Definitely. This is one where when your coworker/neighbor comes and says, "Hey, do you have a good spirit level?" The answer is always, "nope, so sorry, cant help".
Lesson learned: i was once given s sphrere containing a tiny shrimp aquarium. I thought it would make for a nice item to keep on my desk at the office. However, every once in a while, a new employee would come into my office and shake it like a snow globe. Those shrimp managed to last a few years, but i hope they didn’t suffer ptsd!
Bonding with Van der Waals forces, you have shown that it can be done!!! It was an experiment that I would have liked to try but on, for example, the fracture of a piece of marble or granite, this has been my work for 45 years, the videos on glass polishing are also interesting, on a different scale, I also polished granite and marble but with diamond abrasives, sometimes a good result is difficult, but with a lot of patience and passion no one has ever complained, congratulations on your knowledge, it's a real joy to see your videos!
Optical contacting typically has 15kg/cm^2 adhesion force if the surfaces are clean and flat to 1/10 lambda. The problem is it is often not resisting to liquid especially low surface tension like methanol or acetone. One way to make optical contacting permanent is to heat it to slightly over annealing temperature or activate the surface with chemicals to provide dangling chemical bond so when the surface contacts it will form a covalent bond over time. So the bond probably failed because of water penetrating the bond line rather than the thermal expansion. You can do a simple test, take two piece of 5mm Borofloat and contact them together, soak in methanol over night, you’ll see the bond degrade or even separate.
Thanks, I actually already made the bond permanent using the method you described. However, I took the temperature a little too high, which caused slight deformation of the glass. And that was also the end of the spirit levels' accuracy...☹️
Huygens Optics too bad it was over heated... another way is chemical activation, you can use a mild sodium hydroxide solution so when the solution drys it react with the glass slightly and create siloxane bond as the water evaporates away.
Beautiful work! Thanks for the awesome content on this channel! edit: There are ways IIRC to calibrate an accelerometer (the device in a smart phone that enables its use as a bevel) that do not require a flat surface at all. There's a fair amount of math involved, but the jist of it is that you collect the uncompensated data from the sensor and rotate the device. If you plot the vectors in 3D you should, in an ideal world get a sphere. In reality, you will get an ovoid. Mathematically you create a compensation to correct it to a sphere. It's tricky, but it can be done and is done to magnetomers in compasses (search for 'hard iron compensation').
Thanks for the insights. You are right. In fact, for a bubble level (I mean the analog one) there are several easy methods to callibrate the zero point. Most of these involve some kind of rotation and comparison/averaging. For this video, I just wanted to show how you can build accuracy intrinsically in your device by tight tolerancing. The smart phone clip in the end was just to demonstrate how inaccurate some accelerometers are mounted inside a smartphone device.
Quite interesting to see all your process. If it was able to have a 6 arcsecond precison with a semi metallized top surface it would probably do as an artificial horizon for getting sextant altitudes on land. I got to your video while searching for ideas for such a tool. All the best!
@@helpmaboab7 Mercury surface “artificial horizons” have been used historically to reflect stars and the sun and get sextant heights, but mercury is nasty stuff to handle.. so the answer to your questions is yes if is big enough but.. not the best substance to handle around even contained in glass. This is just an opinion obviously.
@@tasheido I have a selection of mercury filled glass capsules that I bought to measure temperature. Am I at risk? But you are right that the capsule would have to be big enough to avoid meniscus errors.
Wow, great job! Was the level useful for the carom table? Most of the table mechanics use a Starrett 12" machinist's level because it spans a longer distance of the slate. A shorter level such as a 6" machinist's level keeps them chasing their tail. I'm curious how your level worked out at about 100mm [4"] diameter.
Excellent video! The 'molecular bonding' of the flat surfaces at 9:54 reminds me of how gauge-blocks (or "Jo-Blocks") are assembled by "wringing" them together. I wonder if it is the same process.
I wondered that, too. But then gauge blocks have a micro-thin layer of oil on them. So my next idea was why not clean the glass edge with oil and wring them together. I'll have to watch the video he linked showing the process with glass. Fascinating.
