Something all should read.Expecialey newbies

[lane]

http://www.practicalmachinist.com/vb/showthread.php?t=173891

[Liger Zero]

Interesting. I make pipe-fittings and other less-than-complicated stuff at home... I don't want or need four-five digit accuracy.

While there is the temptation to have the "best" in many situations you are just paying for accuracy you never use.

Thanks for the link.

[lazlo]

http://www.practicalmachinist.com/vb/showthread.php?t=173891

"The spec sheet for that lathe had the spindle runout at .0003 when measured using an indicator placed inside the tapered spindle hole. The lathe was a new 13x40 Kent TRL"

He should sent it up to British Columbia. It'll measure 2 micron up there :)

[BobWarfield]

Yeah, but they actually build stuff and show their projects in BC. Not sure the same is true of Austin, TX.

Cheers,

BW

[lazlo]

Or hey, I've got a better idea Bob -- I can copy everyone else's projects to my web page :)

[MickeyD]

Most of the stuff that I make is pretty low tolerance. My mill is theoretically good for .0003" and in real life it hits within .001 depending on tool wear and temperature, and most of my customers would be happy with 1/64th of an inch!

[dan s]

Or hey, I've got a better idea Bob -- I can copy everyone else's projects to my web page :)

Or you could act like an obnoxious AS..... Wait never mind :D:D

[Evan]

It'll measure 2 micron up there

1 micron. You can't buy the 1 micron dial guages any more. The best they make now is 2 microns per division.

BTW, 0.0003" is 7.62 microns, not 2.

[lazlo]

Or you could act like an obnoxious AS..... Wait never mind :D:D

No, sorry Dan -- don't want to infringe on your turf there ;)

[Optics Curmudgeon]

Mitutoyo still makes 'em, series 524.

Joe

[lazlo]

Mitutoyo still makes 'em, series 524.

Federal/Mahr (or whatever they're called these days) still makes them too.

[dan s]

No, sorry Dan -- don't want to infringe on your turf there ;)

I freely admit that I am, but I can't hold a candle to you .

[Paul Alciatore]

I have a Mahr that reads in 0.00005" increments. You could probably interpolate to 1/2 or 1/4 of that. The total range on the scale is only 3 or 4 thousanths so you would probably have to use a coarser indicator to get started.

I haven't really used it much but it's there if/when I may need it.

As for this runout issue, I would not be at all surprised to find that different manufacturers measure this spec in different ways. This has been a long standing problem in electronic equipment like amplifiers and it would come as no surprise to find that lathe makers 50 or 100 years apart and 10,000 miles distant would mean different things when they quote runout. To quote a 0.0000" runout is almost meaningless. No bearing is perfect. All of them have runout. The best meaning you can assign to this is the runout is less than 0.0001". But how much less? Would 0.00009" qualify? That's 90 percent of 0.0001". Not really what you might think at a glance. Another maker might hold it to 0.00005" or less. Claiming 0.0003" runout will allow the maker to "pass" more machines at the QC check, if they even do one.

I suspect that other design features will be a better determining factor for the ultimate accuracy of the work you can produce on a given machine. Total mass, stiffness, type of bearings, alignment, accuracy of the ways and lead screws, etc. Don't get hung up on one spec.

[Ed Tipton]

JTG...Why the obsession with such extreme accuracy. There is nothing wrong with wanting a highly accurate and high quality machine tool to work with, and I admire your willingness to seemingly spend any amount of money, but you seem to be obsessing on obtaining the un-obtainable.
You have not told us what type of work you are doing, or what the need is for such extreme accuracy. I will readily admit that I have never made anything requiring the degree of accuracy you are asking for, and I have worked on a few things that required the part to be accurately made.
To work to the tolerances you are trying to achieve, I don't believe is achievable in a home shop environment. Even if your lathe were capeable of working to this standard, there are so may ways for additional error to be introduced into the work...and what about temp control, surface quality etc. Unless every piece of equipment you are going to use to make this part is up to that standard of excellence....including your expertise....you're spinning your wheels. We have all seen excellent work produced on relatively poor equipment...and the reverse is equally true.
I personally believe that you would benefit more from learning more about machining and improving your techniques using the tools you already have rather than spending your hard earned money trying to purchase a high dollar piece of equipment in the hope that it will enable you to do "perfect" work. It's not the tools...it's the craftsman operating the tools.
Your overall statement reminds me of the type of person who will pay out vast amounts of money to a golf pro trying to "buy" a par golf game. No amount of money is going to enable him to play par golf. The approach just wo'nt get him there. The only way for him to get there is to put in his time at the golf course hitting that little white ball.
Buying a new super-whoop-de doo lathe" is not going to make you instantly capeable of turning out zero error parts any more that using Tiger Wood's driver is going to make you drive that ball 300 yards. It just doesn't work that way.
I once worked at a place that permitted the workers to use the company equipment for "government" projects. We had one mechanical engineer who liked to come down to the shop after hours and work on his projects. We could all tell when he'd been there because the tools would all be screwed up. He was good at cleaning up etc. , but he would tweek and turn all the knobs and levers because the machine "wasn't set up right". He was a good engineer, and could talk a good game, and knew all the buzz words, and could use a micrometer...but as a machinist...he was a good engineer.

