Good tools/Bad tools

Topics include, Machine Tools & Tooling, Precision Measuring, Materials and their Properties, Electrical discussions related to machine tools, setups, fixtures and jigs and other general discussion related to amateur machining.

Moderators: Harold_V, websterz, GlennW

adamc
Posts: 6
Joined: Sat Mar 24, 2012 10:13 am

Re: Good tools/Bad tools

Post by adamc » Sun Sep 16, 2012 7:48 pm

I hear your frustration Ron and share it. I don't have the answer, but I'll tell you what I've done; I bought high quality basic tools on eBay. No "real" machinist would buy second hand squares, because he'd have no idea what he was getting. But that's what I did. It's a gamble where you win some and lose some. The key is to find a few tools you can trust. You need something VERY square and very flat to start with.

IMHO, the difference between HSMs and real machinists is that real machinists have access to good metrology gear OR have calibrated tools. We buy second hand or second quality and need to verify our tools. So job one for the HSM on a budget is to get some basic measuring tools and learn how to use them correctly. I don't think real machinists do this or start out this way. But I think it's a valid approach and a good way to learn the craft.

Start with form- flat and square. Then go for measuring tools- Mics first. Then go for DTIs on surface gages.

It's true the stuff I'm talking about is hundreds of USD at the low end. But by doing your homework, and taking your time, you CAN assemble a first rate set of tools on the cheap.

Harold_V
Posts: 18137
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Good tools/Bad tools

Post by Harold_V » Mon Sep 17, 2012 1:43 am

Mr Ron wrote:Harold, Are you telling me my Starrett and Mitutoyo vernier calipers are and never will be capable of accurate measurements?
Pretty much! Sorry, but they simply aren't capable of *reliable* measurements, due mostly to the fact that there are several variables that introduce error, no small part of which is the hands with which they are held.
They cost a lot of money. Should I get rid of them?
No, you shouldn't. You should, instead, gain an understanding of their limitations, and apply them when appropriate. There may even be times when you must use them when they are not appropriate. I've been in that situation on many occasions, but key to success is understanding that they often yield less than acceptable results. If you can't rely on a measurement to be correct, how can you possibly expect your actions to yield proper parts?

Here's the problem. You take a measurement, which you assume to be correct. From this measurement, you take succeeding cuts, hoping to achieve the size required. Case in point might be a bearing fit. You work to the readings you achieve, boring a pocket. I'm going to assume the bearing is 1" in diameter, and you must have a .0005" press fit, no more. Three tenths might be better.

Your caliper can't resolve a half thou, not even if the display can read the half thou. How then, are you going to know where you are to achieve the desired fit?
There are many instances where a mike cannot be used, a step measurement for instance. A depth micrometer might work, or might not. The caliper seems to be the go to tool in many instances.
Yep. Been there, done that, too. Problem is, is the resulting measurement representative of the actual dimension? My money says it's not, and the years of experience working in a facility where calipers were not acceptable for inspection for dimensions with .005" tolerance proved that to be the case. They simply are not capable of making precision readings. Understand that what I'm talking about isn't the ability to achieve the same reading over and over. The problem is, if it reads .063" consistently, but it is actually .061", what good is the reading? You can get that close with a scale (rule).
I am familar with some of the precision instruments used in industrial settings, that depend on very sophisticated tools; surface plate, gage blocks, optical comparators, etc. These are well beyond a place in the home shop.
Depends on what you're doing in the home shop, doesn't it? If what you do demands precision measurements, they aren't a luxury, they are a requirement. If, on the other hand, you are happy with your processing, and accept items as they turn out, perhaps, to you, it doesn't matter, That doesn't mean it isn't important---it simply means you are able to overlook the importance and settle for what ever results you achieve.

If, however, you hope to achieve the level of compliance that might be expected of a talented craftsman, you don't work by hit and miss. You measure correctly, and work accordingly. That choice is yours, and not for me, or anyone, to determine.
I am far form being a "precision" machinist. I regard myself as a hobbyist machinist because I own a metal lathe , mill and a few precision tools. My metalworking is conducted more along the lines of woodworking where I cut, test, sand and test until it fits. I can "sneak" up to an accurate fit, but not with the repeatability required of a precision machinist like yourself.
If you're happy with that process, no problem. However, if you hope to gain the skills necessary so you don't have to work in that fashion, a very important part of that will be learning to make proper measurements, so you can work according to the requirements at hand.

