Second video is the most interesting to me, even though I wouldn't use a co-axial indicator. Looks like it might be a You Tube Saturday for me. Just to prove or disprove my theory once and for all.

Tom

I personally don't think I'd look to YouTube to prove or disprove anything.
My guess is that it could easily be persuaded to pretend to do both these things simultaneously, for pretty much any theory you could think of.

If I didn't have a mill to try it on, a piece of paper to draw a couple of triangles on would quickly serve the purpose.

Maybe 'proof' means something different to different people.

- 0 - 0 - 0 - 0 - 0 - 0 - 0 -

I've got a little side bet on whether you'll climb down graciously, but I won't spoil things by saying which way I put my money! (or on what reasoning)

As time permits I've been working on this little project:

Demystifying Coaxial Indicators @ Metal Working At Home
Demystifying Coaxial Indicators, Part 2 @ Metal Working At Home

Hey Troup, want to lend me a hand down.

Tom

Nice vids dp. At first glance it does seem counterintuitive. Yet, when you think about it, it is easy to see why the readings stay the same for different length arms. You are, in essence, measuring the sine of the angle of tilt of your spindle axis with the surface plate.

Reminds me of a similar exercise in counterintuition:
Take a billiard ball and tie a piece of wire around it so that it is snug. Then try a second piece of wire and tie it off so that it is 1/16" longer than the first. The second (longer) wire loop will be quite loose. You can pull on a section of it and get a gap between wire and ball.

Take a bowling ball and try it. Same result.

Take a ball the size of the earth and try it. Same result.

In each instance the wire is only 1/16" longer yet even for a ball the size as the earth it will be just as loose as it was for the billiard ball.

As is typical with these kinds of problems... the math is left as an exercise to the reader

-DU-

I use an EZ Tram like Ray mentioned on page 2:



We have several mills and often reconfigure the heads, so it was a worthwhile investment. Usually it takes only a minute or two to tram the head to less than 0.001", so it also makes quick work of checking tram before starting an important job.

dual indicator tramming, you zero both indicators to same spot by rotating quill. Some use a parrallel or gage block for the spot to zero so indicators are not touching anything else. They are then set perfectly to each other, it does not matter if collet is eccentric in fact gage can easily be held in a drill chuck. Digital indicators are easier to zero than dial indicators but either work.

point is you do not need to rotate much. 180 degree to calibrate and tram that direct and rotate 90 to tram other direction. If you do not trust dual indicator calibration then for final check use only one dial indicator and sweep it one last time.

10X faster and more accurate. As you tightened screws holding turret/tram you want the dual indicators to see if BOTH maintain same reading. It saves rotating single indicator around after final tightening.

Older machinist usually hate dual indicators but I say if you need more than 60 seconds to tram the mill you are using the wrong tools. I also find most Bridgeport type turret knee mills need tram checking every day if you want to maintain 0.001" per 5" tram or closer. Vibration and heavy milling moves the turret especially if an end mill is broken for any reason. I have never seen a Bridgeport maintain tram closer than 0.001" per 5" for very long.

Is it feasible to make one of these?

SCOTTIE (hereafter known a "tightwad")

Not all machines are as nice as the original BPort.
If the pivot is not proper, the head will not consistently swivel on the same spot...
So, you zero the indicator on the left, swing it to the right, and it is off by .004"
Fine, you think, move the head by .002".... Swing it back to the left. Now you are out more than you were when you started..
Next time you try it, the pivot point is someplace else..
Splitting the indicator readings will not work, the side to side keeps changing..

Finally you learn to very, very "ritualistic" about how you you tighten and loosen the bolts that hold the head to the knuckle.

The dual head indicator solve the problem of the head not swiveling consistently on the knuckle.

If you never had the "privilege" of tramming in such a beast, it would be difficult to imagine the need for the Tram Indicator.

I've often thought this, for once I thought I'd raise it for what it's worth.

Any form of 'tramming' or 'clocking up' which relies on the operator moving machine elements to generate an error signal is always going to take longer and require more skill, focus and mental horsepower than a form which gives an error signal, instant by instant (ie in "real time".)
For the purposes of this post I'm going to call this first (non-instant feedback) situation "Type A".
It requires iterating between adjustment, (using a mixture of inference and intuition to decide how far to move the item being adjusted) and the measurement operation. The hoped-for result is convergence towards zero error.

Any sweeping operation where the machine axes or spindle are moved is clearly Type A

I'm considerably more relaxed about Type A when I have a rapid traverse option, and a quiet shop (or a machine I'm very familiar with. I've managed to get this far through life without crunching an indicator and I'd like to keep it that way). Or, in the case of a spindle, either a sensitive and responsive jog option, a nicely operating clutch, or a very free neutral.