My physics teacher wanted to demonstrate that effect so he put a disc form an old harddisk against the bathroom mirror. It stuck a bit better than he expected. That disk is probably still stuck to that mirror.
Great video. I have a couple of drawing improvement suggestions regarding the geometric tolerancing. You may not care as you were making the parts yourself, however i guess if you don’t care, then why make a drawing at all which includes geometric tolerancing?! Anyway, 1) you have a degree symbol in your parallelism tolerance. Even though it may seem reasonable to think your trying to lock down a maximum angle, parallelism actually works but forcing all points on a surface to lay between to parallel planes which are the tolerance value apart. The two planes are parallel to the datum plane. So no need for a degree symbol as the tolerance zone isn’t an angle. 2) you seem to have combined concentricity and position to govern the location of the curved surface. All you need here is a position tolerance. However to establish the tolerance zone of the sphere shape, you need to add “S ” before the value, and reference 2 datums. 1 datum (A) will only locate the sphere to the OD of the part, but you also need to reference a face to manage the depth the cut will go into the part. Hope all that makes sense. ISO 1101 for more details.
Ha, you caught me! I'm not a real mechanical engineer as you have discovered. Thanks for pointing these out though, I will use the info next time I make a drawing. By the way, I make these drawings for myself. Generally when I design something behind the computer, I take the drawings to the workshop as a memory help. And then as a rule, make parts with slight different dimensions because I'm missing a particular drill or something. ;-)
@@HuygensOptics oh great… we’ll you did a pretty good job then, and certainly know a great deal more about the practice aspects of manufacturing these components than I do. Nice work 👍🏻 I’m sure you don’t need it, but if you’d like any help with the drawings in the future (perhaps you’re getting something made by an outside company), I’d be happy to take a look. Offer is there if you want it, but I won’t be offended if you don’t take it up 😉
The table consists of 3 flat stones and one of these was off level somehow. What you can do is add pieces of paper under the stones to level each stone and match the edges between them exactly. It's actually quite a job.
Yeah, I perfectly know that. Just yesterday I was leveling my snooker table and encountered the exact same problem. One of the stones was leveled just on one end to match the slate on its side, but the other extremity wasn't leveled. So I had to lift this side and put some papers until it was leveled. The thing is that I didn't use a bubble level, I leveled all the slates using a billiard ball because I had no access to a accuracy spirit level.
I would have liked to have seen a comparison with a regular level and compare the two results. The phone level can be calibrated. Maybe I don't understand this enough but I would guess that a regular level would be just as good. Maybe I'm wrong.
How flat a pool table must be is given by the international regulations. The rule is that the maximum distance between the highest and lowest point must be less than 0.5 mm, and way its done is shimming each individual slate section. I use a 12 inch engineer level. This one you made is a little bit too small in diameter for the job, but you could put it into a bigger frame with a couple of adjustment screws to calibrate it.
I couldn't help it... during your awesome video, Rick constantly was in my mind: "I'm familiar with the bubble, Morty. I also dabble in precision, and if you think you can even approach it with your sad, naked, caveman eyeball and a bubble of fucking air, you're the reason this species is a failure, and it makes me angry!" I hope not to offend you. Really great work!
@@HuygensOptics Now I've got your attention, great! :D I have some questions :) How are you dealing with tensions in the glass after the drilling? Is there an increase in tensions in the glass by polishing? And rustling these thoughts, Is this affecting the accuracy? And in minute 8, you show the sphere, and an interference pattern is visible. Do you know the origin? I really like your "wild" idea (optical contact bonding) to connect both sides. Maybe some surface treatment with acid or bases could help. Or using acrylic :D Sorry If my questions are going too much into detail. And thank you for making the video.
Tension due to drilling can only be completely removed by fine grinding and polishing all machined surfaces. I did not do that, so basically I did not worry too much about it. These tensions can deform the parts a few microns, but generally not much more. The interference pattern you mentioned is due to the fact that the part is wet, it is interference in the water film. Optical contact bonding will not work in combination with contact with fluids. Best way is to do thermal bonding of the parts by first contact bonding and then heating to 500-600 deg C to create a permanent bond.
