These are my favorite kind of videos. The ones where you think it's going to be simple and you end up going down a rabbit hole and decide bring us along the way. I love it!
Not at ALL! Doing something and getting an unexpected result just shows you have another problem to solve. Maybe the problem is in the process. Maybe it's the math. Maybe it's an incorrect assumption. Either way, it's just more fun stuff to work out and another opportunity to learn.
I pursued your first approach at 5:52 with Pythagoras’ theorem by factoring R out of the root and approximating it as f(x) (where x=r/R, and r=10mm) by its Taylor expansion around zero. f(0)=1, f’(0)=0 and f’’(0)=-1. Putting it all together, I get a second-order approximation of R~r^2/(2d)=50/d, exactly the same result as you got. This also highlights that the value for R is extremely sensitive on r, which highlights the importance of drilling the holes at accurate spacings and means that one should make these spacings as large as possible.
You have this rare gift that you can talk about (almost) anything and make it interesting. Thank you for taking the time to make these videos and share them with us!
Spherometers are pretty awesome tools, loved the video! Their accuracy can be pretty excellent, though it does vary significantly with the sag of the surface and also radial distance between the indicator tip and contact points. You can calibrate your spherometer on a sphere of know radius to get a good estimate of the latter. Been watching for a while, really cool seeing some optics content!
Occasionally when we get lost in the weeds we find a Four Leaf Clover. Please continue to make this type of video. It inspires and justifies (at least for me) the things we do in our shop and why some (all?) projects take twice or more time to complete. Thank you, sir.
I’m glad that you had the time to contemplate the whole scenario about confirming the accuracy of your new glasses. It was fun to watch you work it out into believable numbers, disregarding that you claim to be none the wiser. I applaud you for exercising those space age math skills. Cool. Thx
Watched from beginning to end and only understood half of it, but honestly, watching your working method and the fact you posted this with your raw results was just relaxing to watch and mentally stimulating. Love your videos man, been here 10 years now, still love them
It might be time to exercise your middle/high school math and physics. We all lose skills if we don't use them often, and most of us don't constantly exercise trig and/or optics. Being out of practice can actually effect the way we look at things, since we lose the intuition. The good news is that the internet has PLENTY of free ways to learn/brush up on topics. Not JUST watching video though. Do some homework. Like, work through some actual problems. The good places to learn will provide those resources.
@@Prophes0r I was half joking, so I guess I understood as much as you did. I'm not a nerd, and math was my least favourite subject in school. I did very well in history and, to this day, my favorite subject, so I watch Matthias for pure entertainment not to learn something new
@@gerhardpet1 It wasn't an insult. We really do lose this stuff. As to the Math not being a favorite? I blame the way we teach it. Unfortunately, teaching Math works a LOT like...teaching tools, since the first decade is almost exclusively tools. The first tools are really simple, and also really useful. Just like a hammer and a screwdriver, they can be useful every single day. Unfortunately, after the simple/useful Math, we need to build up years and years more tools until there is a magical moment where the chole combination of things suddenly clicks and it all becomes useful again. And this moment is usually halfway, or more, through college. We spend a decade being taught new tools, then practicing using that specific tool in situations where there is only one right way to do things. Very few of us ever get to the point where suddenly there is a problem, we need a solution, and it is up to us to use whatever tool we want to get there. It really is a shame that for a LONG time, most of the tools of Math don't do much other than let you adjust other parts of Math. A = B/C --> AC = B Just isn't something we are going to apply to much on a day to day basis.
I think this video should be watched from both directions to fully understand, otherwise you only get the average understanding of one that is approximately half.
Every lens has hidden hallmarks in the lens itself to identify them, usually mounted in the frames such that the hallmarks are positioned near the nose bridge. They are very small (less than 2mm high), but If you look closely with the right lighting you can see the symbols. Then you can look up their meanings in industry tables. The markings specify lots of things including the material type, the base curve/magnification and other characteristics. Using this method I was able to determine that My Eye Doctor had definitively delivered to a cheaper lens (Polycarbonate) than I had purchased (Trivex).
I know you're not a 3d printer guy - but that's where my mind goes for a solution to making a precise holder for pins. Your way was very interesting, however, and I'm glad I watched it and learned about it
I have an fdm 3d printer, and I can tell you they're not all that much more accurate than what his setup was doing. The usual dimensional tolerances are around 0.2mm, depending on material, speed and a thousand other variables.
@@akaHarvesteRI agree a 3D printer isn't perfect, but it avoids a number of errors that Matthias gets drilling wood using a very low end pillar drill and xy table (not least the drill probably wanders a bit).. a 3d printer will yield holes that are *very* consistent layer to layer, and if printed say half a mm undersized and then reamed to final size the reamer will average out the radii to yield a hole centered with surprisingly high precision. On the other hand Matthias is Mr Wood Gears, using wood is kind of the law for him, its part of the game 😂.