A dual indicator setup (whether as discussed here, or as used for squaring vices and workpieces, is clearly Type B
So, or course, are precision machinist's squares (say, used vertically against the side of the quill or horizontally against the side of the work) and precision levels (in the latter case, if you're correcting twist, or levelling in two planes at once, you really want two).

Working with Type B is always quicker, more pleasant, and more elegant. It may sometimes not yield quite the same accuracy, but most people believe you can sometimes have too much of a good thing, even where accuracy is concerned.
Some people have missed the fact that a 2-indicator tramming aid can be qualified (and if necessary recalibrated) simply by turning it 180 deg, on the table, and checking that the error message reverses exactly.

Workpieces on big long-bed mills which don't have a suitable true mounting surface available are often set up using one or more precision levels, rather than trundling a huge head large distances hither and yon, holding a tiny indicator. The jacking screws might be in the next county, for that matter. There's no substitute for being able to watch the bubble (or the digits) move towards zero as you turn (or pump) the jacks. Particularly when one jack influences both x and y levels. Type A is not in the hunt here.
The table tops are almost always set extremely accurately level to assist with this.

However a depth mic, used to set the two ends of a workpiece parallel to a table reference edge, I would argue is more Type A than Type B

Which brings us to hybrids, like the coaxial indicator. Or a single precision level, used to correct twist. It's not necessary to move an axis or turn a machine spindle, but it's not giving a static, self-sufficient error measure in real time.
These could be called Type C.

I would suggest that a swivel vice, provided the pivot is nicely made in the context of the accuracy required, is more Type C than Type A, because having zeroed an indicator at the midpoint, it is then possible to score a "home in one" by moving to one end and swivelling to zero. IN THEORY, a mill head is the same deal, at least for sideways tilt. In practice the head generally moves on unclamping and reclamping, because the deadweight supplies a serious tilting moment which is not present for a vice on a horizontal table.

My favourite? Type D.
Hard stops, taper dowel pins, keys, Newbould stacks, detents and setting jigs, (I made one for squaring my machine vice, inspired by a post last August from forum member "smallplanes" Leics, UK ... thanks, Dave! ):
Basically I'd use Type D to designate anything where no error message is needed, because the machinery just gets lined up by inexorable forces linked with immovable objects.
(So a machinists' square doesn't comply, unless you clamp or weld it in place)

Backed up when necessary with good old "Type A", arguably the most accurate and versatile method of all, for the rare cases where the accuracy envelope has to be pushed, and occasionally to re-qualify the hard stops.

I don't like vice keys because big vices are heavy and I don't like fussing about with keeping the crevices around the keys clean, making sure the keys don't ding the Tee slots, and removing and replacing them whenever I want to deviate from the 90 deg orientation.

Another thing I don't much enjoy is hearing people put down anyone who tries to come up with a non-Type A solution. It's like a zero-sum game take on things: "If what he says is right, what I'm doing must be wrong, and that can't be right..." or alternatively/additionally, a "there can only be one correct solution to each of the problems of the world" mindset.

Although it can sometimes be pretty funny, eh, Ray? -- aka "Home Shop Guy" ;-)


(the post from smallplanes was as follows:
<<If the T Slots are in tram (one time check) then a U shaped piece of steel clamped in the vice (So the legs of the u stick down and into the T slots), pushed up against the t slot will do an instant align, and is very simple.
Dave>>

If anyone's interested, I think i have a photo of my slightly different implementation somewhere, send me a PM)

Troup,

And then there is Type E, or is it Type J (I lost track) on my Rockwell.

As follows: There is no "nod." Tilt in the Y-Z plane was fixed when the machine was made. The cast iron head was cast on to the steel ram. It can be shimmed if way out but it is less than 0.001" over the width of the table, 6-1/2".

For the X-Z plane it is a bit different. The head is mounted on a cylindrical ram and clamped by two bolts that clamp the split bores at the top of the mill column. There is a worm gear type tilt adjuster with degree marks but that is not reliable for tram. If one sets the tram dead nuts and then tightens the clamp bolts, to paraphrase Galileo, "It moves." The head rotates as the clamps are tightened.

What I have to do is tighten the clamps to 5 ft-lbs set the head to approx in tram and then tighten to 45 ft-lbs in two steps of 20. Not how much it has moved. Loosen back to 5 ft-lbs. Set the tram again PLUS the "movement" (or minus, depending on which side of the movement the indicator is on) and then tighten the clamps to 45 ft-lbs again. Then re-check.

