Measuring flatness, parallelism and perpendicularity on a surface plate. 

I’d love to claim the credit for that, but I’m pretty sure that I saw Tom Lipton do it. That gives me an idea, I reckon I’ll have to make a few comparators and put them up for sale.

That's the essence of it, but there are different ways to achieve the same result. Tom Lipton, for example, put together a square with adjustable feet meaning that the two faces which he had ground parallel could be calibrated to a surface plate without further scraping or grinding.

You need to flatten both sides 3 AND 4. AND check that 3 and 4 are also parallel to each other, using the same technique as with 1 and 2. This is vital for the next step. Now you can check for squareness by comparing the measurement at sides 3 and 4. The actual value doesn't matter but it should be the same at both sides. If the sides are leaning, they will be leaning "in sync" because they are parallel. Specifically, if side 3 is leaning towards the indicator, side 4 MUST be leaning away from the indicator. That's what's so cleaver about this method. I fear this wasn't too well explained in this video.

It was nice to meet you aswell, the livestream was pretty interesting and I certainly learned a few things. The plate certainly looks great on camera...

This one is a Mahr Millimess, it is a plunger type indicator but with a very distinctive dial casing.

You know, I don't know if that's really what it's for but it certainly works. A proper stand with a bumper is easier to use but this does the job.

@@ImpracticalMachinistWith the ball positioned at the top of the shaft, it's both for aesthetics or a bit of traditional decoration I guess, but more importantly, it's really meant to be used to park the tip on the bent end of the scriber into that groove below the ball so you don't stab yourself by mistake when your not actually using it. However Starrett in there bean counting wisdom no longer adds that ball tip. I bought a brand new Starrett surface gauge a few years ago for some fairly decent money. Unfortunately I should have noticed in the pictures they no longer have that ball end. Between that and the nice fat run right on the top of the gauge base that never should have made it through any level of QC, I'll buy anything metrology related from Mitutoyo, Mahr or or others.

@@Pete-xe3il I'll have to try that out the next time im using it with a scriber, never occured to me that it might be used for that.

@@ImpracticalMachinistIt wasn't obvious to me either, but I ran across a thread on Practical Machinist a number of years ago that mentioned parking the bent end of the scriber tip in that recess. For what ever reason, few today seem to know about that little detail. Once I did know, I've since run across older drawings of various brands of surface gages that sometimes show that tip set into that position as well.

Because we need to use the surface plate as our flatness reference, otherwise we would be checking the face for parallelism against its opposite face.

That’s a good point, I wish I’d mentioned that now.

Well, you would be "nicely" surprised, how "round" are balls from China ball bearings !

Ball bearings are graded from 3-2000. The smaller the number, the more precise. Typically noted as a G-value. A decent ball bearing for this application would be G25. That's what I use on occasion, FWIW. Bearings made for, say, slingshots can vary by a couple of thou. That's not round.

@@vaidaspetraitis9353 Always check the bearing's G-value. COO has nothing to do with this.

@@felixf5211 Even a G200 bearing is spherical within a tenth, which is the resolution of the best indicator I own. And they're really cheap.

Within one-ten-thousandth of an inch, or .0001" Usually you'll hear tenths, or thou (.001"), and occasionally millionths but not in my workshop.

@@ImpracticalMachinist Ahhh, of course. That makes sense now. Many thanks. 👍😀

Yes, the mounting hardware tightens against the shaft on the holder and the shaft on the gauge simultaneously.

It’s a couple of attachments borrowed from a dial test indicator set, nothing special or proprietary.

Wouldn't it! Unfortunately the workshop is full of the contents of our loft at the moment so I can't even get out there to work, never mind making videos.

In this case because it's a reference not a bearing surface I want the scraped faces as flat as possible.

@@ImpracticalMachinist right, that makes sense. did you have any particular technique to scrape it that way? I'm not sure how I would tackle that to be honest. Have you started with a precision ground surface?

Get close with the mill, i face milled the block but a fly cutter may be a better tool for the job. Then scrape, let the blue on the plate thin out and gradually ease off the pressure on the scraper as the dips get shallower. When stoning off after each pass just go over the piece a bit more than you usually would for knocking the burrs off. Once you’ve done that and you’ve got it to where you think it’s flat, leave the piece alone for the night, come back in the morning and check it again, scrape as needed. There’s a reason that straight edges, squares etc are stress relieved, because the metal will move, so if you can start with a cast iron square that’s been treated and stress relieved that would be a way better start. I’m going to do a video covering a proper square build at some point.