trav-a-dial accuracy

Facts: I picked up a used trav-a-dial for my lathe. It came will all instructions so I know about the curvature on the wheel for calibrating it, the 'springy' mounting block and the springs are in good shape. I made a decently stout mount and went through the whole procedure for aligning it including making sure it's parallel to the direction of travel, I only got to about .0005 (half a thou) but sheesh.... Using my Chineese made 2 inch dial indicator the closest I can get it is -0.008 over 2 inchs.. that's 8 thou short on a 2 inch travel. To get this it's angled at about 20 deg. If I set it level it's -0.010 (10 thou) short on the 2 inch travel. FWIW I would have figured that tilting it would have made the error GREATER since it's effectivly making the wheel smaller. These numbers are ROCK solid, (well as rock solid as I wanna test) meaning I whipped the carrige back and forth for 5 minutes and it dialed in EXACTLY to

0 so it's no loosing or gainig anything.

Questions: Is this as good as they get? Is this about the ammount of difference you'd expect from level to 20 deg tilt (2 thou)? And yes for what I do -10 thou over 2 inchs is acceptable and that's where I'll leave it, easier to remember 5 thou in 1 inch to compensate( if I EVER need that much accuracy).

--.- Dave

Reply to
Dave August
Loading thread data ...

Dave,something is far wrong.I do not know what,but as you said yourself tilting it 20 deg only reduces the error 0.002 in 2".I fitted a lot of these to horizontal borers back in the seventies and was calibrating them to about 0.002 in 36" and while it took a lot of time it was attainable.Are you confident the unit is not damaged?If not that I would start from the beginning on the set up and as you made the mounting bracket I would put a clock on that while you`re setting the spring load on the mounting to ensure the bracket is not springing.Other than that I have no idea.Sorry. Mark.

Reply to
mark

The older used Model 6 TAD I have seems to be accurate to about +/- 0.001" over around 8" of travel on a Clausing lathe. What did you use to calibrate it for travel length? My manual refers to a "measuring standard" of either

6" or 24" length, but I used an aluminum rod turned to a shoulder at the HS end. The length from the end of the rod to the shoulder was measured with a height gage, which was in turn calibrated against a stack of gage blocks. The procedure seemed to work well, but took a couple of hours as I recall.

Mike

Reply to
Mike Henry

Reply to
William Noble

I usually don't have any great expectations when I buy used stuff. If you aren't willing to open the unit to inspect for worn or mis-adjusted parts, then your best choice might be to send it to SWI (or other instrument repair shop) for repair. I believe you should be able to expect very good accuracy from a properly calibrated unit.

The general operating principle of the T-a-Ds is that they imbed the pattern of the surface of the wheel into the smooth face of the machine's moving surface to create a tracking motion like that of a gear and pinion. That's the purpose of the pressure tension parts in the mounting block.

If your mounting method meets the manufacturer's recommended setup specifications, then the unit most likely needs some work and/or repair parts.

When you watch the pointer on a dial caliper move as you roll the thumbwheel, that's essentially the same operation.

You've seen that you can obtain repeatable results by changing the pitch of the wheel, so I would conclude that the unit needs cleaned, and possibly repaired and/or calibrated.

I've disassembled the electronic digital readout encoder versions of T-a-Ds, and found a small internal mechanical adjustment in them. The internal parts of the mechanical dial and pointer versions are probably very similar, except that the encoders have an encoder disk instead of a pointer. There are likely to be needle bearings for the wheel spindle/axle that might have crud in them, or possibly worn spots on the axle.

See Bill Noble's comments.

WB ............

Dave August wrote:

Reply to
Wild Bill

Thanks for the answer guys. I'll see about cleaning/rebuild.

Just for the record (in case anyone googles this thread in the future). The mount I made is plenty stout, 3/8 plate and it ain't moving. The T-A-D dosen't apear/feel gummed or sticky but it is obviously used. I'm using a 2 inch dial to calibrate it, and I haven't checked the accuracy of the dial(it IS an import)

--.- Dave

Reply to
Dave August

FWIW: A dial indicator's tolerence is likely less than the trav-a-dial. A dial indicator is more an indicator of variance than actual distance. Granted it's close enough for any work I've needed to perform. But I think you'd want to couple some gage blocks and use the DI to give you the zero reference. IOW any known reference - gauge block or micrometer gauge is better than any DI.

Reply to
John Hofstad-Parkhill

Greetings Dave, Trav-A-Dials have an anti-baclash gear set in them. This gear set is basically two gears on a common axle with a spring or springs connecting them. One gear is turned to tension the spring and then the gear set is meshed with another gear. Once meshed the gear set won't unwind. The gear set elimanates backlash because it is always making contact on both sides of the gear that contacts the anti-backlash gear set. Trav-A-Dials have a hole in the case that is under the knob that you turn to set zero. This hole is for a pin you insert to keep the anti-backlash gear wound up until it is meshed with another gear. There is a set screw that holds the knob in place. If you look at the knob you will see a hole for an allen wrench. If the setscrew isn't visible then turn the knob while holding the contact still and eventually the hole will line up with another. Removing this knob will reveal the hole. ERS

Reply to
Eric R Snow

your description is correct for the silver bodied trav-a-dials, but as I recall, the later ones use a different setup

Reply to
William Noble

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.