For those operations where you want accuracy with rough tools you can always make a precise template using a 3d printer with the holes where they need to be and use the template to position the drill being already chucked in the machine. I used this technique for drilling the backplate of a lathe chuck without having a mill and it turned out pretty accurate.
I found it interesting, Maybe if you setup your dial indicator to show table travel of your X/Z table you could get more accurate placement of your holes. Would need 2 indicators to show both directions.
That is a super useful shop-made tool. Being able to reasonably accurately measure curves like this is very hard, particularly when you are measuring transitions in things and don't have a nice simple cylinder or sphere to read. It seems surprisingly consistent for something made out of wood. Good work!
Regardless of any practical application to me, it was still interesting to follow along with your logic and how you were approaching the problem. Thanks for sharing.
certainly fun to watch the process, even if ill never make the gadget. Seeing the "engineering" process of making a thing and diagnosing which of the assumptions you made are wrong is certainly a crucial step and difficult to learn. so watching people do it in a variety of topics is certainly useful, and interesting!
I never believed in the ability to read other"s minds, but now I wonder. You posted this on the day that I had a need for this tool. As we have had the first real snow here in New Brunswick this year I'll just stay in the shop and make another tool.
Nice catch on the roundness of your indicator legs. Would've been a different comment left. I genuinely enjoy watching you work things out. More than one occasion I've learn valuable information from you and avoided mistakes myself.
Yes. It's also an issue that his hand ground roundish legs dont have any precise or even consistent radius tips. I think an improved approach is to use a lathe and center to drill a countersink in the end of the legs and the glue in a ball bearing. They can be had very cheap but are highly polished and hard with precise (thus consistent) known radius
A radio based channel that I watch dropped a line about "FT8 destroying amateur radio" in the middle of a technical presentation on antennas - The comments section went absolutely crazy. As an avid detractor of FT8, I approve of the key in the chuck.
Absolutely fascinating! Us tradesman that barely graduated high school use a Contruction Master calculator to figure all the triangle and circle math. They are absolutely indispensable.
Another idea on accuracy is instead of relying on the hand wheels, rig up a pair of dial indicators in x/y and move the table against them so they read the correct values. i.e.a mini 2 axis analog read out.
yeah, that takes the backlash of the screws and nuts out of it, and measures the true position of the table, not relying on the gauges on the hand wheels that don't correct for backlash
@@gorak9000 it looked to me like Matthias was winding the backlash out. If you turn opposite of the direction you want to go and then wind back that eliminates the backlash.
@@costarich8029 in theory there's no difference between theory and practice, in practice there is. Matthias clearly got good results doing what he did. So he must be doing something right.
@@1pcfred Totally get that. I'm not knocking him, he did great. It was just a suggestion. Second time's the charm. But even winding the backlash out carefully like he was, I'm sure he'd admit that the lead screw is not made to any particular tolerance (based on last month's video of him fussing with the xy table). Imagine for example if the rolling process on that screw formed the threads a bit irregularly. So globally it has say 6TPI but locally it could be pretty erratic and would still 'function'. More or less. A dial indicator on the table itself would let him disregard any sort of error in the screw itself, backlash or not.
most modern lenses are aspherical which makes it impossible to accurately calculate the refractive index without knowing the geometry of the front of the lens. I can't remember the formula but your outer probes are measuring a chord of a circle and your center probe is giving the sagitta. cool video
Great video. Love the math machining and general nerdiness. An idea to get more accuracy from your sliding table is to set up a dial indicator and measure the distance the table actually moves. This would take out any inaccuracies introduced by the dials and lead screws and nuts.
VERY interesting! I've been wondering how I can measure some curved surfaces, but I only have a contour gauge that I thought of for it, but figured it wouldn't be precise enough. Your gauge jig there could be the answer I'm looking for as I just recently bought a digital caliper like the one you used.
Good strategy Matthias! Nothing triggers the comment section like a key left in a chuck :-) Nice work with XY positioning. Very interesting study - worth re-watching!
That was interesting to me as well. I worked in a US Army calibration lab where I used to calculate the inside diameter and taper of ammunition gages by measuring the diameter near each end using gage blocks and two steel balls. I used trig to calculate the diameter and height where the balls touched the inside of the bore. I could then use those values to calculate the diameters at the ends of the bore. We also used a similar device to check the flatness of surface plates.
You must live a hard life. Never seen a drill press without a chuck on it. Onless all your drills are big ones with morse tapers, a drill press without a chuck is not much use.
Just never know what's coming next from Matthias and that's why I love these videos. Will I rush out and make this stuff? No - but it's a breath of fresh air to watch.