[oldtiffie]

Originally Posted by lane
http://www.practicalmachinist.com/vb...d.php?t=173891


"The spec sheet for that lathe had the spindle runout at .0003 when measured using an indicator placed inside the tapered spindle hole. The lathe was a new 13x40 Kent TRL"

He should sent it up to British Columbia. It'll measure 2 micron up there :)


1 micron. You can't buy the 1 micron dial guages any more. The best they make now is 2 microns per division.

BTW, 0.0003" is 7.62 microns, not 2.

https://www.machineryhouse.com.au/Products?stockCode=Q223

[Circlip]

Strange how the"normal" preoccupation with bigger has now reversed to smaller :rolleyes:

[oldtiffie]

As is all too often the case - either wrong or side-tracked or both. The thread was side-tracked (hi-jacked?) to a discusion on indicators instread of determining the accuracy and alignment of the spindle bore taper.

Putting an indicator on only one part of a bore is an exercise in futility. If that part is OK, the rest needs to be checked as well. It has to be a straight-sided symmetrical cone with its axis concentric with the spindle bearings as well as aligned to the lathe bed longitudinal axis.

So, first off, the concentricity needs to be checked in several places along the cone. It is hard to differentiate between eccentricity of a perfect circle on the cone and "out of round" on/of the cone.

What is really required is a good test-bar which is essentially a precision-ground bar with a taper which the lathe spindle has to accurately match as well as a plain straight cylindrical parallel extension.

The tapers of the lathe spindle and the test bar should be thoroughly cleaned and the test-bar taper covered with a very fine smear of Prussian Blue along its length and the two tapers "wrung" together and then separated. Check the taper on the lathe spindle to see whether it is accurately or not. If OK, proceed to the alignment test, otherwise consider what accuracy you require and what you are going to have to do to achieve it.

The alignment test is simply to put a good test dial indicator on the test bar near the spindle end and rotate the spindle slowly. Read and record the run-out. Repeat at the end nearest the tail-stock. If the run-outs at both ends are the same and have the same "zero" amount and points then all is OK. If the second test varies from the first the bore is not aligned to the spindle axis as the parallel section is describing/"sweeping" a cone instead of a cylinder.

Next, with the spindle stopped, run carriage left and right along the length of the parallel section of the test bar. Turn the spindle 90 degrees and repeat - 3 times. If the indicator deflects the spindle bore axis is not aligned to the lathe bed axis and needs to be (re?)adjusted.

[Willy]

As is all too often the case - either wrong or side-tracked or both. The thread was side-tracked (hi-jacked?) to a discusion on indicators instread of determining the accuracy and alignment of the spindle bore taper.


I believe the point Lane is trying to make is that it isn't Important if your lathe has .0000 or .005 spindle bore run out. Don't worry about getting or having the perfect machine.

Instead concentrate on reducing the error inherent in the operator of the equipment. Much better to have a skilled operator on an imperfect machine, than to have an unskilled operator on a perfect machine.

[MickeyD]

I would like to add to what Willy said. The imperfect operator will bring the perfect machine down to his own level. Sort of like the Peter Principle but with machinists and things that break.

[Your Old Dog]

1 micron. You can't buy the 1 micron dial guages any more. The best they make now is 2 microns per division.

