English is my first language so bare with me…

I have a couple of used .0005” and .0001” dial indicators, Starrett and Mititoyo respectively.

On a quality magnetic base I detect some “creep” or lag. In other words, when I am taking a reading by rotating the chuck manually and stop, the indicator reading slightly decreases as if the work is getting smaller.

The work has been cleaned thoroughly so it is not stopping and “wading” into cutting oil or the like.

If this is enough to make sense out of what I am trying to relay, what is the scoop, what’s going on here?

Do I have worn out or damaged dial indicators or what?

Those are very sensitive, are they on top dead center? How fine is the finish of the part?

Friction between a moving part your indicator point can cause what you are seeing.Poor rough finishes will accentuate what you are seeing.Sometimes I tap the indicator holder LIGHTLY with a pencil to help settle it out. This can also be an indication of a worn stem or side bearings depending on the type of indicator.

No, not TDC but within 10-15 degrees.

Finish is smooth (Hot rolled steel) that was turned with great finish followed by 220 grit. I don't see how the stem could be "riding the waves" when there is really a smooth surface.

I need to look at the stem under magnification...there might be something missing...

Your method and/or tool for holding the indicator is far more suspect. I think you may have a moving clamp or something.

I'm going to take a different tack on this. I've quite often seen this, and tracked it down to clearance in the spindle bearings.

This is especially evident if the probe is on top of the workpiece, because the spindle drops in the first half second or so after stopping rotation, as the hydrodynamic wedge of lubricant which has moved the spindle closer to the nominal centre height oozes away. A small movement is inevitable unless the bearings are preloaded.

What sort of spindle bearings do you have? Plain cylindrical, plain with front bearing tapered, taper roller, angular contact ball, or cylindrical roller on a tapered journal?

In all but the first and last of these, radial play in the bearings will be accompanied by endfloat, either of which, if you can establish and report a figure, would act as tentative confirmation of this theory.

check for side play on your indicator point. I've been using my Mititoyo .0001 indicator daily for about 15 years now, and the pivot screw has worked loose several times with the same results. I can always tell right away when I reverse direction or stop rotating and the needle jumps to a different position as if there's backlash. If this is the case, be careful not to over tighten it.
Terry

I'm with Troup on this one. I've seen exactly the same thing many times, but using the same indicator on a small lathe I have with new preloaded angular contacts it doesn't happen.

Here is something I need to triple check. The movement is duplicate so I had ruled it out. Won't hurt to re-verify.

I haven't got a clue...it is a 1986 South Bend 10L (CL81887RB s/n 02894R). But something appeared to be floating momentarily.

As a general rule, these indicators are pretty much weevil-proof unless there is visual physical damage (bent/damaged/worn indicator point, broken/dinged case/crystal, etc), am I somewhat correct?

Check the stylsis. Sometimes they come loose and your readings will get "soft" and "sloppy".
That is what that little tiny wrench is for.
Regards Walt...

You don't need to know the bearing type to diagnose the problem, just to fix it. If there is significant play (you don't say how much 'creep' you are seeing), the problem you describe is the least of your worries and you need to fix the play before you can get decent finishes.

If the 'creep' is only a few tenths, I'd say the problem is that you are using indicators which are too sensitive for the style of lathe you have.

You can eliminate the indicator as the source of the creep by putting the probe on the side of the workpiece, advancing the cross-slide a small distance, and stopping abruptly.
Check for under- or overshoot followed by creep. I doubt you'll find that, but it's worth eliminating.

You'll have to eliminate sources of slop one by one. First two are easy to check:

- Deflection in the indicator holder
- indicator point not fully seated in its threads

My guess is you've already checked those. Another common source of error, and my best guess of your problem:

- Side loading on the indicator as you're rotating the part causes a bit of binding. Possibly aggravated by slightly gummed up works. When you stop, the (perhaps somewhat weak) spring gradually overcomes the previous side loading, gumminess, and gear lash to add a few tenths to the reading. Quick solution, as noted above, is to be sure your indicator has settled, perhaps by tapping. Better solution is to refurbish the indicator.

The spindle bearing problem suggested above may also contribute.

I have to deal with this all the time on my South Bend. Yours probably also has plain bronze bearings, or some of the smaller ones do not even have bearings, but run in the cast iron bore. With the indicator resting on top, pry under the spindle with a stick. A couple of thousandths clearance is normal. If you get much more than that, you have bigger problems than an indicator. A common problem with these lathes is that people treat them as if they had sealed bearings and don't oil them. I have seen a 10" south Bend that I could move the chuck up and down with a pronounced clunk. Even if, like mine, it has been always lubricated, the spindle will float up almost .001" as I turn it because the oil wiped around the bearing will move the spindle to the center of the bearing. To indicate a part accurately, I turn a little and stop, repeating around the part, taking the readings after it settles each time.

Bill

Who, where and how much? Is is cost effective?

A late South Bend has a bronze plain split bearing. It is supposed to be adjusted to around .0005 - .0007 radial clearance (cold). Find a manual for it and read the procedure for adjusting the clearance. It isn't difficult, but there are a couple of things not obvious from observation of the housing (there is a wedge in the bearing split on the upper shell which must be loosened first, or you chance wrecking the bearing).

By that I assume you mean that you will get .001" to .0014" movement total when you pry the spindle up. Mine was closer to .002" total from the factory, but it is a 14 1/2" which is the same spindle as the more common 16". In fact, I have never figured out why they made that size. With the larger journal, you would expect to have a larger clearance. If you don't have a manual, I can scan the page from mine.

Bill

Long Island Indicator. Google their site for info and recommendations on which indicators are worth fixing. If it's a cheap indicator or really beat -- not worth doing. You can often find good used indicators on Ebay, though it's a craps shoot.

If the indicator is just sticky you can sometimes carefully clean it yourself, then add just a tiny drop of watchmaker's oil on the internal pivots. I believe there have been old threads on cleaning dial indicators. In general, you don't want much oil on these -- keep the rack, barrel, etc. clean.