That is awesome! I made a similar tester a couple of weeks ago when I had to duplicate a part and needed to calculate a radius on it. Only I kept it to the first version with two feet.
What an entertaining video. More please. Matthias, I know you're not really interested in 3d printing, however, I've had read fun with the precision I can get from some careful prints. I can use them the create accurate spacers, radii, and drilling jigs. Also, I've had some success (better than +/- 0.1mm) with laser printed CAD drawing suck to the work piece. Alignment is with a centre drill and an XY bed mounted on the drill press. I examine the drill point touching the printed template via a phone camera zoomed right in close on the drill tip.
Always love your videos. Loved the homemade spherometer as well. If lense making is something that interests you, you should look into the Amateur Telescope Making community. I’d love to see your approach to grinding a mirror and building a telescope.
I did some telescope mirror making a while ago. I made a couple spherometers, a 3 point like yours and a ring type. I am a machinist, so mine were a little more accurate 😊
Love getting a guided tour of the weeds you got lost in. Sometimes I already know what you will find, where you went wrong etc and others some detail or other was completely new to me. But always satisfying entertaining and informative. (Plus it makes a great distraction from the weeds I've misplaced myself in, for some reason the obviously good idea that refuses to come to you always arrives 5 mins after you 'stop' thinking about it... Or is that just me?)
Interesing, Matthias. I have been thinking of purchasing the x-y base on Vevor myself. I appears to help when hole locations must be pretty accurate. Thanks!
Hi Matthias, interesting to watch a video like this as an Ophthalmologist 😀 What you have build is an ophthalmologic clock just with three standoffs instead of two. Keep in mind however, Trivex is an Index of 1.53 whereas standard polycarbonate is at 1.59 so your measurements have been very accurate indeed!
It's been a long time since I worked in optics but I think you should be able to measure the focal length in air and then measure it in water (in a glass container with straight sides) and then calculate the refractive index from the focal length difference and the difference in refractive index of air vs water. Maybe.
Occasionally, I need to replicate a curve when woodworking. Your device would allow me to compute the radius I need to lay out a good facsimile. I think it would also make a good height gauge for setting up small bits in my router table with a higher degree of repeatability than some of the cheap commercially available ones. Just a couple of potential uses.
There are a lot of easier/cheaper/free ways to solve this particular problem though. The key is to minimize measurement > movement > rounding inaccuracies. Think about how a garage woodworker would cut a circle. Put a pin in the center and use it as the center of rotation to keep the radius constant against the blade. You only need to measure one thing (radius) and that measurement is direct. No conversion inaccuracy. No applying it to another measurement. You can use the same idea to cut [x] holes around a circle at a fixed radius. Use the center hole as the rotation center. Attach a metal ruler to the piece so you can get a better measurement of angle (further out). This should result in precision and repeatability far in excess of what is necessary, considering the dimensional stability of wood, and the deflection of a small drill bit. It isn't the machinist way to do it. But sometimes the machinist way isn't always the best way. (Though "best" is pretty vague)
As a person who bought a dro… that sounds very expensive compared to how cheap that table is. Personally I would maybe consider those drill press quill travel bolt on caliper adaptations. Or actually modding some calipers. Though drilling calipers is no easy feat that ss is hard and work hardens to impossible in my experience.
7:10 bowling ball, perhaps? I imagine a Mohs pick set would probably work as well, but obviously the lenses wouldn't fare too well. This is so far down the rabbit hole that the rabbit is cooked through from the heat. I love it.
Totally as clear as mud! Is this the start of a new & improved turbo-incabulator? I love your woodworking videos. Your attention to detail is beyond me.
Using spherical pins will change your measurement as the point of contact will change depending on the radius. With points you know the point of contact is consistent or at least as minimally different as the point is sharp.
When using these 'value' XY tables you ideally always want to wind past your desired value in one direction so you always arrive on the final value with a clockwise rotation (or anticlockwise, doesn't matter so long as you're consistent). That removes any backlash issues.
XY table accuracy had me consider an idea i had for possibly DIYing high accuracy DIY slide way dials... a Differental windless is a simply way of providing an extremely high mechanical advantage, between rotating a handle, and linear motion. Working the other way... tiny linear motion from lots or revolutions...exactly what's needed for high accuracy dials. With a differential pulley there'd need be some additional guide pullies to ensure angles of the cable loop to "hook pulley" remain constant over range of linear motion. Also.. drive such from headstock of plain lathe... and that would be a way of moving a carriage to cut an accurate screw thread without a master screw.
You tube is dying a slow painful death. I searched for something to watch, found nothing . I guess I’ll watch this video again No disrespect, but this is about the only good channel on here.