BTW, 0.0003" is 7.62 microns, not 2.

Would you believe me if I told you I had one? I can't use it though, everytime I get it too close to my crude shop it starts to glow red and vibrate like hell :D

I got it from a piece of equipment the company got rid of to measure oxide drop out on video tape at speed. I think mine is a Mahr.

Thanks Lane, that was a useful thread for me and I don't normally read that board.

[J Tiers]

A few "newbie rules", that seem to be programmed-in to many folks.

1) The specs that are printed on the paper are all true, and represent accurately the machine, no matter what the source.

2) The machine must ALWAYS be much better and "more accurate" than the best work you ever want to do on it.

3) There is such a thing as a half of a tenth of a thousandth of an inch when using an ordinary lathe.

4) Errors represent the machine, since a good machine makes the work independent of the operator.

5) if there is anything wrong with the machine, it will ALWAYS be a serious and disabling defect in the most expensive and central part of the machine, typically the spindle and spindle bearings.

.
.
.
My favorite is of course #2, which makes me then wonder where the machine came from that was used to make all the Monarch 10EEs and bearing grinders. And where it is now.

[S_J_H]

I think it is quite useful and a very good learning experience to have a very sensitive DI. A lot of fun has been poked at how true a SB9 spindle will run. And yes they will run very true as it is a plain bearing.
Put an indicator on it, get the oil film established and don't touch or breath on the lathe and you'll see a remarkably low reading of spindle runout.
Now do the same while taking your pinky finger and just put a little pressure on the tailstock end of the ways and watch the needle defect.
A real eye opener...You can if nothing else, at least learn the best way to use your machine and where the weak links are that will cause unwanted deflections.
Steve

[MickeyD]

I had a 10L (heavy ten) Southbend with the plain bronze bearing headstock. When setup correctly it would give a beautiful finish except for having a pattern of little marks repeating on the finish. I finally figured out that they were coming from the metal clip on the old flat leather belt. That little bit of change in pressure on the spindle showed up in the work. I finally replaced it with a glue together belt and the marks went away.

[oldtiffie]

Originally Posted by oldtiffie
As is all too often the case - either wrong or side-tracked or both. The thread was side-tracked (hi-jacked?) to a discussion on indicators instead of determining the accuracy and alignment of the spindle bore taper.


I believe the point Lane is trying to make is that it isn't Important if your lathe has .0000 or .005 spindle bore run out. Don't worry about getting or having the perfect machine.

Instead concentrate on reducing the error inherent in the operator of the equipment. Much better to have a skilled operator on an imperfect machine, than to have an unskilled operator on a perfect machine.

Thanks for the prompt Willy.

I could see what Lane was driving at.

I was just showing how a good (or any) indicator is neither the be-all or end-all of determining the accuracy of a lathe spindle if used in limited form or in isolation. Some-times those "good ideas" create more problems and misconceptions than they solve. They are often ill thought-out "knee-jerk" solutions. It gets worse when nobody challenges them as a faulty solution becomes "the" solution by default.

The taper really doesn't matter unless you are relying on something which in turn relies on the bore taper - such as collets (MT or sometimes R8) that bear directly on the bore taper or something such as collet adaptors which rely on the bore taper as well as their own internal taper (think ER and C collets).

A good set of soft jaws will solve most problems - even if the chuck is out by say 5 degrees!! It simply does not matter as so long as the finished job is within the material that is to be removed, as the job will be aligned to the spindle axis.

Now, whether the spindle axis is both aligned to the lathe bed and whether the cross-slide is normal to the lathe bed axis are other but very related matters.

I will mention but not pursue the matters of the ways of the lathe bed, cross-slide and top/compound slides being straight and true which are also very related to the accuracy of the lathe - as is wear on the lead-screw (for thread-cutting).

That brings me to your point about the competence of the lathe operator and his appreciation of tolerances (sizes and finish) as well as "working around" any "problems" with the lathe.

On several occasions, Forrest Addy has mentioned the matters of rigidity in an indicator set-up and the possible comprising of the accuracy of them. Everything is "elastic" to a degree and things do "give", bend, deflect, sag etc. as the reaction of the thrust of the indicator as well as its mounting may well play a significant but hidden role in the "accuracy" or otherwise of the indicator readings. This gets both harder and more important as the degree of accuracy rises.