Usually I nail it within the first sequence. However, it can vary for some reason. Sometimes the movement as it is tightened varies a bit. Enough so that I have to go through two or more sequences. Since I spend far more time torquing bolts and noting movement the tram indicators being discussed off no advantage over the old beam-to-spindle-with-indicator-on-the-end. Though I haven't tried any of the dual head products mentioned above I have mounted two indicators on either end of a beam and it was no help. The head still moves and it has to be compensated for.

-DU-

I hear you, David....

I always wondered why the machine operator in this photo was looking so .... wistful !

Attached Thumbnails

 

I think it's the shiny metal.

Some girls like that.

- Leigh

Correct me if I'm wrong, but I was always under the impression that using Co-Ax Indicators depended on not only the spindle bearings (same as in the case of using a spindle-mounted indicator set-up like an Indicol), but the accuracy of the collet and the alignment of the Co-Ax shank to same.

Both of these last factors could/would introduce error as it is imperative the bottom surface or plane of the Co-Ax body be perpendicular to the rotating spindle axis for the relative readings to be correct as the arm is swept.

It might be an interesting experiment to use a Co-Ax to set tram, then check it by using an Indicol setup; both of them using the same sweep diameter.

Older machinist usually hate dual indicators but I say if you need more than 60 seconds to tram the mill you are using the wrong tools. I also find most Bridgeport type turret knee mills need tram checking every day if you want to maintain 0.001" per 5" tram or closer. Vibration and heavy milling moves the turret especially if an end mill is broken for any reason. I have never seen a Bridgeport maintain tram closer than 0.001" per 5" for very long.

I must be one of those old machinists.....have always used an Indicol.

Worked in a shop with 13 Bridgeports. Then later with Kondias and DoAlls which were BP clones.

I'd like to see a head BP trammed in less than 60 seconds start to finish; don't care what method is used!!

I got a bit of interest in the vice squaring jig, but could only find one photo (camera's down right now) - unfortunately it was badly overexposed, and there was a limit to how successfully I could tweak it.

It's just two legs TIG welded (aiming for lowest heat consistent with moderate strength) to a flat bar previously ground flat on front face. Shallow channels premachined on back face to define inboard edges of weld zone (aging eyes). Space between welds needs to be wider than vice.

Resulting distortion of flat face (minimal) corrected by a few wipes on wet n dry on a piece of float glass.

Set up on front face on surface grinder with legs upwards, rear face dusted parallel to front face between legs. Contact face of legs dusted at same setting as each other.

Lacking access to a big enough grinder/wheel, this could all be done with a bit of care (re distortion and clamping) by milling.

Would pay to make it longer for clamping, and trim after, perhaps.

Mine's not hardened, so I make sure to wipe the jaws and the jig carefully before clamping it in place. This just about doubles the time it takes to get the vice square, but I can live with the extra seconds. Don't mind cleaning things I made....

This generally gets the vice close enough for anything not involving arming devices for nuclear warheads (which doesn't bother us woosy pussies)

Attached Thumbnails

 

I dunno Troup. What if a BIG chip got between the jaw and bar? What if the weld broke? What if there was a big gouge in the T slot? What if ???. Actually, I think it's quite clever and easy to make. Have I been one upped?

Ray

Hi Troup,

Neat thingy.

- Leigh

This can be a problem - the trick is to see if it is a problem. In my case it's not but I haven't abused my machine yet

Here's a good video from Tubal Cain showing the 'old fashioned' way of tramming a BP head - YouTube - BRIDGEPORT MILL Tramming or Indicating the Head to Zero

Corm

Corm

I hadn't even watched the clip (dial up, y'understand) but I couldn't help my eye being drawn to this comment at the foot of the youTube page
(I blame my parents, they had me proof reading stuff they were writing before the age of ten, and my eye behaves like a tongue searching out chipped teeth)

<<Great video. Very informative. It's splitting more hairs, but if it's important to anyone, (those in the gage/fixture trade) lowering the quill all the way down will exaggerate your reading. Meaning, if your indicator is showing out of square about .001", with the quill all the way down you'll be showing two or three thousandths out. I notice in the video that he had the quill down, but did not mention it's importance.

My trade is all about splitting hairs. Sorry!>>

The funniest part is, the guy who made the video bought it, and thanked him for the enlightenment!

(Obviously not "The" Tubal Cain, aka Tom D. Walshaw, who would not have entertained this for more than a nano-second were it not for the unfortunate detail of being dead)

Like I said earlier in the thread, you can 'prove' any contention on youTube, regardless of the merits ! Even when you're not trying to...

(I'd log in and do my civic duty to put the opposite view, but I'm not so dutiful as to register with youTube for that sole purpose. Any takers?)