I found it interesting! Interestingly I think this is how more advanced maths should be taught, at least for people of my mindset. It's more engaging and more practically based
With spherical contact points of radius r lying on a circle of diameter D, the spherical radius of the test surface (R) is given by: R = H/2 + D^2/8/H -r Where H is the gauge reading when compared to a flat datum. You can make good, accurately spherical contact points using 3 ball bearings sitting in recesses drilled on a known diameter.
You can say Conclusively, that you're measurements were "inconclusive" well done Matthias! kept me hooked for the whole duration! so who's the folly here? you or me! haha. ECF
Modern lenses are almost all cut with all of the astigmatism cut on the back. Also you could get the measurements without having to average the astigmatism on the back of the lens as long as you can line up your instrument( 2 prongs not the three pronged one you made) with the axis of the astigmatism. If you didn’t use an average, your measurements for index of refraction would be more accurate. I could show you how to find the axis but it is hard to explain with words. Maybe I could make a video. Also a lot of lenses made today are aspheric which will be another source of inaccuracy by measuring with a lens indicator (in my industry we call them lens clocks). Any ways that’s my 2 cents.
I'd pretty much use a non permanant marker to mark the axis on the lens first, otherwise trying to get it lined up and measure at the same time would be too much. But I's till have to be very careful to get it perpendicular. Three prongs is just easier. I tried to stay near the center of the lens. Aspheric or not, some distortion towards the edges is not unusual.
The dial scales on the X-Y table are not easy to read accurately. You would get better accuracy by setting up two dial indicators, one for X and one for Y axis. This is what machinists do for precise machining dimensions on metal lathes which do not have DRO's.
"Probably not useful to anybody, but it was interesting to me." That's why we like you, Matthias. Keep it up.
Wouldn't have it any other way!
It was interesting to me, and probably heaps of other viewers, too :)
Matthias, I am pretty sure that this is a very good general description of your whole UA-cam career.
Pretty much sums up all my projects.
I didn't know I was until I heard his genuine interest in the subject and was drawn into the video.
I think this is the nerdiest video by far. And I’ve been watching your videos for 14 years.
for sure, and the nerd in me loved it.
It's sooo good. I don't even understand half of it.
I like it too.
Homer, that isnt very nice.
@@TheLaoruga I'm quite certain it was meant as aa compliment!
"Honey I need to use our kitchen aid to measure the index refraction of my new glasses."
These are my favorite kind of videos. The ones where you think it's going to be simple and you end up going down a rabbit hole and decide bring us along the way. I love it!
Content like this is why I've been watching your channel for the past 15 years. Simple, interesting, and not even clickbaity titles or thumbnails.
Man goes above and beyond, only to realize the data is inconclusive - a tragedy in nineteen parts.
Not at ALL!
Doing something and getting an unexpected result just shows you have another problem to solve.
Maybe the problem is in the process. Maybe it's the math. Maybe it's an incorrect assumption.
Either way, it's just more fun stuff to work out and another opportunity to learn.
Real science is experiments not turning out like expected most of the time xD
Crying all the way to the bank.
@@Prophes0r Tools! We need more tools! Better tools! More Tools!
Spoiler 😢
“Maximum viewer engagement!” Haha! Love it!!
Yeah, i'll engage
He should add some kind of easter egg safety violation in every video, to get the safety nazis riled up. Ha!
@@douglasedward4041 yeeeeaaahhh!! Haha! I’d love that! Heehee!
9:32: "Reading distaance..."
Way over my comprehension, but I still watch and admire every single video.
@@douglasedward4041 REEEEEEEEeeeeeEEEEEE!!!!
As an avid longtime viewer I feel very engaged by that drillchuck key! Thanks for that!
I pursued your first approach at 5:52 with Pythagoras’ theorem by factoring R out of the root and approximating it as f(x) (where x=r/R, and r=10mm) by its Taylor expansion around zero. f(0)=1, f’(0)=0 and f’’(0)=-1. Putting it all together, I get a second-order approximation of R~r^2/(2d)=50/d, exactly the same result as you got. This also highlights that the value for R is extremely sensitive on r, which highlights the importance of drilling the holes at accurate spacings and means that one should make these spacings as large as possible.
And for the tri posts to have a precision point instead of a round point.
Can we not just solve the equation at 5:52 exactly for R? R=(d^2+100)/(2d). The expansion is nice to see the dependence on r though.
The hole are drilled to an accuracy of .01mm... maybe, and the the tip of the rods are hand made with no control. Oh well, it was fun to do.
You have this rare gift that you can talk about (almost) anything and make it interesting. Thank you for taking the time to make these videos and share them with us!
At first glance of the thumbnail I thought you made an electrical plug. Love the video!
I should make one, then use it electroboom style!