[lane]

Hay every one . I was trying to not say any thing . It`s just a good read.
Every one thinks their lathe are what ever is just so precision. I say diddle dam. You may once in a lifetime turn .00001 round are to size but that was just a fluke. When you are working to 10ths you use grinders. I am beginning th thank a lot of folks do not know what .0001 is are less . Yes I put a .000050 indicator on it and the needle did not move so it was that close . I say BULL $hit plain and simple. Their may be some machinery out their that accurate but a home shop guy sure as heck don`t have one. Besides if your parts fit together and measure right what the heck does it matter.

[oldtiffie]

Agreed Lane.

I do know what a "tenth" is - and what it is not.

I get by on tolerances as regards both fit and finish. If the fit and finish are OK the job is OK. But fit and finish are looked at first.

I, as Lane alludes to, find that a thou (occassionlly) or more (frequently) a "wider" tolerance gets the job done.

I have Chinese lathes (2), mills (3) and grinders (3). The grinders hardly ever get used as my challenge is to make do with the least accurate machines and tools. Its quite surprising how well and how often it works. My T&C, universal and surface grinders (all of which are very well made) can "hold a tenth" if need be go months without being used unless I think I need to "give them a run to keep my hand in" - and sharpen bloody cutters!!

I use rulers, calipers (digital and old-fashioned "spring") as well as marking out and milling to scribed lines as I can.

I am more than passably good at marking out and centre-punching hole centres and line, arcs etc.

I rarely use the mill for ordinary drilling as that is what the pedestal drill and portable drills are for.

I fabricate a lot of stuff and find that I can use the side of the pedestal grinder wheel or the angle grinders to "clean-up" most times as "looks" and "finish" may be more a requirement than accuracy.

If I need to bore accurately I can repeatedly "hold" half a thou with relative ease and "2 tenths" with a bit more effort and care - with spring calipers or "expanding ball" guages. I rarely have to work to closer than a thou or two.

Because I clamp directly to the mill table most times,I don't have or get a lot of the problems associated with vices. My vices aren't big but they can do the job pretty well.

If something goes on the grinders, it is mainly because of the hardness of the job or the "clean-up" required - and sometimes, but not often, for the accuracy of it.

My machines are generally "light" as I don't "push" them, I just take my time and get the job done so that it works.

There are really on two reasons why a job has to have high accuracy - neither of which are mutually exclusive:
- the job really does require it; and
- the operator just wants it that way for personal reasons.

[10KPete]

Back about a century ago (:D ) there was a "Bull of the Woods" comic showing an old guy running a big old lathe. A few yards behind the old guy the Bull is talking to one of the other shop hands saying something like..."When that guy retires tomorrow we'll have to scrap that lathe 'cause no one else will be able to give it the bump and tweak necessary to get good parts out of it".

Well, words like that anyway.

Ya just have ta know where to shim, bump, tweak and 'adjust' during the cut to make the part come out right!!

That's a real machinist.

Pete

[Optics Curmudgeon]

If you haven't yet, Google "adventures with klunkers". I can't remember Meridian's owners name, but he has it right.

Joe

[doctor demo]

Hay every one . I was trying to not say any thing . It`s just a good read.
Every one thinks their lathe are what ever is just so precision. I say diddle dam. You may once in a lifetime turn .00001 round are to size but that was just a fluke. When you are working to 10ths you use grinders. I am beginning th thank a lot of folks do not know what .0001 is are less . Yes I put a .000050 indicator on it and the needle did not move so it was that close . I say BULL $hit plain and simple. Their may be some machinery out their that accurate but a home shop guy sure as heck don`t have one. Besides if your parts fit together and measure right what the heck does it matter.

I don't know about the rest of you guys, but I have a radial arm saw that has been converted to a grinder that can hold .00005:D .

Just kidding

Steve

[tony ennis]

If I ever get my clunker lathe running I'd be delighted with .001"

As a wanna-be woodworker, I can't really visualize/understand anything smaller than a 64th anyway.

[Circlip]

Don't understand the problem, once you fit the "Standard" machines with DRO's you IMMEDIATELY move into the realms of super precision. Wish we'd had them many moons ago in the toolroom I worked in, would certainly have saved in temperature control and clean room conditions. :rolleyes:

Regards Ian