Our good friend J mentioned, in a different thread, that working to tenths on a lathe is, more or less, a myth. Well, I worked in the trade for 26 years, full time, and did that very thing for years on end. I've known any number of talented machinists who can work to .0002" with success. None of them are hacks, and none of them do it randomly. They do it by proper workmanship. It requires fine feeds, good tool geometry, and proper measurements, but it can be done, and has been done. It is not accomplished with calipers.

If a person hopes to learn to work precisely, there are things that must change in their minds. Equal cuts are critical to success, as is precision measuring. It's something you can learn to do, and do with success. It's up to you. There are many on this forum who are capable of providing guidance. You simply must be open to their thoughts, and willing to discard things you (wrongly) believe to be true.

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.

Harold_V
Posts: 18137
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Good tools/Bad tools

Post by Harold_V » Mon Sep 17, 2012 1:52 am

Chuck K wrote:"Are you telling me my Starrett and Mitutoyo vernier calipers are and never will be capable of accurate measurements?"

You really have to define your idea of "accurate Measurements". If you want to work closer than .002 - .003, probably not.

Chuck
True---and the harsh reality is, .002" or .003" isn't *working close*. That won't yield fits that are acceptable for almost anything aside from clearances. If machinists of any stripe can't understand the significance of two tenths, they aren't going to be successful. When called upon to do critical work (fits---slip, press, snug things any machinist MUST face at least occasionally) to think a thou is "close enough" simply is not acceptable.

To be clear----I do not mean it isn't acceptable to me. It isn't acceptable, because the results are not functional. It's much like playing a musical instrument. If you don't care if it just makes noise, anything goes, but if you hope to make music, you must play the instrument properly. Same with machining, it's just a lot easier than making music.

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.

JTiers
Posts: 422
Joined: Fri Jan 22, 2010 12:55 am
Location: Missouri

Re: Good tools/Bad tools

Post by JTiers » Mon Sep 17, 2012 7:12 am

Harold_V wrote: Our good friend J mentioned, in a different thread, that working to tenths on a lathe is, more or less, a myth. Well, I worked in the trade for 26 years, full time, and did that very thing for years on end.
Harold
Not HARDLY......

Wow is THAT a DISTORTION of what I said...........

I said it was pretty hard to do, UNLESS you took care of several issues..... If you did that for years, and there is NO reason why you cannot, it means that you DID take care of those things.

Name me ONE THING I said that is untrue............. (text quoted below)

Yep, didn't think you could.........

The thing is, a lot of people consider it to be no problem, and believe they ARE "working to thousandths", when they may have errors of several thousandths in reality.....

Uncalibrated equipment, temperature, measurement technique, using the wrong measuring instrument, trying to measure a diameter to tenths when the surface roughnesss is several thou, there are lots of possible problems.

It is SO easy to "fool oneself", unless and until you really DO have to satisfy someone who doesn't know you, has never even seen you, but who has your parts in front of them in the incoming QC dept....... Trying to interpolate the caliper dial to half thous is not going to help you to satisfy that inspector.... quite the opposite.

The average home shopper probably is working to between 0.003 to 0.01" in actual fact. There are exceptions......


J Tiers wrote:I suppose it might be "unkind" to point out that "working to thousandths" has a number of "components" to it.....

1) you must be able to easily measure to that, using an instrument which has 0.001" comfortably within its "range", and not the "finest graduation visible".*

2) You must be able to make SURFACES that are smooth to better than 0.001", or you don't have anything TO measure.... the typical lathe finish is bumpy at least to 0.002 or 0.003", and speaking of single thousandths or tenths at that point is not productive.

3) you must consider part temperature.... hot off the machine will be large, shrinking as it cools.

4) you must consider mic temperature.... the mic you have been holding in your hot sweaty paw for a while is larger than when you took it out of the drawer.

All in all, you really have some work to do to actually measure and make parts in the 0.001" range, let alone to tenths.