@@HuygensOptics Thank you, for the answer. Okay, I guess the temperature thing might also work. The glass blower told me for removing the tension after drilling, tempering might be a solution, that's why I was curious. Keep up with the great work and videos. Highly informative :)
What you need for ultimate level is an EDS Precision Systems water level setup. I've used one before to level a 4' x 8' gantry router at a local hackerspace, HackPGH in Pittsburgh, PA. This is pretty awesome too though! I love your dedication to flatten pool tables
With deference to the premise of "what's sufficiently level" ... there's also a philosophy of thought that -- a good player can (within reason) adjust their style to minimize the TIME (as I believe the "roll off" is exclusively an issue of how long the ball is rolling before hitting the the 3-rails and the 2nd ball for a billiard) ... Also -- as I'm sure you know -- you can level the table (at the legs) but you can also level areas of the table by drilling through the table and pulling it down in spots.
My question to you is did you perhaps make a mistake in saying microns when you meant "hundreds". At 5:05 drawing you specify 0.005 parallelity. 1 micron is 1/1000 ("thousand") = 0.001 mm. But at 6:18 you said you have 5 microns or less at the edge when your micrometer can only measure up to 0.01 mm of accuracy which is 1 "hundred" and not a micron. Anyways your videos are really awesome. So educational and concise. Thank you for your hard work.
Wow, very sharp. You are right about this type of micrometer, the accuracy is only 0.01 mm (on the scale). I actually also measured the height of the edge and the flatness using a flat granite plate in combination with a vertical 0.0005mm (0.5um) scale micrometer, but unfortunately I did not record any video on that. So, I cheated a bit with the video, showing a shot of the raw thickness measurement. I thought nobody would notice. Damn... ;-).
you said to calibrate a phone you need a horizontal plane that deviates less than 0.01 degree, well, that is probably could be achieved on any reasonably leveled plane, just rotate the phone 180 degrees and you will get amount you need to compensate phone leveling
Keep posting excellent content like this... Soon you will get subscribers and views will go up. Eventually you can afford to purchase some pockets for your dads pool table
Very nice level! I once took a small glass tube, filled it up with water and sealed it up. The bubble is relatively large and sluggish, but if you would replace it with a lower viscosity fluid and a bigger tube, couldn't you make a simple, infinitely resolving bubble level? Because there is no curvature, the bubble would move with the slightest tilt. If it wouldn't move at all, you would be perfectly level, right?!
This assumes that the inside of the tube is perfectly straight and nothing physical is perfectly straight. Light paths in a perfect vacuum come very close, but between diffraction and the uncertainty principle, how would you measure it to be exactly straight? Another problem is that any slight tilt would drive the bubble to one end of the level, which would make it hard to estimate how much you would need to adjust the tilt. This would make it harder to adjust. Finally, gravity is not perfectly uniform on the surface of the earth. There is a curvature equal to the radius of the earth and then distortions due to the local mass distribution. The closest concrete wall would introduce an error.
@@bretcannon3826 Okay, you're right. I shouldn't have used infinity. Perfectness and infinity always makes it difficult ;-) The bubble level I made is relatively short, but If the surface is roughly level, the bubble still needs quite a bit of time to reach one end. You can estimate the error by the speed the bubble is moving. It isn't actually too bad to use. Still, a traditional level is more convenient to use. Yes, finding ideal parts would be impossible and even if you would have a ideal setup, you couldn't simply touch it, because of heat expansion etc. Even your body would be emitting infrared and heat the bubble level unevenly, as well as the surface to be measured.. But roughly speaking for the practical case of leveling the billiard table, this could make it a much easier and precise approach. But admittedly, it wouldn't look as nice.
Nice! I got the chance to play with a spirit level used to align CNC plasma cutters. I could tell if someone farted across the city. It was impossible to align it perfectly.
Impossible to align likely because of thermal expansion. When you get really sensitive with things, your own body heat makes the machine change shape while you're moving around.
This sort of raises the question what is the true shape of a thing or if the true shape of thing actually exists! The shape of an object at that level is more like a probability wave telling us the probability of any given point being on the surface of the part or not. Just a thought.
