Most home machinists don't need and can't achieve any better accuracy than your countertop cutoff. Great demo! I'm working on naval guns that were made between 1910 and 1942 and the quality of machining on the original parts is amazing but as long as I can hit within .0005" the parts fit and work.
Great video mate! I am learning to hand grind knives and the last thing I need is some moron telling me that I have to have tooling that is accurate to less that a tenth or "Its a waste of time". Your proposal is a practical solution to a real world problem that isn't going to cost $500. Its good to see someone using their brain and not just parroting what they heard on the net. Thanks.
This one just cropped up on my feed, so yeah, here's me commenting 5 years down the line, not so much for Benjamin, but for those coming later. There's people here commenting that this is not a good test, and to a certain extent they are right. If this is a flat(ish) surface with a maximum deviation of ±25 microns over 10 cm, or a concave spheroid(ish) surface with maximum deviation of ±25 microns over 10 cm, , or a convex spheroid(ish)… you won't be able to tell using this setup. A machinists level can tell you, and an autocollimator can tell you, but that's about it. However, it's unlikely to be a spheroid, which is usually achieved by lapping (see telescope mirrors, etc), it's far more likely to be flat(ish) But. If what you want to do with it does not require more accuracy than that, it's fine. And most home shops do not require more accuracy than that for most uses of a "surface plate". Maybe if you're reconditioning machines, or measuring pieces that require high accuracy. Most of us aren't doing that. Most of us will be using it with a height gauge for measuring or marking out. It's fine in that respect. It'll get 90% of what you want done. What I would be more worried about is surface durability. Anyway, it's a good example of what you can find "over the counter" (if you'll excuse the pun).
Thanks for the clear explanation. I probably don’t need a surface plate but I will keep my eyes open for granite like this. Might be fun to figure out how inaccurate it is. If I can’t figure out a problem with it, I probably don’t need a better one. :)
@@gf2e Oh, you probably do need a surface plate, even if you don't know it yet. The same way everyone needs a lathe :) Headstones are often made from granite, and offcuts are cheap.
@@wibblywobblyidiotvision I feel like there is some sort of morbid connection between people getting lathes they don’t need, and tombstones. I do want a lathe, because I often want weird screws that I can’t easily find locally. Clearly, I need a Tornos :)
@@gf2e You can measure how flat the surface is with your indicator and a 'reference bar' that covers the surface's diagonals. The bar does not have to be a perfect straightedge, just straight enough to be measured within the range of your indicator. First, measure the bar's thickness variations at the ends and center with a micrometer, or by setting-up the indicator over a single point of contact in the middle of your surface and sliding the bar under the indicator. You don't need to read everywhere (just the ends and middle) and you don't need the actual thicknesses (just the variations measured with your set-up). The difference between the end readings is the bar's 'taper', and the center reading minus the average of the ends is the bar's 'bulge' (or 'neck' if it is negative). Second, support the bar on its ends and zero the indicator at the bar's center. Rotate the bar 180 degrees around the long axis, and read the indicator's change. That is your bar's 'bend'. From the bulge and the bend you can calculate the bow/sag at the center of each of the bar's surfaces. Now you can use the bar to measure the surface's bow/sag in the center, and the 'twist' between the diagonals. You can also characterise the same bar over other distances for the surface's length and width, or the sub-span of a measurement grid.
@@gf2e I find my lathe incredibly useful. One thing to do is start any project by asking yourself 'can it be round'. This is usually a new way of thinking about stuff. If the answer is yes then making it can be joyous lathe work.
I made 1 from scratch, shooting for .005 for woodworking, but easily got .002. The indicator setup you used cannot pick up widely-spaced holes and high places. I used a home-made precision spirit lever with 3-point contact.
I use counter pieces for my sandpaper lapping plates as you can generally average out the surface imperfections. Also makes for generally good enough rough inspections plates for like cast things so you don’t scratch the good one Also your DIY repeatometer is not enough to inspect to even like B grade, a really wide tenths dial works enough to see if you have big dips (anything over like 3 tenths means toast for real flatness)
I've got several grave masonry offcuts. They are all terrible. I can tell they are bad even using a long school ruler. I'm thinking about getting a new float glass piece. My previous one broke.
