|Geoff Manship||05/01/2012 15:12:17|
|23 forum posts|| |
Some time ago I brought two engineer's squares. On checking the squareness of each, I found that one was appalling and the other just not very good. I now want to buy a 4” and 6” square of reasonable quality for use in the workshop. I don't think I need inspection standard, just reasonable accuracy and quality.
I did intend buying Moore & Wright 400 series squares, but on the Tool Fast website I noticed their Fisher range.
According to the spec, accuracy appears to be comparable with the Moore & Wright equivalents but the cost is significantly lower.
If anyone has had experience of Fisher squares I would be glad of their comments.
|chris stephens||05/01/2012 17:24:14|
|1045 forum posts|
Cannot comment on Fisher, but M&W seem to have gone downhill in recent decades.
Have you considered one of these;
Made from that wonderful stuff, if you have a damp old shed to work in, Stainless Steel. They are also of a one piece construction, so there is no risk of the blade coming loose as might happen on a cheapo piece of Tat.
|304 forum posts|
Just a thought - I brought a nice, new Cromwell square of eBay - can't complain about the qualitiy and squareness (not that I've measured it !) - BUT - it's a bit of a silly design as the cut away that allows a scriber to go right to the edge goes to the top of the square if that makes sense.
So it's easy to have the square not sitting quite square . . . .
The 'notch' scribber cut away (like the old beaten up M&W one I have) doesn't suffer from this and is certainly the type I'd recomend.
|Jim Greethead||05/01/2012 19:56:34|
131 forum posts
Interesting. I don't recall seeing specifications before like those in the reference URL but if they refer to the accuracy at the end of the blade then they agree with measurements I took the other day of the squares in my workshop.
At the time, I thought the squares were of low quality, expecting better then plus/minus 0.02mm but it appears that they are suitable for purpose.
|Martin Walsh 1||05/01/2012 21:21:15|
|113 forum posts|
1935 forum posts
The easiest way to check a square is to set it on a reference edge (side of surface plate etc) and scribe a line against the blade. Flip it over 180 and scribe another line against your first. If not accurate draw file and stone the blade until it is. - A very good exercise in benchwork which we all had to do in our apprenticeships. You never accept that things will be accurate enough and you need to know how to remedy faults as engineers did in the past, not simply use a chequebook as the answer.
p.s. have a look at the article in MEW number 1 or perhaps 2 regarding French try squares.
Edited By Terryd on 05/01/2012 21:51:11
|Jim Greethead||05/01/2012 21:57:52|
131 forum posts
With my hand skills, I would have a very thin square at the end of that exercise. But I guess it would not hurt to practice on an old square that is already stuffed.
Might give it a go.
|1008 forum posts|
I have a mild disagreement - the turnaround method described by Terry is very good, but the most accurate way to check a square is to use the lathe bed (or other suitable) as a reference, and put it up against one of Michaels cylindrical squares, which are so easy to make. Some kind of light behind is best.
Put a bit of (true) bar of sensible size into a 4 jaw, and clock up true. Face the end, and then turn a recess. How deep doesn't matter, but about 1/8 is fine - to relieve the base is all that is needed. Just make sure that the edge of the recess is as thin as is sane - say 1/16". Dismount from the chuck and wrap carefully, because it will be square to within pretty much nothing. If one thinks of geometries of generated circles its easy to see why.
That was in ME years ago, when it still did tooling and printed in hot metal/B&W!
I accept that the thing can only be truly accurate if the edge is infinitely thin, such that it represents 2 points, but we don't need to worry about that, especially as we can't achieve it, but it will be the rightest angle one has in the workshop, or is ever likely to posses.
A 3/4/5 triangle is also absolutely a right angle, but there is the possibiltity of introduced errors if ones drilling and measuring are not spot on. Still, made properly such a triangle should be very accurate.
|John McNamara||05/01/2012 23:32:03|
1311 forum posts
"Put a bit of (true) bar of sensible size into a 4 jaw, and clock up true" ? in the post above.