Pool is one version of a game that uses a table like this. Theres obviously American 8 and 9 ball pool, but there's also 8 ball pool on a smaller table (I want to say its british, maybe it isn't) There's snooker, which Americans often call billiards. Its bigger in europe, australia, and some south east asian countries. It uses a much bigger table, but has very small pockets compared to American pool tables. And then theres billiards, like this, where you dont get points for pocketing balls, you get them for how balls contact eachother or the cushions. Each player can hit their ball (either red or yellow, decided before the game), or the cue ball, and it must hit another ball. Its probably more complicated than I'm making out, but I only know the basics
I play pool in Australia and found this level fascinating. Players are always complaining because they lost a game due to the table not being level. I so wish your level was in production. Or is it?
So, when and where can we buy our custom billiard/pool table levels? 🤣 Seriously, if these were available at a price that was sufficient for you to make some profit, and for some of us to be able to afford, I think you might have a winner!
Beautiful work and really interesting. Do you squash the adhesive between the two parts or is there an adhesive of low enough viscosity to have capillary action draw an adhesive in from the sides/ outer edge / seam ? I'm just wondering how that all levels out ?
I first tried the capillary method but my UV glue was way too viscous for that. So I ended up applying a very minimal layer of glue and use a lot of pressure to make the layer as thin as possible before curing.
A ball bearing might be better than a bubble. However I suppose if you use an electrical insulating type like aluminium oxide you have to worry about electrical charges and electrostatic force, rather than surface tension forces. Some kind of pendulum device might be easier to free from errors. Maybe with a laser diode at the end of the pendulum pointing at a webcam sensor.
I curious if wouldn’t be better thin CA glue to join them together or to avoid any divination “ring” them at the beginning but follow with epoxy around afterwards
That's a lovely lovely product (and maybe something you could sell to billiard players if you could manufacture it more easily)...but you never mentioned the flatness of the billiard table. How did you determine this was the right diameter for the tool?
Actually there's way more to the complexity of the mobile levelling margin of error. Most of it comes from the phone butts themselves being bevelled in all sorts of weird ways and having camera bumps and all the stuff on them. Sides have buttons and aren't always flat too. In addition to that the gyro sensor may not be soldered flush to the PCB. And the butt often is angled against the plane of the frame and the PCB bolted to it. Most chances are that the PCB is parallel to the screen (thanks to the fact that it's usually glued to the same frame) making it the only trustworthy surface, but using the level by plopping the phone on the screen seems impractical to say the least.
Awesome. Your craftsmanship is off the charts. I’m curious about an observation I made during the video. Do the rules of this game differ in Germany? American pool tables have pockets in each corner and halfway along the length. I don’t see those in the example featured in the video. If the goal isn’t to sink ones own billiards into the pockets before the opponent does so with theirs, what is it?
I always appreciate replies from the channels I’m subscribed to. You speak very good English. One could barely tell it’s a second language, and you speak so eloquently.
Very nicely made. Did you have to make any design considerations on how big the contact area between the level and the table needed to be, to average out the (non precision, yielding) surface of the cloth?
If you wanted a tube style spirit level, how long would it have to be to achieve 1/100th of a degree? I assume the same length as the diamiter of your level? Great video, really enjoyed it.
Wondering if maybe it would be possible to use holographic non-destructive testing techniques to make a purely optical leveling system. An exercise for the student.
If you want a seriously accurate level you need to look at machinist's levels. Starrett and/or Mitutoyo are some of the major reputable brands still manufacturing them but you better be ready to throw down some serious cash if you want one of them.
I believe professional snooker/pool referees just use conventional engineers' precision levels (Starrett etc). Some of the super-accurate ones have to have settling times measured in 10s of minutes!
@Andrew H Precisely (pun intended). As the accuracy goes up, the magnification of the angular position to bubble position goes up. This is achieved by making the radius of the curve greater; thus the gravitational force acting on the bubble gets less and less.
Everyone in the comments is like wow that molecular bonding is amazing and yes I agree but I really need one of these to level the pool tables at my local pool hall. Any idea if I can buy one this accurate from somewhere?
