Sorry Chris Mate what is this all about?

It may help if you tell us the machine make and model, or supply a photo, as there are two basic layouts of small milling-machine; and to show how your DRO is fitted.

– One keeps the table at fixed height, on the machine's plinth, and raises / lowers the head.

– The other is more like the industrial pattern with the table on a "knee" that can be raised and lowered.

Both have a quill as well, but that is normally for drilling, not high-accuracy depth-milling.

.

The DRO's vertical axis may measure the movement of the quill, but usually is for the knee or the head..

Not the spindle.

This does depend on machine type and how the DRO is installed, of course. For example, my mill is a Myford VMC and I have fitted the Z-reader to measure the knee travel.

Backlash is less of a problem with the Z-axis because the moving unit's weight tends to keep the same thread-flank in permanent engagement, but in any case a normal DRO's long and cross-travel detectors measure the table not the lead-screw.

'

Other than that I am really rather puzzled what you are asking, but To Mill to Depth accurately, using a DRO or handwheel dial:

– If possible work with the quill fully raised, for maximum rigidity.

– Use the head or knee control to approach the cutter tip to just touch the work. You can use a witness cut but a safer way, more likely to err on the side of shallow rather than over-depth, is to "glue" a slip of thin paper to the surface with oil, and bring the cutter down to just remove that.

– Set the DRO's 'Z' reading or the elevating hand-wheel dial to 0.

Now cut down from that 0 setting.

Easy peasy. Hold a ball bearing mounted on an arbor in the dpindle. The inner race will rotate eith the spindle. The outer race can be held still by the pressure from a dial indicator plunger which can measure running spindle movement. Compare that with your DRO and you can see any error between the two and make allowances.

But worrying about such precision on a milling machine is a waste of time. They are not a super precision machine. If you can get within a thou or two thats all you need.

Apologies if I've misunderstood what Chris is saying.

I come at this another way. I don't bother with the absolute position of the table (X,Y) or of the spindle (Z). Instead, I measure relative to the cutting edges of the tool.

Typically,

• I plonk a vice on the table without worrying about it's exact position, apart from making sure the jaws are aligned parallel with table. This is done with a DTI held in the spindle.
• I position the job in the vice, again not worrying about its exact position. The vice ensures alignment, and height is fixed by parallels or by vice features.
• A fixed point on the job is selected as a datum, from which the DRO is set to 0,0,0. Corners are convenient when the job is a flat plate:
  • With an edge-finder in the spindle, the table is positioned to move the corner under the vertical axis and the DRO set to X=0 and Y=0. Thereafter all X,Y movements measured by the DRO take place relative to this reference. (Until it's convenient to work from a different datum on a complicated workpiece.)
  • Z (Spindle) is different because changing the cutter on a manual machine requires the Z DRO to be reset again from a reference datum. I usually set Z=0 by tacking a piece of hand-rolled cigarette paper to the job at the datum with a dab of light oil, spinning the cutter slowly, and gently lowering the head/quill until it just whips the paper off. As cigarette paper is less than 0.02mm / 1thou thick, this is usually 'good-enough' to set the DRO Z=0. All Z movements with this particular cutter are then done relative to this DRO setting.

The process uses the DRO to move cutters relative to fixed references on the job. Doesn't matter what the quill and head do provided the operator knows where the cutter is relative to a datum, as understood by the DRO. If the cutter or table moves because of backlash, the DRO is still correct and the operator can trust it to reposition the job under the cutter.

The DRO can also be reset on a reference datum at any time to make sure accuracy is maintained: errors can accumulate! On a simple job, the DRO might be set once but anything remotely complicated will have several datums and much shifting between them.

In short, always locate the job under the spindle relative to a fixed point on the job, as touched by the tool, resetting X,Y and Z to new zeros as necessary.

Dave

Blimey I’m still trying to work out what the question is, but DRO——how accurate? I can answer that, good enough for industry to use for the past 40 years so plenty good enough for chasing microns in your garden shed.

Put the DRO on the quill like the rest of the world and read the quill movement. If the spindle moves any differently in the vertical direction to the quill at all then the problem isn't with the DRO placement it's with the spindle bearings.

No idea what you are talking about now. Pictures needed to explain your problem. I am totally unclear on where you have the two dial indicators mounted and where they are measuring.

Are you having problems with your mill not producing accurate work? If not, then you are probably overthinking the problem.

Best to have your machinist friend look at it in person. It seems whatever your problem is, it is not commonly known about among the mill owners on here or one would have jumped in with a solution by now.

This all sounds like an excuse for not getting any work done to me.

The only time I've put a DTI near either of my lathes is to dial a part into the 4 jaw chuck.

What is the OP intending to make that justifies this sort of inspection?

Chris, thanks for posting the diagram. That makes the problem clearer now. But the solution could be more tricky!

It seems strange, as you know, that the quill can move downwards but the spindle does not.

I think Martin is right in the above post, the first think to check would be the spindle bearings. The spindle bearings should be holding both quill and spindle solidly so they move up and down together with no variation. The only movement should be the rotation of the spindle, not up and down movement.

It may be that if there is slack in the end float of the bearings, that the spindle is held in place by where it passes through the splines in the pulley at the top end that drives the spindle. Depending on how your mill works. I am not familiar with your model machine. But if there is slack in the bearings and the spindle is a tightish fit in the splined pulley that drives it, it could be possible the quill moves a small amount but the spindle does not.

I am having the same problem with an HM50 mill from Hare and Forbes or Machinery house.. The clearance between the cast housing bore and the diameter of the spindle cartridge assembly. Its not too hard to remove the spindle. In my one, there is about 0.05 mm to 0.06mm between the housing and the casting. It should be less than 0.01mm. I have taking apart my assembly and getting a company to hard chrome the outside with a build up of 0.05mm. I will lap it if needed to clean it up and will then make a lap for the casting bores if needed to be loosened off any. It looks like there is no wear on the spindle body and none to be seen or measured on the main castings. It seems that this is a very common problem with cheaper mills. A bridgeport mill spindle is made with around 5um of clearance and are matched fit assemblies. In my spindle there was casting sand and fine cast iron powder in there wearing out the bottom bearing. In mine is a taper roller bearing. Some have a sealed ball bearing. To check out my theory, I spray painted the outside and re assembled the spindle sleeve. At it lowest point it only had 0.02mm of movement. Before taking it apart , it had 0.12mm of movement. So flash chrome it will be. I am also going to check the concentricity of the bearing housings to outer housing before it goes for plating.

If they are not right I will have to think of the plan to correct them.