An interesting proposition. If the point to be measured on the bar is adjusted to be to be spot on the axis of the lathe spindle the indicator will "clock" perfectly; all good.
But what if at the same time the bar is actually off axis as it passes through the chuck jaws? the indicator will read a regular rise and fall as it travels the path of the ellipse generated, however the rise and fall will be very small, and maybe not noticed.
The small error however will be multiplied by the length of the square. it may well be significant.
A possible solution particularly for a longer cylindrical square is to finish face the ends using the fixed steady well oiled and fitted carefully with no play at one end. and only gripped by the last few mm in the chuck jaws at the other end (perfectly clocked true at that point near the jaws)
As only a small finishing cut is required a light grip will be sufficient.
This method effectively assures that the axis of rotation (of the bar) is correct at 90 degrees to the cut. while facing the ends.
The fixed steady is another source of possible errors, if off axis it stresses the chuck jaws, if it is not correctly aligned with the chuck, so this setup should be done carefully. However if the fixed steady is slightly off axis the most likely outcome is the bar will slightly flex the chuck jaws in and out (it is constrained in the fixed steady at 3 points). however the work will still rotate on its own axis.
A good analogy of this setup is the drive shaft of the car. The shaft rotates on its own axis while it can move up and down.
The setup of the fixed steady should ideally be centered on the lathe axis of rotation within .001 inches or less that way the errors generated by the less than perfect connection at the chuck will be minimised.
Edited By John McNamara on 06/01/2012 00:00:17
|Chris Trice||06/01/2012 00:14:48|
1362 forum posts
|You don't need to clock the bar at all. Just take a light skim off the surface and providing your lathe turns parallel (easily checked by measuring each end of the bar), once you've faced the end as described, it must be square. If you centre drill the end of the bar, you can use a rotating centre in the tailstock for support providing it's set centrally, again easily checked by measuring each end of the cylinder for parallelism|
Edited By Chris Trice on 06/01/2012 00:20:33
|John McNamara||06/01/2012 00:57:17|
1311 forum posts
True if you have a near new lathe and you are happy with a turned finish.
On the other hand if you have a nice piece of precision ground stock you want to turn into a square you will not want to damage the fine finish.
I doubt if there are many lathes used by members that can turn parallel over the entire length, say 300mm to within .001". you can get the ends to match but what about the middle? and over longer lengths spring of the work starts to starts to cause errors
If the ground stock was not available it is not hard to hand lap a turned shaft to parallel
Turning on accurately located centres using the fixed steady as above (it is unlikely the bar will fit through the spindle 75 to 100mm is a good size) will work as an alternative to the method I proposed in the previous post. However if the stock is a tube quite often used as a square plugs will have to be made and centred then fitted in the tube.
My "best" square is a piece of precision ground drum 5mm thick rescued from a scrapped printing machine
100 mm diameter and 350mm long.
1935 forum posts
Hi Jim, if you have duff squares what's to lose? it can be good experience and you may well end up surprising yourself with improved skills. File off the bulk of the error across blade with a fine 4 to 6 inch file, then carefully and deliberately file along to remove marks and flatten along the blade, then draw file and finish by draw filing with fine emery cloth held on file. Do it a bit at a time and keep checking. You may well surprise yourself, but think and analyse what is happening as you are working, it ain't so hard.
|Geoff Manship||06/01/2012 17:27:30|
|23 forum posts|
I thought my question was too trivial to merit more than perhaps 1 or 2 replies.
Thank you to everyone for their answers and comments.
I think I shall try one of the Fisher squares and see what they are like. I'll post here with my findings.
Thanks again for your replies.
|1008 forum posts|
I agree about the limitations of making cylindrical squares. Technically you don't need to clock the thing at all. - as long as it holds parallel.
But reality intervenes.
Ideally one wants to clock the thing true and set it parallel in a fixed steady with a proper DTI - I did that with my Quorn spindle, and eventually got it true to .00002 in 4"., before machining the bearing seats. Madness approached.
So you have to accept a point where good enough is good enough, and one can make it as well as it need be.