But you don't need a perfectly level surface to calibrate a level such as a phones one. You just need a solid unmoving surface and you calibrate by rotating the phone, a accurate level would show same tilt for surface no matter level orientation so that's your source of calibration. No bets on how accurate you can get the results this way, but I suspect it's surprisingly accurate.
you should just project a fringe pattern on the surface and adjust it until the fringes disappear. the fringe pattern would have to be somehow connected to gravity of course. like a grating hanging from a wire or something like that. or even better: make it so the fringes are connected to a water surface, so that if properly aligned, the table surface is exactly parallel to the water surface.
Flatness =/= level. Fringe pattern will give you the flatness, but not the level. Correcting for level can be done by adjusting the legs until a satisfactory result is obtained, provided the table has sufficient rigidity. Flatness would require a surface finishing effort to correct any deviations, which would be substantially more time consuming.
I have a Fell All-Way bullseye level that is accurate to about 8.6 arc seconds over 300mm. It has a much larger bubble than your design. What are the design considerations for bubble size?
That is pretty good, about 4 times better than I designed this one for. About the bubble size: No design parameters here, it was purely trial and error. I quickly found that small bubbles can actually stop moving because of irregularities such as grease or dust particles especially when using water. In my experience the bubble needs to be at least 10-15mm in diameter when using alcohol as liquid. By the way, the bubble size is not constant when varying the temperature.
i think not using glue would have been very cool. maybe in a somewhat reduced pressure while closing the hole? i think of a transparent acrylic "bucket" glued on, and 2 syringes glued into the bucket (one for decompressing, the other to inject epoxy)... just a sudden idea, i guess there are options.
Of course, but probably not in the very near future. I think that a tour would be more appropriate when I have discussed more subjects / equipment in detail.
It doesn't actually require that many tools... if you look around on YT, you will eventually find plenty of solid gold videos on making your own telescope (Newtonian, I assume).
omg. i'm learning so much from your channel. yet the videos are surprisingly well made and satisfying too! thank you and please keep on with the great work.
I’m an astrophotographer (hobbyist) and when I travel I have to make sure my mounting setup is as level as possible. I have been doing this with a carpenter’s level but this is error-prone. Something like this would be amazing Do you know where I could get a sprit level like this?
Starrett makes an inexpensive bulls eye spirit level. I’m sure you can pay more if you need better accuracy. www.starrett.com/metrology/product-detail/BE-1
If the spherical shape in the top disk is not exactly centered to the edges of the disk, it's highes point will not be in the center of the disk. Hence the bubble will also not be exactly centered when the bubble level is placed in the horizontal position. Hope this clarifies it.
@@HuygensOptics Ah, i see, so your goal wasn't just make sure top part is not tilted against bottom part, but also to make sure you can use outer edges of your level as a reference surface for marking) Nice! Where did you learned/got all this knowledge of relationship of metrology and geometry - especially the one you used to pick specific tolerances to meet requirements?
That is actually a nice side effect. You are right about the bubble always being in the center of the sphere if you have a parallel shape as the bottom disk. But in this case, I also wanted the bubble exactly in the center of the disk, not just in the center of the sphere. I guess I acquired the knowledge by making a lot of optics. I'm actually a chemist by education.
But, now the question arises WHY did the table shift? Did the floor shift? Did the foundation of the house shift? Did the earth under the house shift? I'm from an area where sink holes are common, we see something shift? We go looking for why ....and hope we don't get swallered up when we do it.
This is one of those tools, you get a nice wood box lined with velveteen, and keep it hidden in your office drawer so no one else touches it.
Definitely. This is one where when your coworker/neighbor comes and says, "Hey, do you have a good spirit level?" The answer is always, "nope, so sorry, cant help".
Lesson learned: i was once given s sphrere containing a tiny shrimp aquarium. I thought it would make for a nice item to keep on my desk at the office. However, every once in a while, a new employee would come into my office and shake it like a snow globe. Those shrimp managed to last a few years, but i hope they didn’t suffer ptsd!
i know I'm getting old and boring because this video was fascinating. great work
You don't need a flat surface to calibrate a level you just have to make sure that it reads the same angle whenever you rotate it 180°
love this. micron-level precision for something that's going to be laid atop a bumpy felt surface.