*if the smallest graduation is 0.001", you have no particular "right" to suppose that the instrument was even made to be accurate below that....It might be, or it might not. Estimating tenths by interpolation is "ok", but not to be taken too seriously, and surely not requiring a magnifying glass.
I have seen folks claim to interpolate a caliper dial with a magnifier, and all I can do is shake my head in "wonder"..... at the utter fatuousness of the whole idea.

adamc
Posts: 6
Joined: Sat Mar 24, 2012 10:13 am

Re: Good tools/Bad tools

Post by adamc » Mon Sep 17, 2012 8:14 am

Just a quick reiteration from this http://www.chaski.org/homemachinist/vie ... 12#p255569 post:

I have a $30 set of digital calipers I bought from Little Machine Shop. I suspect they are identical to all the other Chinese models available for a wide range of prices. Their published accuracy was .001" and this is exactly what I observed calibrating my set using mitutoyo micrometer standards. I checked every 1" from 1 to 6, then used my AS-1 Mitutoyo gage blocks to check every 1/4" to 1".

My LMS calipers are off no more than .0005" for any given dimension. With care, they are very repeatable. I would consider my set accurate to +/-.0005". I don't consider calipers to be high precision tools. I don't handle them with the care I handle my micrometers. I use them for in-process spot checks and other low precision work. Due to my use and poor treatment they receive, I think they return +/-.001" in real world use.

Just wanted to add that while I get the point of the experienced machinists here, I worry that the rhetoric is getting ahead of science. I recommend folks do what I have done- buy some standards (they are cheap second hand) and check your tools using them. The process also teaches you how to take accurate measurements using your tools. Based on my experience, anyone can return better than .002" accuracy with Chinese digital calipers.

Torch
Posts: 1548
Joined: Wed Mar 17, 2010 7:58 am
Location: Muskoka

Re: Good tools/Bad tools

Post by Torch » Mon Sep 17, 2012 11:05 am

Mr Ron wrote: surface plate, gage blocks, optical comparators, etc. These are well beyond a place in the home shop.
I'm not sure I fully agree with that. Calipers and micrometers are unquestionably a more urgent acquisition, but that is not quite the same as saying other instruments have no place in a home shop. A lot depends upon what one is trying to accomplish of course. The HSM may not have a burning need for traceable calibration certificates, but a few strategic gauge blocks are still almost essential for ensuring that one's own measuring tools are consistent with each other. For example, both the 1"-2" and the 2"-3" micrometers should provide the same measurement of a 1" block.

A surface plate is virtually essential for measuring tasks that require a plane of reference, such as checking flatness or (when used in conjunction with a height gauge) measuring an existing piece or marking up preparatory to making a piece.

The optical comparator is probably the most luxurious of the 3 mentioned, and if truth be told, probably no more accurate than a set of calipers for linear measurement, but ohh so useful for tiny parts with angles. I had all sorts of fun sharpening single-point threading tools to an accurate angle before the OC entered my home shop! Luxury or not, the OC gets a lot of use.

I'm sure you could pick any 3 other items and find them in somebody's home shop because of their usefulness to that individual's work.

And that, to me, is really the key: what is the work being attempted? If I'm making a mounting bracket then the calipers are probably overkill -- I could get away with a steel rule and some punch marks. If I am turning a pen, the calipers are adequate. If I am making a gear shaft then I need a micrometer. My philosophy is to buy the tools that match your requirements, but understand their limitations if the requirements change.

User avatar
ctwo
Posts: 2779
Joined: Tue Mar 27, 2012 12:37 pm
Location: Silly Cone Valley

Re: Good tools/Bad tools

Post by ctwo » Mon Sep 17, 2012 2:18 pm

I learned in school, or at least my general recollection is, that an instrument's reading tolerance is generally +/- its smallest reading digit. A dial caliper graduated to the thousands then would be whatever reading +/- one thou...on top of that one needs to add the actual tolerance of the instrument, which to me seems around +/- three thousands.

I also learned that you cannot have greater accuracy than you have precision, but you can have a very precise instrument that is not accurate at all. If an instrument is very precise, then it seems you can work out the accuracy issue.

My newest caliper is an SPI and I seem to recall its worst reading was +0.001 from the cal lab.