Spherometers are pretty awesome tools, loved the video! Their accuracy can be pretty excellent, though it does vary significantly with the sag of the surface and also radial distance between the indicator tip and contact points. You can calibrate your spherometer on a sphere of know radius to get a good estimate of the latter. Been watching for a while, really cool seeing some optics content!
You can’t check for circularity by measuring diameters (in old school submarines, for example). Curves of constant width are avoided with this method.
Occasionally when we get lost in the weeds we find a Four Leaf Clover. Please continue to make this type of video. It inspires and justifies (at least for me) the things we do in our shop and why some (all?) projects take twice or more time to complete. Thank you, sir.
Matthias, you don't need me to tell you this, but you're a damned genius thinker.
I’m glad that you had the time to contemplate the whole scenario about confirming the accuracy of your new glasses. It was fun to watch you work it out into believable numbers, disregarding that you claim to be none the wiser. I applaud you for exercising those space age math skills. Cool. Thx
Watched from beginning to end and only understood half of it, but honestly, watching your working method and the fact you posted this with your raw results was just relaxing to watch and mentally stimulating. Love your videos man, been here 10 years now, still love them
I watched the whole video too and didn't understand any of it 😂
It might be time to exercise your middle/high school math and physics.
We all lose skills if we don't use them often, and most of us don't constantly exercise trig and/or optics.
Being out of practice can actually effect the way we look at things, since we lose the intuition.
The good news is that the internet has PLENTY of free ways to learn/brush up on topics.
Not JUST watching video though. Do some homework. Like, work through some actual problems.
The good places to learn will provide those resources.
@@Prophes0r I was half joking, so I guess I understood as much as you did. I'm not a nerd, and math was my least favourite subject in school. I did very well in history and, to this day, my favorite subject, so I watch Matthias for pure entertainment not to learn something new
@@gerhardpet1 It wasn't an insult. We really do lose this stuff.
As to the Math not being a favorite? I blame the way we teach it.
Unfortunately, teaching Math works a LOT like...teaching tools, since the first decade is almost exclusively tools.
The first tools are really simple, and also really useful. Just like a hammer and a screwdriver, they can be useful every single day.
Unfortunately, after the simple/useful Math, we need to build up years and years more tools until there is a magical moment where the chole combination of things suddenly clicks and it all becomes useful again. And this moment is usually halfway, or more, through college.
We spend a decade being taught new tools, then practicing using that specific tool in situations where there is only one right way to do things.
Very few of us ever get to the point where suddenly there is a problem, we need a solution, and it is up to us to use whatever tool we want to get there.
It really is a shame that for a LONG time, most of the tools of Math don't do much other than let you adjust other parts of Math.
A = B/C --> AC = B Just isn't something we are going to apply to much on a day to day basis.
I think this video should be watched from both directions to fully understand, otherwise you only get the average understanding of one that is approximately half.
Every lens has hidden hallmarks in the lens itself to identify them, usually mounted in the frames such that the hallmarks are positioned near the nose bridge. They are very small (less than 2mm high), but If you look closely with the right lighting you can see the symbols. Then you can look up their meanings in industry tables. The markings specify lots of things including the material type, the base curve/magnification and other characteristics. Using this method I was able to determine that My Eye Doctor had definitively delivered to a cheaper lens (Polycarbonate) than I had purchased (Trivex).
I love your mind, sir. I love watching your thinking process and how you get from a to b. I’ve learned a lot from watching you. Thanks man.
I know you're not a 3d printer guy - but that's where my mind goes for a solution to making a precise holder for pins. Your way was very interesting, however, and I'm glad I watched it and learned about it
I have an fdm 3d printer, and I can tell you they're not all that much more accurate than what his setup was doing.
The usual dimensional tolerances are around 0.2mm, depending on material, speed and a thousand other variables.
@@akaHarvesteRI agree a 3D printer isn't perfect, but it avoids a number of errors that Matthias gets drilling wood using a very low end pillar drill and xy table (not least the drill probably wanders a bit).. a 3d printer will yield holes that are *very* consistent layer to layer, and if printed say half a mm undersized and then reamed to final size the reamer will average out the radii to yield a hole centered with surprisingly high precision.
On the other hand Matthias is Mr Wood Gears, using wood is kind of the law for him, its part of the game 😂.
For those operations where you want accuracy with rough tools you can always make a precise template using a 3d printer with the holes where they need to be and use the template to position the drill being already chucked in the machine. I used this technique for drilling the backplate of a lathe chuck without having a mill and it turned out pretty accurate.
I found it interesting,
Maybe if you setup your dial indicator to show table travel of your X/Z table you could get more accurate placement of your holes. Would need 2 indicators to show both directions.
Was going to say this too, although you would need an indicator with a fairly long travel for the x axis.