Bonding with Van der Waals forces, you have shown that it can be done!!! It was an experiment that I would have liked to try but on, for example, the fracture of a piece of marble or granite, this has been my work for 45 years, the videos on glass polishing are also interesting, on a different scale, I also polished granite and marble but with diamond abrasives, sometimes a good result is difficult, but with a lot of patience and passion no one has ever complained, congratulations on your knowledge, it's a real joy to see your videos!
A carpenter told me when I was a young child... that the bubble is a "Fish Fart" collected by divers in a jar..
Thats fantastic XD
An engineer once told me that carpenters only use levels to make sure nothing they build is level
@@tonybutters1953 An engineer once told me his drawings will work.
Optical contacting typically has 15kg/cm^2 adhesion force if the surfaces are clean and flat to 1/10 lambda. The problem is it is often not resisting to liquid especially low surface tension like methanol or acetone. One way to make optical contacting permanent is to heat it to slightly over annealing temperature or activate the surface with chemicals to provide dangling chemical bond so when the surface contacts it will form a covalent bond over time. So the bond probably failed because of water penetrating the bond line rather than the thermal expansion. You can do a simple test, take two piece of 5mm Borofloat and contact them together, soak in methanol over night, you’ll see the bond degrade or even separate.
Thanks, I actually already made the bond permanent using the method you described. However, I took the temperature a little too high, which caused slight deformation of the glass. And that was also the end of the spirit levels' accuracy...☹️
Huygens Optics too bad it was over heated... another way is chemical activation, you can use a mild sodium hydroxide solution so when the solution drys it react with the glass slightly and create siloxane bond as the water evaporates away.
Yeah...... What that guy said. 🤷🏼♂️ Lol!
Aww. Poor spirit level.
No more excuses for missed shots. Love these videos!
Beautiful work! Thanks for the awesome content on this channel!
edit: There are ways IIRC to calibrate an accelerometer (the device in a smart phone that enables its use as a bevel) that do not require a flat surface at all. There's a fair amount of math involved, but the jist of it is that you collect the uncompensated data from the sensor and rotate the device. If you plot the vectors in 3D you should, in an ideal world get a sphere. In reality, you will get an ovoid. Mathematically you create a compensation to correct it to a sphere. It's tricky, but it can be done and is done to magnetomers in compasses (search for 'hard iron compensation').
Thanks for the insights. You are right. In fact, for a bubble level (I mean the analog one) there are several easy methods to callibrate the zero point. Most of these involve some kind of rotation and comparison/averaging. For this video, I just wanted to show how you can build accuracy intrinsically in your device by tight tolerancing. The smart phone clip in the end was just to demonstrate how inaccurate some accelerometers are mounted inside a smartphone device.
That was impressive enough that it deserves a comment for the algorithm!! Great job 👏
This is such a great engineering problem.. thanks for the explanation
You provided so much truly brilliant information. Thank you for sharing and well done!!!
Quite interesting to see all your process. If it was able to have a 6 arcsecond precison with a semi metallized top surface it would probably do as an artificial horizon for getting sextant altitudes on land. I got to your video while searching for ideas for such a tool.
All the best!
Would a capsule of mercury do?
@@helpmaboab7 Mercury surface “artificial horizons” have been used historically to reflect stars and the sun and get sextant heights, but mercury is nasty stuff to handle.. so the answer to your questions is yes if is big enough but.. not the best substance to handle around even contained in glass. This is just an opinion obviously.
@@tasheido I have a selection of mercury filled glass capsules that I bought to measure temperature. Am I at risk? But you are right that the capsule would have to be big enough to avoid meniscus errors.
So awesome. You need to make some billiard pockets next!
Wow, great job! Was the level useful for the carom table? Most of the table mechanics use a Starrett 12" machinist's level because it spans a longer distance of the slate. A shorter level such as a 6" machinist's level keeps them chasing their tail. I'm curious how your level worked out at about 100mm [4"] diameter.
Excellent video! The 'molecular bonding' of the flat surfaces at 9:54 reminds me of how gauge-blocks (or "Jo-Blocks") are assembled by "wringing" them together. I wonder if it is the same process.