A sphere would only show no movement if he only rotated his dial indicator holder, he translated (lateral movement) the indicator holder as well as rotate. BUT the calibration process for one of these certified tables involves lasers not indicators. It is generally accepted that the surface grinding of granite countertops is within 1 mil though! Good enough for hobbyists
I wonder if using a straight edge and light would help settle your worries for the precision he seems to be going for since .001 over a plate is acceptable for him. (And reasonably so.)
@@christopherjones7191 Yes, you can measure the deviation with an accurate straight edge and feeler gauges. Another method is to pivot a known accurate straight edge on the plate; if it spins in the middle - the middle is high; if it pivots around the end, you have some convexity. A good straight edge will pivot about 30% from each end on a true flat surface. The tool and method Benjamin shows is just one method used to test plates but it's not definitive and as others have mentioned, other tools such as optical autocollimators, lasers, differential digital levels and straight edges can provide more precise results. ASTM has specific test procedures for calibrating plates.
@@flyingmonkey3822 You are incorrect. This is how convex and hyperbolic / parabolic mirrors and lenses are measured. Benjamin's "Repeat-O-Meter" tool is similar to tools used to measure mirrors for telescopes. If you have a spherical surface, this tool will not show any deviation unless you calibrate the tool against a known reference plane.
Most home machinists don't need and can't achieve any better accuracy than your countertop cutoff. Great demo! I'm working on naval guns that were made between 1910 and 1942 and the quality of machining on the original parts is amazing but as long as I can hit within .0005" the parts fit and work.
Great video mate!
I am learning to hand grind knives and the last thing I need is some moron telling me that I have to have tooling that is accurate to less that a tenth or "Its a waste of time". Your proposal is a practical solution to a real world problem that isn't going to cost $500. Its good to see someone using their brain and not just parroting what they heard on the net.
Thanks.
This one just cropped up on my feed, so yeah, here's me commenting 5 years down the line, not so much for Benjamin, but for those coming later. There's people here commenting that this is not a good test, and to a certain extent they are right. If this is a flat(ish) surface with a maximum deviation of ±25 microns over 10 cm, or a concave spheroid(ish) surface with maximum deviation of ±25 microns over 10 cm, , or a convex spheroid(ish)… you won't be able to tell using this setup. A machinists level can tell you, and an autocollimator can tell you, but that's about it. However, it's unlikely to be a spheroid, which is usually achieved by lapping (see telescope mirrors, etc), it's far more likely to be flat(ish)
But.
If what you want to do with it does not require more accuracy than that, it's fine. And most home shops do not require more accuracy than that for most uses of a "surface plate". Maybe if you're reconditioning machines, or measuring pieces that require high accuracy. Most of us aren't doing that. Most of us will be using it with a height gauge for measuring or marking out. It's fine in that respect. It'll get 90% of what you want done. What I would be more worried about is surface durability.
Anyway, it's a good example of what you can find "over the counter" (if you'll excuse the pun).
Thanks for the clear explanation. I probably don’t need a surface plate but I will keep my eyes open for granite like this.
Might be fun to figure out how inaccurate it is. If I can’t figure out a problem with it, I probably don’t need a better one. :)
@@gf2e Oh, you probably do need a surface plate, even if you don't know it yet. The same way everyone needs a lathe :)
Headstones are often made from granite, and offcuts are cheap.
@@wibblywobblyidiotvision I feel like there is some sort of morbid connection between people getting lathes they don’t need, and tombstones.
I do want a lathe, because I often want weird screws that I can’t easily find locally.