That is a super useful shop-made tool. Being able to reasonably accurately measure curves like this is very hard, particularly when you are measuring transitions in things and don't have a nice simple cylinder or sphere to read. It seems surprisingly consistent for something made out of wood. Good work!
Regardless of any practical application to me, it was still interesting to follow along with your logic and how you were approaching the problem. Thanks for sharing.
certainly fun to watch the process, even if ill never make the gadget. Seeing the "engineering" process of making a thing and diagnosing which of the assumptions you made are wrong is certainly a crucial step and difficult to learn. so watching people do it in a variety of topics is certainly useful, and interesting!
I never believed in the ability to read other"s minds, but now I wonder. You posted this on the day that I had a need for this tool. As we have had the first real snow here in New Brunswick this year I'll just stay in the shop and make another tool.
Nice catch on the roundness of your indicator legs. Would've been a different comment left. I genuinely enjoy watching you work things out. More than one occasion I've learn valuable information from you and avoided mistakes myself.
Yes. It's also an issue that his hand ground roundish legs dont have any precise or even consistent radius tips. I think an improved approach is to use a lathe and center to drill a countersink in the end of the legs and the glue in a ball bearing. They can be had very cheap but are highly polished and hard with precise (thus consistent) known radius
A radio based channel that I watch dropped a line about "FT8 destroying amateur radio" in the middle of a technical presentation on antennas - The comments section went absolutely crazy.
As an avid detractor of FT8, I approve of the key in the chuck.
Absolutely fascinating! Us tradesman that barely graduated high school use a Contruction Master calculator to figure all the triangle and circle math. They are absolutely indispensable.
VERY ENTERTAINING. I love science, math and wood working. Your videos are perfect.
I like tinkering and sometimes surprise myself with figuring stuff out but you sir, take it to a whole new level.
You would use a Geneva Lens Measure which reads out directly in diopters.
Yes! This should be the top comment. Geneva Lens Measure. This measures surface curvature of lenses and that is what Matias is looking for.
The adventures you get into when exploring these kinds of things... I love it.
The journey is why were here. Inconclusive results are still entertaining. Thanks Matthias!
I loved this video. Love the super accurate drill press. And enjoyed the adventure. Nice job!
Another idea on accuracy is instead of relying on the hand wheels, rig up a pair of dial indicators in x/y and move the table against them so they read the correct values. i.e.a mini 2 axis analog read out.
yeah, that takes the backlash of the screws and nuts out of it, and measures the true position of the table, not relying on the gauges on the hand wheels that don't correct for backlash
@@gorak9000 it looked to me like Matthias was winding the backlash out. If you turn opposite of the direction you want to go and then wind back that eliminates the backlash.
@@1pcfred Yeah he was. But still it would take some of the variability out.
@@costarich8029 in theory there's no difference between theory and practice, in practice there is. Matthias clearly got good results doing what he did. So he must be doing something right.
@@1pcfred Totally get that. I'm not knocking him, he did great. It was just a suggestion. Second time's the charm. But even winding the backlash out carefully like he was, I'm sure he'd admit that the lead screw is not made to any particular tolerance (based on last month's video of him fussing with the xy table). Imagine for example if the rolling process on that screw formed the threads a bit irregularly. So globally it has say 6TPI but locally it could be pretty erratic and would still 'function'. More or less. A dial indicator on the table itself would let him disregard any sort of error in the screw itself, backlash or not.
most modern lenses are aspherical which makes it impossible to accurately calculate the refractive index without knowing the geometry of the front of the lens. I can't remember the formula but your outer probes are measuring a chord of a circle and your center probe is giving the sagitta. cool video
Great video. Love the math machining and general nerdiness. An idea to get more accuracy from your sliding table is to set up a dial indicator and measure the distance the table actually moves. This would take out any inaccuracies introduced by the dials and lead screws and nuts.
This was absolutely fascinating to watch. I like to get wrapped around the axle doing this kind of thing myself - drives my wife nuts!
Sometimes it's about the journey, not the destination. Thanks for letting us tag along.
VERY interesting! I've been wondering how I can measure some curved surfaces, but I only have a contour gauge that I thought of for it, but figured it wouldn't be precise enough. Your gauge jig there could be the answer I'm looking for as I just recently bought a digital caliper like the one you used.
Very interesting measurements! I have never thought of doing anything like this.
Good strategy Matthias! Nothing triggers the comment section like a key left in a chuck :-) Nice work with XY positioning. Very interesting study - worth re-watching!
That was interesting to me as well. I worked in a US Army calibration lab where I used to calculate the inside diameter and taper of ammunition gages by measuring the diameter near each end using gage blocks and two steel balls. I used trig to calculate the diameter and height where the balls touched the inside of the bore. I could then use those values to calculate the diameters at the ends of the bore. We also used a similar device to check the flatness of surface plates.
