PDA

View Full Version : Lathe Readout Idea, Shoot holes in it.



Spin Doctor
12-04-2016, 11:35 AM
In general I don't like DROs on lathes. Especially on the Z axis. X is fine but on the Z sucks in my mind especially when coming up to a zero or end point on a length. I find that a lot of operators feel the same. Hence why some people prefer Trav-A-Dials. But what if you want to build your own mechanical readout. Imagine a "clock face" with 100 divisions instead of 60 with a "minute hand" to read the thousandths. The "hour hand" geared to run at 10:1 reads the inches. The dial assembly is driven by a cable connected to the saddle that runs over idler pulleys to a master drum with a diameter of 3.1831 (10/pi). This cable then continues on to either another idler pulley with a weight attached to keep a constant tension or a set of compound pulleys carrying the tension weight to limit the vertical travel of the end of the cable. I have tried all kinds of goggle search combos and the only thing that comes up is either displays for cable playout in digital or the wheel type "tape" measures you push on a stick. In some ways I think an idea like this, provided it works, could be useful on a smaller lathe where you quite frankly do not have room for a Trav-A-Dial. And as I stated I don't like readouts. Right now this is just some sketches on paper and idea ii have had kicking around for a while. The biggest problem I see is dirt/contamination in the system. Aside from getting the master drum the right size. I'll do a rough sketch and post it farther along.

CalM
12-04-2016, 11:45 AM
Spin doctor

The set up you are describing has been promoted some years ago on the web. Somewhere.
As I recall, the cable drum turned a Potentiometer with the display through a BASIC programmed PC
(HERE! http://www.lindsayengraving.com/other_interests/dro.html)

At any rate, it is the hardware that is the parallel to what you wish to do.
These days, one of those "postage stamp" processors would replace the PC, But you would invest the most effort in producing the display.

A trav-a-dial turned out to be the better option for me, Everything in one nice package!

Spin Doctor
12-04-2016, 11:53 AM
I think it would make a neat project. I really like Trav-A-Dials but as I said on a small lathe (10x27) they can be a little too much. Plus a nice big "clock" face about 10 to 12 inches (250 to 300mm) in diameter would be a lot easier for us old farts to see. I'm working on the sketch now.

Baz
12-04-2016, 01:08 PM
Is the aim to make something mechanical or to get a dial type readout that is easier than numbers to see the approach to the end point?
A DTI on a bed stop seems the easiest.
These people (http://bwelectronics.co.uk/)do a small wire pull DRO. Maybe an interface to a PDA to create a dial picture is the modern way to go.

The Artful Bodger
12-04-2016, 01:52 PM
Spin Doctor, I am sure you could make something useful following your concept but I do wonder how fast your 'thousands' dial would have to spin when doing a rapid saddle traverse.



This was my 'poor man's DRO'..

http://farm4.static.flickr.com/3486/3294903470_996ce149f4.jpg?v=0

http://madmodder.net/index.php/topic,781.msg5030.html#msg5030

Flexible cables for X and Y. It did work reasonably well but was taken off when I moved workshop and has not been put back on because I never got around to doing the 'improvements' I had in mind!

Instead I fitted this for the cross slide with the scale protected inside the slide.
https://c7.staticflickr.com/8/7320/8994708326_ec8632fe5d_z.jpg

LKeithR
12-04-2016, 02:48 PM
In general I don't like DROs on lathes. Especially on the Z axis. X is fine but on the Z sucks in my mind...

I'm not sure what you're missing but your "dislike" of DROs doesn't make sense. I've run a lot of lathes over the years and I can tell you that the first time I ran one with a DRO it was like a revelation. Absolutely the best productivity tool you could put on one of those machines. I know there's something to be said for "old school" but the DRO on my lathe allows me to run it faster, more accurately and more consistently than I ever could with just the dials. And just so you know, I'm 70 years old so I'm not some young 'egghead" kid who's never done it the hard way...

J Tiers
12-04-2016, 02:51 PM
Anything (nearly) is possible....

What if you used a regular indicator for the dial?

The pulley wheel on a threaded rod that moved the indicator plunger as it turned. With the right thread and wheel size, you might get a useful ratio so that the indicator would read 10:1 reduction and show up a 0.100" on the dial as an inch of actual travel.

or you could "get over it" and put on a DRO..... Even the i-gaging stuff would probably be more accurate than the pulley system unless you were extremely careful.

cameron
12-04-2016, 03:06 PM
Well, JTiers posted #7, so that's 6 guys who obviously know where the Z axis is on a lathe, and why somebody would want an indicator on it. Would someone please re-orient me?

danlb
12-04-2016, 03:12 PM
It's in relation to the axis of the spindle. In essence, the movement of the saddle



https://www.google.com/search?q=lathe+Z+axis

Mcgyver
12-04-2016, 04:56 PM
Well, JTiers posted #7, so that's 6 guys who obviously know where the Z axis is on a lathe, and why somebody would want an indicator on it. Would someone please re-orient me?