Edited By mgj on 06/01/2012 18:20:54
|Jim Greethead||06/01/2012 19:19:23|
131 forum posts
Michael: Your method of making a 3-4-5 square sounds like a good use for a bit of ground bar salvaged from a printer or copier. And there are other sources of cylinders that are accurate in diameter over sufficient distance; some drawn pipes might serve. Sounds like a job for a time when everything else is going wrong and one needs a break.
Terry: Thanks for the encouragement; when I did a Tech course a couple of years ago, I was obliged to file the usual drill gauge etc. and surprised myself by being able to do it. Filing a square is another order of accuracy but using your advice, I think I might be able to do it.
I did think about clamping the worst of my squares to the milling machine and taking a light cut. But of course, that would only make it as accurate as my mill.
All the discussion about cylindrical squares set me thinking (oh no!): flat plates are made by testing three against one another. That should work for three squares shouldn't it? Of course they would have to be set on a flat surface. The good news is that you would end up with three really accurate squares and could sell the other two.
Just a thought
|Stub Mandrel||06/01/2012 21:09:02|
4307 forum posts
I have three small squares (1", 2", 3") from Chronos, all spot on as far as I can tell.
I have another, much older square apparently of english make, patent design and high quality - it's bent as a dog's hind leg and IO can't see how it could ever have been straight.
Two woodworking squares I have are close, but not good enough for metalwork.
|Jim Greethead||07/01/2012 00:08:49|
131 forum posts
" Another absolute method is to make 3 ordinary squares but with hinged blades . By testing them all against each other on a straight edge and adjusting so that they all touch full length on the blades in any combination you end up with three perfect 90 deg squares ."
Thank you Michael, that is a great refinement on my thought about the three squares. Three steel rules - the ones with the holes in the end - for the blades, a bit of flat rectangular stock and some work with the slitting saw and Robert is your father's brother.
And no filing (sorry Terry)
Sounds like a plan
1935 forum posts
The filing method is simply good practice. I agree about the three squares method, but that only gives you a test square. An adjustable square is just that - adjustable, and if only using one hole to assemble, how will you prevent it moving?. One knock or drop and you have to go through the process again. If you try to drill an extra hole to lock blade how do you ensure no movement. You can't silver solder, that can cause movement or distortion. Loctite may work but only if you're quick in your three square testing.
After any attempt to make the blade fixed permanently you have to check the square and then make it accurate,- - by grinding on a precision machine - or hand file. Square one!
Good luck and share your experiments and results, we are all keen to learn new tricks.
|Gordon W||07/01/2012 10:11:33|
|2011 forum posts|
We had to make a square in apprentice training ( here we go again ) made the actual square 6" x 4" x 1" from gauge plate by filing, then riveted two flats for the handle bit. This is accurate and cannot move. I'm still using it almost every day.
|Nicholas Farr||07/01/2012 12:05:48|
2262 forum posts
Hi, I also agree that the turn around test is the best method for general use. There are a couple of things to consider though, firstly using this method you are only checking the inside edge of the stock to the outside and/or inside edge of the blade, therefore you will have to either accurately measure parallelism of the inside to outside edges of the stock, or have a reference for the outside of the stock to rest onto which has a rebate wide enough for the stock to lay into, to at least the depth of the stock to blade sides. Also the reference edge must be reasonably flat and straight and square to top of the side you will scribe your test line onto.
Checking a square against a cylinder square has some of its own problems to get a true interpretation, apart from the cylinder squares accuracy. Firstly the blade must not have any significant curve long its length in its flat dimension, and secondly the base of the stocks face, sideways, must be square with the sides of the blade, because the side of the blade will have to stand exactly in the same plain as the side of the cylinder square when the stock is sitting on the same surface as the cylinder square, i.e. assuming that the cylinder square is perpendicular to the reference surface, the side of the blade must be also, to get a true reading.
Please login to post a reply.
Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!
You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.
Click THIS LINK for full contact details.
For subscription issues please see THIS LINK.