Making a reference square

[Conrad_R_Hoffman]

Got fascinated by squares and decided to make one. Probably the 837th write-up on the matter, but added a page on it anyway- http://www.conradhoffman.com/refsquare.htm

[ctwo]

Thanks, Conrad. I think I've learned something from all your posted projects. I did not consider what the ball on the end of the old height scribers rod could be used for. I was peculiarly looking at that particular measurement setup and wondered what you were using for the lower reference point.

[Conrad_R_Hoffman]

Yes, when I saw the ball being used like that it was a slap-your-forehead moment. I'd like to make a dedicated comparator, but it's one more project I don't have time for. Being able to use what I had helped progress a lot. I've added some comments and changes on the page, notably that to increase the baseline and accuracy, one could also make the same thing from a piece of iron pipe, say 3"-5" with a flat down each side. I'm all about cheap readily available materials, as buying large pieces of tool steel is getting too expensive. It also occurs to me that one can buy some really cheap 123 blocks. They could be used, but are too small. If 2-4-6 blocks can be had (without being drilled like Swiss cheese), they can be checked for parallel and squareness. A bit of luck, or a bit of lapping and Bob's your uncle.

[ctwo]

Well, my first thought on the reference block was, why not just pick up 123 blocks, etc. I think you can get them for less than the raw material, if you consider the cost of hardening.

I used both 3" and 4" iron pipes to make cylindrical squares - they are probably cast steel, but not sure. I have not been able to face the edges to much better than a thou (8") and was thinking of welding in end-plates so I could mount them between centers and try again, or try to lap them in as they are. At least two sides should be square, though they do not seem as practical as blocks.

[pete]

That web page is more than well done Conrad, good well researched information is never a waste of time.

I'm still unsure if the ball on those old school surface gauges was ever purposly designed so squareness checking could also be done but it obviously does work. At the price I paid for a brand new Starrett surface gauge I was less than impressed to find they no longer have the ball. I do know the general practice is to park the bent needle scribing tip inside the undercut below the ball as an easy way to make it a bit safer when picking up the gauge. I'm not even sure that's what the undercut was for and suspect the ball was more decorative than anything else since Starrett seem to now judge us as unworthy of needing it. For those without a surface grinder Randy Richard on Youtube mentioned using a fiber cut off disc in a mill to slot the end of the hardened surface gauges base so a light press fit radiused plate could be used instead of the ball. Because it has a lot larger radius it's much easier to set up while checking squareness with a dti. There's a couple of commercialy made indicator bases that already have that radiused plate, but there more than a little proud of them for what there priced at.

I used a couple of large wrist pins out of a V 12 Cummins for years before I finally bought a half decent hardened and ground cylindrical square. There still not cheap when you can find them, but I also have one of the B & S 558 cylindrical squares you mention on your webpage Conrad. In hindsite as good as they are I'm not sure I'd really recommend buying one. There still a subjective method of measurement where your judging by eye exactly where on the cylinder you get a zero light gap fit. I do know the human eye can judge light gaps well under .001" At times I'll use Zig Zag cigarrette rolling papers as edge or tool tip finders and there almost exactly .001" thick. It's not hard at all to see something that thick. With good eyes I'd think under a 10th light gap is more than possible. Personaly I think a surface gauge with that radiused plate, a good 10ths dti, plus a known to be accurate shop made or commercial cylindrical square to preset the dti's zero at the required height might be the fastest, cheapest and easiest method in most home shops unless an actual squaremaster or equivelent can be justified.

Of course it all depends on what someone is using there home shop for. But I consider at least a small cheap surface plate and a shop made cylindrical square as basic references. The longer you have them the more uses you find for what they can tell you. Don Baily at Suburban Tools did a great video showing where even a cylindrical square with an inaccurate base it's still in fact perfect at two points around it's circumference. If those are known and marked with something like a sharpie you can still get very good results. Like most I've got a couple of machinist squares, but even Starrett's and Mitutoyo's absolute best still have allowable tolerances and again are a subjective method with no real numbers for how far out something might be. I find I'm using them a lot less than I thought I would.