Z is always the same axis as the spindle.....I've never put an an indicator on it on the lathe as I have micrometer depth stops, but photos of the them suggest they would be used in the same fashion, ie to where to stop along the Z

J Tiers
12-04-2016, 05:25 PM
I made my depth stop to take either an indicator or a micrometer stop.

So far I have never put an indicator on it except to verify it fits....... but I could...... I just never have felt the need. Yet.

boslab
12-04-2016, 06:26 PM
I tend to stick a mag base and clock on when I need it myself, mostly the stuff I do at home is to the class of "rough" to "not too bad"
Mark

Paul Alciatore
12-04-2016, 06:44 PM
Unless I am mistaken, many of us have indicators with mounts for our lathe. The problem there is they are limited to 1", 2"' or perhaps 4". I have never seen one that has a greater travel than that. Yes, you could gear them down, but to keep the accuracy level down to a thousandth, you would need a tenths reading indicator or something like that if you are going to read over a distance of as little as 20 inches. I don't know about you, but I don't find that attractive.

I have given some thought to some other, inexpensive mechanisms for reading these distances. I did this when I was first starting to get serious about the hobby so I was definitely interested in things that did not cost hundreds of dollars. I do not want to discourage anyone from trying anything, but I found reasons to avoid many approaches. Here are some of the thoughts I had.

A wire could be used to turn a dial and with appropriate gears, there could be a turns counting mechanism. So a basic 0.100" dial with 100 divisions could indicate thousandths and you could probably even read down to 1/4 or 1/5 of a thousandth if that dial was large enough. It could be implemented with the dial fixed to either the ways or to the carriage, either would work. But I see two problem areas. First, the diameter of the wheel that the wire runs on would need an EFFECTIVE diameter of 0.100" / pi. That translates to 0.0318+". That is awful small and more gearing or more divisions on the scale would be needed. At 200 or 500 divisions that scale will get awful big if you want to read it without a magnifier. So gearing it would be. Perhaps a 1/10 gear ratio so your wheel would now need an EFFECTIVE diameter of 0.3183+". That's better, but still quite small. I would want at least double that which would require a 1/20 gear ratio for our 100 division, thousandths reading dial. OK, that is essentially what is done in a DI so it can't be that hard, can it? Well, do the math on a gear that is, say, 1/4" in diameter and must have an accuracy on the location of all points on the faces of the tooth flanks that is calculated as follows:

Assume 1" scale diameter and 100 divisions
Each divisions = 3.141" / 100 = 0.0314"
So +/- half that amount or +/- 0.0157" is the tolerance on the needle position for a "correct" reading
Now we are gearing down 1/20 so that number is divided by 20 and we get 0.000785".
But we are not finished, I said a gear (pitch) diameter of 1/4" not 1" so again we divide by 4
And we get a face tolerance of +/-0.000196"

That's a tolerance of about two tenths. Now, that is not an absolute tolerance. If all the teeth were uniformly small or large by that amount, it would not matter. But they must ALL be within that value of some theoretical position. They must be uniform to that tolerance. That is not an impossible job, but I doubt that I could meet it in my shop.

This is why good DIs cost a fair amount and inexpensive ones have larger errors in their linearity. The companies that make DIs can afford to set up a line where gears with this kind of tolerance can be made all day and for a reasonable price. For a home shop or even a run-of-the-mill commercial shop it can be quite a challenge. And even if you did make one that accurate, just how would you verify that? What do you have that will measure gear teeth positions down to tenths? I would have to make or buy a rather special instrument.

The next problem I see with a wire on a wheel is the EFFECTIVE diameter of that wheel and wire combination. You may notice I have been capitalizing the word effective. That is because a wire wrapped around a wheel/drum will not actually turn that wheel in the exact amount you would expect from a measurement of the wheel itself. The part of that wire which is closest to the wheel will be compressed so it becomes shorter. And the part of it that is opposite the wheel will stretch and becomes longer. As an example, with a 1" diameter wheel, the part of a 0.050" diameter wire that is against the wheel will be 3.1416" long while the outer part of that wire will be 3.4558" long. That is a difference of 0.3142" which is not trivial. And there is no guarantee that the half way point will be at the the EFFECTIVE diameter. The wire may compress more easily than it expands or vice-versa. Worse, if you measure and get the actual OD of the wheel exactly right, the wire may age and change it's characteristics. My feeling is that maintaining an accuracy over a distance of 20 inches or more is going to be difficult.