Seeing that chuck in the drill press just ruined my day, and I doubt I will sleep tonight
You must live a hard life. Never seen a drill press without a chuck on it. Onless all your drills are big ones with morse tapers, a drill press without a chuck is not much use.
@@matthiaswandel Nah, my press has a special clip holder next to the handle for the chuck to sit in when not in use. Otherwise it's OCD overload
@@matthiaswandelI suspect they mean the chuck key.
Just never know what's coming next from Matthias and that's why I love these videos. Will I rush out and make this stuff? No - but it's a breath of fresh air to watch.
That is awesome! I made a similar tester a couple of weeks ago when I had to duplicate a part and needed to calculate a radius on it. Only I kept it to the first version with two feet.
You are genuinely my single biggest inspiration. I love these videos!
What an entertaining video. More please. Matthias, I know you're not really interested in 3d printing, however, I've had read fun with the precision I can get from some careful prints. I can use them the create accurate spacers, radii, and drilling jigs. Also, I've had some success (better than +/- 0.1mm) with laser printed CAD drawing suck to the work piece. Alignment is with a centre drill and an XY bed mounted on the drill press. I examine the drill point touching the printed template via a phone camera zoomed right in close on the drill tip.
I'm grateful there are people like you in the world
Brilliant, shows the depth of thinking that is needed in this world
Always love your videos. Loved the homemade spherometer as well. If lense making is something that interests you, you should look into the Amateur Telescope Making community. I’d love to see your approach to grinding a mirror and building a telescope.
One of your best videos. I enjoyed it immensely and got to learn new stuff about refraction!
I did some telescope mirror making a while ago. I made a couple spherometers, a 3 point like yours and a ring type. I am a machinist, so mine were a little more accurate 😊
Love getting a guided tour of the weeds you got lost in. Sometimes I already know what you will find, where you went wrong etc and others some detail or other was completely new to me. But always satisfying entertaining and informative. (Plus it makes a great distraction from the weeds I've misplaced myself in, for some reason the obviously good idea that refuses to come to you always arrives 5 mins after you 'stop' thinking about it... Or is that just me?)
I thoroughly enjoy glimpses of how intelligent people's minds work.
Interesing, Matthias. I have been thinking of purchasing the x-y base on Vevor myself. I appears to help when hole locations must be pretty accurate. Thanks!
You are like a REALLY smart person. . . Right? Loved it.
Not useful? Maybe not but interesting as hell. Your mind is a joy to behold.
This is actually a really cool video, never even considered such a thing before
I'm currently studying all of this math in school and it's so cool to see applications for it
Hi Matthias, interesting to watch a video like this as an Ophthalmologist 😀 What you have build is an ophthalmologic clock just with three standoffs instead of two.
Keep in mind however, Trivex is an Index of 1.53 whereas standard polycarbonate is at 1.59 so your measurements have been very accurate indeed!
I can usually keep up with your thinking on your videos, but you left me in the weeds on this one.
I enjoyed it none the less
Fascinating indeed, Matthias! 😊
Stay safe there with your family! 🖖😊
It's been a long time since I worked in optics but I think you should be able to measure the focal length in air and then measure it in water (in a glass container with straight sides) and then calculate the refractive index from the focal length difference and the difference in refractive index of air vs water. Maybe.
Occasionally, I need to replicate a curve when woodworking. Your device would allow me to compute the radius I need to lay out a good facsimile. I think it would also make a good height gauge for setting up small bits in my router table with a higher degree of repeatability than some of the cheap commercially available ones. Just a couple of potential uses.
The easiest way to fix that tables accuracy is to just get a decent 2 axis DRO, with that you should be able to get extremely accurate hole placement.
There are a lot of easier/cheaper/free ways to solve this particular problem though.
The key is to minimize measurement > movement > rounding inaccuracies.
Think about how a garage woodworker would cut a circle. Put a pin in the center and use it as the center of rotation to keep the radius constant against the blade. You only need to measure one thing (radius) and that measurement is direct. No conversion inaccuracy. No applying it to another measurement.
You can use the same idea to cut [x] holes around a circle at a fixed radius.
Use the center hole as the rotation center.
Attach a metal ruler to the piece so you can get a better measurement of angle (further out).
This should result in precision and repeatability far in excess of what is necessary, considering the dimensional stability of wood, and the deflection of a small drill bit.
It isn't the machinist way to do it.
But sometimes the machinist way isn't always the best way. (Though "best" is pretty vague)
As a person who bought a dro… that sounds very expensive compared to how cheap that table is. Personally I would maybe consider those drill press quill travel bolt on caliper adaptations. Or actually modding some calipers. Though drilling calipers is no easy feat that ss is hard and work hardens to impossible in my experience.