One can make a career out of measurement, considering tolerances, accuracy, precision, and uncertainty...that last item can be a very complex subject. The company where I now work employs a full time Metrology manager, he's earned a doctorate in Metrology...this is not a subject to be taken loosely... :lol:

BTW, I need for my "shop" at least a 2x2 surface plate and a decent set of gauge blocks (I might settle for a B grade set)...especially if anything of my work is destined for someone else, I will need to know...
Standards are so important that everyone must have their own...
To measure is to know - Lord Kelvin
Disclaimer: I'm just a guy with a few machines...

Torch
Posts: 1548
Joined: Wed Mar 17, 2010 7:58 am
Location: Muskoka

Re: Good tools/Bad tools

Post by Torch » Mon Sep 17, 2012 3:39 pm

ctwo wrote:I also learned that you cannot have greater accuracy than you have precision, but you can have a very precise instrument that is not accurate at all. If an instrument is very precise, then it seems you can work out the accuracy issue.
Accuracy vs Precision:

I forget where I heard this explanation, but it always stuck with me:

A watch that loses 1 second per hour is less accurate than a watch that doesn't work at all, since the slow watch is correct once in 43,200 days but the stopped watch is correct twice a day.

However, they are equally precise, since they both indicate the correct time +/- 6 hours.

Similarly, a watch that advances exactly 3600 seconds per hour but is set 5 minutes fast is less accurate than the stopped watch, since the fast watch is never right. But the fast watch is more precise than either of the other two, since it always indicates the correct time +/- 5 minutes.

Jose Rivera
Posts: 3803
Joined: Wed Feb 21, 2007 9:21 pm
Location: Vallejo California

Re: Good tools/Bad tools

Post by Jose Rivera » Mon Sep 17, 2012 8:43 pm

I agree with Harold 100%.

Calipers should be used as reference and a micrometer as the last word ... " AND ONLY IF THEY'RE CALIBRATED ".

Calipers can flex and give different reading depending on how much force is applied.
Also if the "gib" in the caliper is not tight this will cause the jaw to rock giving an incorrect reading.

I still stick to my comment that in measuring tools ( calipers-micrometers-indicators ) Chinese give a high value for the money.
There are no problems, only solutions.
--------------
Retired journeyman machinist and 3D CAD mechanical designer - hobbyist - grandpa

John Evans
Posts: 2110
Joined: Tue Jan 20, 2009 9:33 pm
Location: Phoenix ,AZ

Re: Good tools/Bad tools

Post by John Evans » Mon Sep 17, 2012 9:16 pm

I have to agree with you Jose on Chinese measuring tools. I have a set of 0-6 tenth reading mics and they read spot on to various gauge block stacks. Dial and digital calipers OK but the best dial caliper I have is 20 year old one from ENCO. IT does real good with various block stacks ,my HF digital not so much mainly use it for quick inch-metric conversions. My advise is watch Craigslist as I have got some great deals on quality US made measuring tools. And pawn shops on occasion.
www.chaski.com

Harold_V
Posts: 18137
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Good tools/Bad tools

Post by Harold_V » Tue Sep 18, 2012 12:15 am

JTiers wrote:
Harold_V wrote: Our good friend J mentioned, in a different thread, that working to tenths on a lathe is, more or less, a myth. Well, I worked in the trade for 26 years, full time, and did that very thing for years on end.
Harold
Not HARDLY......

Wow is THAT a DISTORTION of what I said...........
Careful, J, you're going to pop that vein in your neck.

I think you know the message I was trying to convey. It's important for these guys to understand that working closely is reality, and that surface finish, tool geometry, fine feeds, procedures and careful (proper) measurements all go hand in hand. Machinists who achieve the results of which I spoke do not do so by accident. They do so because they follow prescribed procedures, using acceptable methods. And, it's not done with calipers.

Or---the work is sent to the grinding department! :wink:

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.

JTiers
Posts: 422
Joined: Fri Jan 22, 2010 12:55 am
Location: Missouri

Re: Good tools/Bad tools

Post by JTiers » Tue Sep 18, 2012 7:32 am

Harold_V wrote:
Or---the work is sent to the grinding department! :wink:

Harold
Well, that will certainly take care of the surface finish........... we hope......... and grinding is the means of choice for holding tenths.... the machines are made to do it all day every day as a matter of course.

On most lathes... not so much.

Post Reply