Finally, that wire MUST stay at the same place (axial location) on the wheel. So some flanges or other mechanism must be employed to keep it there. Otherwise, it will walk toward one end or the other as the wheel turns: one wire width per revolution. So some slippage is guaranteed. And this slippage must be constrained to be axial only and not tangential. Old fashioned radio dials used a string linkage from the tuning knob to the tuning capacitor, both of which use rotational motion. And somewhere along the line, there was an indicator scale which was usually linear. This is a similar situation. These radio dial mechanisms ALWAYS used a wheel on the tuning capacitor that was large enough to allow the cord to be tied down at one position on it's circumference because the necessary slippage would eventually allow some tangential movement in addition to the axial movement that was unavoidable. That was the only way to keep the tuning capacitor and the indicator in sync. So it would also be with a lathe DRO but with accuracies reduced from 1/16s of an inch down to below the 1/000" range. For a 20" linear travel you may need a wheel that is over 6.5" in diameter to allow a fixed point of attachment but then your gearing would need to be all that much more accurate.

These are not necessarily barriers, but they must be taken into account.

Other mechanisms are possible. A band could be used. Personally I would favor a timing belt. But they would also have problems.

loose nut
12-04-2016, 06:48 PM
I'm not sure what you're missing but your "dislike" of DROs doesn't make sense. I've run a lot of lathes over the years and I can tell you that the first time I ran one with a DRO it was like a revelation. Absolutely the best productivity tool you could put on one of those machines. I know there's something to be said for "old school" but the DRO on my lathe allows me to run it faster, more accurately and more consistently than I ever could with just the dials. And just so you know, I'm 70 years old so I'm not some young 'egghead" kid who's never done it the hard way...

+1..........................

But to each his own.

gzig5
12-05-2016, 11:23 AM
I like this one. Simple and mechanical, seems to be direct reading with no linkages or gear trains. Bullet proof. With a bit of a radius on the drive wheel you can tune the accuracy. I think the Trav-a-Dial uses similar concept on the wheel.

http://www.ebay.com/itm/252668188427?_trksid=p2060353.m1438.l2661&ssPageName=STRK%3AMEBIDX%3AIT

Magicniner
12-05-2016, 02:02 PM
If you prefer the analogue display of a clock when turning to a stop point it's easy to add a mag indicator and a mag stop to a lathe with a digital DRO and use the digital to set the analogue, that gives you repeatability over the full travel of your Z axis with a clock display that can have it's zero anywhere you like ;-)

- Nick

Spin Doctor
12-05-2016, 09:13 PM
I ran this idea past a couple of enginerrs at work. Neither saw a major stumbling block aside from careful execution*. The biggest issue is of course the drum diameter needs to dead nuts. The question is take the mean diameter or the inner diameter. I've come to the conclusion that the wire/cable needs to run in groove similiar to a hoist sheave. If one were to use a 1/16th cable running in groove with a pitch of .125" with a circumfranced of ten inches the helix angle is .7 something degrees. Difference in helix length versus the wire diameter is 3.1415 versus 3.1413 when you crunch the numbers. I'd make the cable groove to the inner side of the wire and check for repeatability and accuracy over a known distance. The error if any (there will be) should allow you to calculate the needed diameter. Recut the "thread" and check over pins. The big thing here is repeatability. In terms of accuracy issues due to angle changes on the wires/cables entry into the sheave if the wire/cable exits a guide at least ten inches from the sheave and would transition from say left to center to right the possible difference is in tenths by the math. Oh, by the way. One of the engineers used to design machines for making mattresses. So he's familiar with working with wire to form coils of specific length. Plus I have a vague memory of seeing something similiar used on old stereo photography machines used in making maps from airial photographs circa 1940. Big machines that looked complicated as all get out.

softtail
12-06-2016, 09:14 AM
Cables send up a red flag as far as repeatable accuracy imo.

Magicniner
12-06-2016, 09:20 AM
You'd either need a cable that's the same length on cold and warm days or a thermal compensation system, probably not required on mattresses though :D

- Nick

J Tiers
12-06-2016, 10:49 AM
You'd either need a cable that's the same length on cold and warm days or a thermal compensation system, probably not required on mattresses though :D

- Nick

Not for relative measurements, the cable is just a cable (I'd use thin music wire, which is a single strand). But the drum might cause errors of the same sort if it expands. The Trav-A-Dial would have the same issue. Good thing all that is probably at the same temp as the work and lathe itself............