Clearly, I need a Tornos :)
@@gf2e You can measure how flat the surface is with your indicator and a 'reference bar' that covers the surface's diagonals. The bar does not have to be a perfect straightedge, just straight enough to be measured within the range of your indicator. First, measure the bar's thickness variations at the ends and center with a micrometer, or by setting-up the indicator over a single point of contact in the middle of your surface and sliding the bar under the indicator. You don't need to read everywhere (just the ends and middle) and you don't need the actual thicknesses (just the variations measured with your set-up). The difference between the end readings is the bar's 'taper', and the center reading minus the average of the ends is the bar's 'bulge' (or 'neck' if it is negative). Second, support the bar on its ends and zero the indicator at the bar's center. Rotate the bar 180 degrees around the long axis, and read the indicator's change. That is your bar's 'bend'. From the bulge and the bend you can calculate the bow/sag at the center of each of the bar's surfaces. Now you can use the bar to measure the surface's bow/sag in the center, and the 'twist' between the diagonals. You can also characterise the same bar over other distances for the surface's length and width, or the sub-span of a measurement grid.
@@gf2e I find my lathe incredibly useful. One thing to do is start any project by asking yourself 'can it be round'. This is usually a new way of thinking about stuff. If the answer is yes then making it can be joyous lathe work.
Thanks for that. I have a granite cut out from a counter top and it has been very flat. I like the idea of using it for a lapping plate.
Excellent video. I went to the local stone shop and got some of the samples for free. They are excellent for home shops.
Awesome video. I'll have to find a place to get some granite near me. Thanks for sharing.
Picking up 2 cut offs to use for home shop use as well, same no videos before I saw yours, great reference plate!!!
I made 1 from scratch, shooting for .005 for woodworking, but easily got .002. The indicator setup you used cannot pick up widely-spaced holes and high places. I used a home-made precision spirit lever with 3-point contact.
It is most certainly a rough look. Would not pick up a large sphere either as another commenter pointed out.
Lawrence Sparey said a peice of plate glass was what he used. Probably good enough for who it’s for!
Just found this video... Freakin hilarious!! That's great.
Thanks!
I use counter pieces for my sandpaper lapping plates as you can generally average out the surface imperfections. Also makes for generally good enough rough inspections plates for like cast things so you don’t scratch the good one
Also your DIY repeatometer is not enough to inspect to even like B grade, a really wide tenths dial works enough to see if you have big dips (anything over like 3 tenths means toast for real flatness)
I've got several grave masonry offcuts. They are all terrible. I can tell they are bad even using a long school ruler. I'm thinking about getting a new float glass piece. My previous one broke.
You could get 3 and rub them together for even greater accuracy!
Outstanding!
This just a piece of scrap he aquired without lapping it . Lucky that flat
*not good enough* _for the space program_ LOL
A sphere would show the same indicator reading. And, your indicator's resolution is not enough for any meaningful inspection.
I had not thought of it being part of a sphere, nice one.
A sphere would only show no movement if he only rotated his dial indicator holder, he translated (lateral movement) the indicator holder as well as rotate. BUT the calibration process for one of these certified tables involves lasers not indicators. It is generally accepted that the surface grinding of granite countertops is within 1 mil though! Good enough for hobbyists
I wonder if using a straight edge and light would help settle your worries for the precision he seems to be going for since .001 over a plate is acceptable for him. (And reasonably so.)
@@christopherjones7191 Yes, you can measure the deviation with an accurate straight edge and feeler gauges. Another method is to pivot a known accurate straight edge on the plate; if it spins in the middle - the middle is high; if it pivots around the end, you have some convexity. A good straight edge will pivot about 30% from each end on a true flat surface.
The tool and method Benjamin shows is just one method used to test plates but it's not definitive and as others have mentioned, other tools such as optical autocollimators, lasers, differential digital levels and straight edges can provide more precise results. ASTM has specific test procedures for calibrating plates.
@@flyingmonkey3822 You are incorrect. This is how convex and hyperbolic / parabolic mirrors and lenses are measured. Benjamin's "Repeat-O-Meter" tool is similar to tools used to measure mirrors for telescopes. If you have a spherical surface, this tool will not show any deviation unless you calibrate the tool against a known reference plane.
😢😢😢