7:10 bowling ball, perhaps?
I imagine a Mohs pick set would probably work as well, but obviously the lenses wouldn't fare too well.
This is so far down the rabbit hole that the rabbit is cooked through from the heat. I love it.
Totally as clear as mud! Is this the start of a new & improved turbo-incabulator? I love your woodworking videos. Your attention to detail is beyond me.
Using spherical pins will change your measurement as the point of contact will change depending on the radius. With points you know the point of contact is consistent or at least as minimally different as the point is sharp.
8:34
You are too smart for me...way over my head. I did enjoy watching you work your magic.
This video was incredibly unuseful to me, and I loved it.
I'll mention the chuck key as you went to so much trouble to engage me :)
You may have went down the rabbit hole, but we're right behind you.
When using these 'value' XY tables you ideally always want to wind past your desired value in one direction so you always arrive on the final value with a clockwise rotation (or anticlockwise, doesn't matter so long as you're consistent). That removes any backlash issues.
"Viewer engagement" -- great trolling Matthias
Trivex is an excellent material. Switched for chroma and never looked back.
XY table accuracy had me consider an idea i had for possibly DIYing high accuracy DIY slide way dials...
a Differental windless is a simply way of providing an extremely high mechanical advantage, between rotating a handle, and linear motion. Working the other way... tiny linear motion from lots or revolutions...exactly what's needed for high accuracy dials.
With a differential pulley there'd need be some additional guide pullies to ensure angles of the cable loop to "hook pulley" remain constant over range of linear motion.
Also.. drive such from headstock of plain lathe... and that would be a way of moving a carriage to cut an accurate screw thread without a master screw.
You tube is dying a slow painful death. I searched for something to watch, found nothing . I guess I’ll watch this video again
No disrespect, but this is about the only good channel on here.
I loved the journey. Thanks for taking us along for the ride 😂😂😂❤
Time to buy a milling machine for accurate holes and was painful watching you making a round rod manually with a lathe at home. Thanks for the video
A milling machine is no more accurate than that X-Y positioning table is.
Good to have you back!
I found it interesting! Interestingly I think this is how more advanced maths should be taught, at least for people of my mindset. It's more engaging and more practically based
Watching Matthias vidoes is truly humbling. 😆
I FRICKIN KNEW IT! I hate the ol' youtuber "leave the chuck key in the drill" to get annoyed comments to feed the algorithm gambit.
DIABOLICAL sir!
This is amazing, Mr. Wandel.
With spherical contact points of radius r lying on a circle of diameter D, the spherical radius of the test surface (R) is given by:
R = H/2 + D^2/8/H -r
Where H is the gauge reading when compared to a flat datum.
You can make good, accurately spherical contact points using 3 ball bearings sitting in recesses drilled on a known diameter.
Im really impressed with your math skills for your age. Be myself im interested in math especially calculus
Have no idea what I watched but, enjoyed it as usual.
You are out of your mind.
That's why I love your videos! :)
You can say Conclusively, that you're measurements were "inconclusive" well done Matthias! kept me hooked for the whole duration! so who's the folly here? you or me! haha. ECF
Modern lenses are almost all cut with all of the astigmatism cut on the back. Also you could get the measurements without having to average the astigmatism on the back of the lens as long as you can line up your instrument( 2 prongs not the three pronged one you made) with the axis of the astigmatism. If you didn’t use an average, your measurements for index of refraction would be more accurate. I could show you how to find the axis but it is hard to explain with words. Maybe I could make a video. Also a lot of lenses made today are aspheric which will be another source of inaccuracy by measuring with a lens indicator (in my industry we call them lens clocks). Any ways that’s my 2 cents.
I'd pretty much use a non permanant marker to mark the axis on the lens first, otherwise trying to get it lined up and measure at the same time would be too much. But I's till have to be very careful to get it perpendicular. Three prongs is just easier. I tried to stay near the center of the lens. Aspheric or not, some distortion towards the edges is not unusual.
Ladies & gentlemen, the smartest man on youtube.
The dial scales on the X-Y table are not easy to read accurately. You would get better accuracy by setting up two dial indicators, one for X and one for Y axis. This is what machinists do for precise machining dimensions on metal lathes which do not have DRO's.
Very interesting, thanks for posting it. Btw "lost in the weeds" is something I do quite often😂
To state the obvious.....Matthias is a smart man!!!
They say I might be crazy. I'm glad that I'm not the only one. great video.
Love this sort of content , reminds me of old UA-cam
Thumbs up for key in the chuck video engagement.
I like that it began with equations for spheres and ended with triangles in a spreadsheet.Though, who doesn't love spreadsheets?
I appreciate your effort, and enjoyed watching. My inner underachiever made me go to ebay and buy a diopter gauge.