6” Yuasa Tilting Indexer, which was just set in the mill vice. I located the edge with an edge finder and measured the diameter and moved over to center and zeroed the DRO.
http://i109.photobucket.com/albums/n50/boucherbyron/IMG_0068.jpg

I put a center drill in the drill chuck and just eyeballed it to center. The DRO indicated that I was off .018”

Next I used the center drill and 6”scale to locate the top as shown.
http://i109.photobucket.com/albums/n50/boucherbyron/IMG_0067.jpg
The DRO indicated that I was off .010

I realize that a single test is not very meaningful but it does kind of indicate rough accuracy. For non-critical set ups The scale and center drill are accurate enough and are significantly faster than using the edge finder.

lazlo
01-31-2010, 04:04 PM
That first picture is great copy for our sponsor :) Those GMT vises sure look nice!

Boucher
01-31-2010, 04:27 PM
That is also one of their keyless drill chucks. Both good tools.

MotorradMike
01-31-2010, 07:07 PM
Hi Boucher:

In your first pic it looks like you didn't turn the OD on your part where you are finding the edge.
Is it possible the edge finder picked up on an eccentricity or scratch?

Just asking, what do I know?


Mike

toolmaker76
01-31-2010, 07:34 PM
One thing that I noticed was the edge of the indexer in the vise. Is it a machined edge, or cast? If the edge is as it has been cast, it is not necessarily square to the diameter piece it holds.

If you pick up the edge and don't move in the X direction, it will stay somewhat centered, but I have noticed on the next picture that your X setting has changed. This may explain the results you are getting (changing center readings in the y).

You can either clamp the indexer to the table and indicate it, or you may want to square the edges of the indexer to the face in order to use it in a vise. Of course, that's probably not the point of your post, but I think you are probably eyeballing it more accurately than the reading shows!

Positioning over the scale is probably more than accurate enough for a set screw!

Robo
01-31-2010, 08:50 PM
That is also one of their keyless drill chucks. Both good tools.

How do you like that vise?

Carld
01-31-2010, 08:52 PM
I have never seen an edge finder like that one. Are you sure it is reliable? Does it repeat if you "0" it and back it off and test for "0" several times?

When in doubt I always check the edge a few times. Measure twice, cut once.

Machinist-Guide
01-31-2010, 09:09 PM
Now that I am in my 50's I have noticed my eye ball is not calibrated as well as it was in my 20's
:rolleyes: :o

Ken_Shea
01-31-2010, 09:50 PM
I have never seen an edge finder like that one. Are you sure it is reliable? Does it repeat if you "0" it and back it off and test for "0" several times?

When in doubt I always check the edge a few times. Measure twice, cut once.
Carl, that looks just like the unit here, a Herman Schmidt, probably not a better kick off type edge finder made.

Also, to throw another factor into the pot is the run-out of the chuck in test 2 and 3.

Boucher
01-31-2010, 09:52 PM
Carl: I think that you are yanking my chain. That is the Herman Schmidt that Bob Ford recommended in another thread. It is as good as he said it was. I thought you got one. If you didn’t you should!

Robo: The vice is very good.

Machinist-Guide: My friend you are on a slippery slope and it is all down hill from where you are. I keep installing more lights in the shop and wearing the magnifiers more.

Adressing the other replies:

This piece of stock was 1.5” 7075 T6 round bar that measured 1.504” OD. A 6” bar was held in the 6 Jaw chuck. The ID was drilled and bored first then the working end was machined to size in the same setup. The end was faced and both corners dressed. The Setscrew position and OAL were marked with the turning tool, before parting off. The part was then held by the machined surface to face the other end and dress the corners. The machined surface was held as shown in the chuck of the Indexer. The base of the Indexer is as cast. It was set on a couple of 5/8 wooden dowels for parallels to hold the Indexing keys above the vice bed. When I came to the house for lunch I took the test in the Thread (Mark 1 eyeball game). That caused me to consider how accurate I could visually locate the center drill. There was a burr marking the plane of the setscrew holes, there may have been other bumps and bruises on the factory mill surface. What I learned is that I can probably eyeball the TDC less than 1/32” and with the scale locate it within .010. As the expression goes, “That is good enough for the Girls that I go with.”

gwilson
02-01-2010, 12:31 AM
I think you should clamp the indexer directly to the machine's table. you have 3 sources of error the way you have it clamped in the vise : Vise's accuracy,the parallel's accuracy, and the accuracy of the indexer's bottom surface.

DaHui
02-01-2010, 04:09 AM
I like this thread! There seems to be an obsession with maximum accuracy at all times but isn't good enough "good enough?" If it needs to be eyeball close I say eyeball it! If it needs to be tenths close, use one of those Hermann-Schmidt finders.