I wondered that, too. But then gauge blocks have a micro-thin layer of oil on them. So my next idea was why not clean the glass edge with oil and wring them together. I'll have to watch the video he linked showing the process with glass. Fascinating.
My physics teacher wanted to demonstrate that effect so he put a disc form an old harddisk against the bathroom mirror. It stuck a bit better than he expected. That disk is probably still stuck to that mirror.
Great video. I have a couple of drawing improvement suggestions regarding the geometric tolerancing. You may not care as you were making the parts yourself, however i guess if you don’t care, then why make a drawing at all which includes geometric tolerancing?! Anyway, 1) you have a degree symbol in your parallelism tolerance. Even though it may seem reasonable to think your trying to lock down a maximum angle, parallelism actually works but forcing all points on a surface to lay between to parallel planes which are the tolerance value apart. The two planes are parallel to the datum plane. So no need for a degree symbol as the tolerance zone isn’t an angle.
2) you seem to have combined concentricity and position to govern the location of the curved surface. All you need here is a position tolerance. However to establish the tolerance zone of the sphere shape, you need to add “S ” before the value, and reference 2 datums. 1 datum (A) will only locate the sphere to the OD of the part, but you also need to reference a face to manage the depth the cut will go into the part. Hope all that makes sense. ISO 1101 for more details.
Ha, you caught me! I'm not a real mechanical engineer as you have discovered. Thanks for pointing these out though, I will use the info next time I make a drawing. By the way, I make these drawings for myself. Generally when I design something behind the computer, I take the drawings to the workshop as a memory help. And then as a rule, make parts with slight different dimensions because I'm missing a particular drill or something. ;-)
@@HuygensOptics oh great… we’ll you did a pretty good job then, and certainly know a great deal more about the practice aspects of manufacturing these components than I do. Nice work 👍🏻 I’m sure you don’t need it, but if you’d like any help with the drawings in the future (perhaps you’re getting something made by an outside company), I’d be happy to take a look. Offer is there if you want it, but I won’t be offended if you don’t take it up 😉
The glue-free electromagnetic bonding also works with metals on a vacuum (cold welding)!
And there's me expecting a laser with various optics as your leveling device!!
Glass surfaces as parallel as gauge blocks. Nice.
optical tools are the best. This is so satisfying to watch and very informative!
I was actually expecting to see you leveling the carom table. How unleveled was the table? We will never know heheh. Great content. Thank you.
The table consists of 3 flat stones and one of these was off level somehow. What you can do is add pieces of paper under the stones to level each stone and match the edges between them exactly. It's actually quite a job.
Yeah, I perfectly know that. Just yesterday I was leveling my snooker table and encountered the exact same problem. One of the stones was leveled just on one end to match the slate on its side, but the other extremity wasn't leveled. So I had to lift this side and put some papers until it was leveled. The thing is that I didn't use a bubble level, I leveled all the slates using a billiard ball because I had no access to a accuracy spirit level.
I would have liked to have seen a comparison with a regular level and compare the two results. The phone level can be calibrated. Maybe I don't understand this enough but I would guess that a regular level would be just as good. Maybe I'm wrong.
How flat a pool table must be is given by the international regulations. The rule is that the maximum distance between the highest and lowest point must be less than 0.5 mm, and way its done is shimming each individual slate section. I use a 12 inch engineer level. This one you made is a little bit too small in diameter for the job, but you could put it into a bigger frame with a couple of adjustment screws to calibrate it.
What a wonderful video. Such a pleasant display of skills and creativity together with diy and astounding results.
TY for showing/sharing it.
Your father must be very proud of you for leveling his table.
Lovely video, thanks for sharing this rarely seen processes!
Really love your videos, thanks for posting this.
I couldn't help it... during your awesome video, Rick constantly was in my mind:
"I'm familiar with the bubble, Morty. I also dabble in precision, and if you think you can even approach it with your sad, naked, caveman eyeball and a bubble of fucking air, you're the reason this species is a failure, and it makes me angry!"
I hope not to offend you. Really great work!
Everything is crooked. Reality is poison.
@@HuygensOptics Now I've got your attention, great! :D
I have some questions :)
How are you dealing with tensions in the glass after the drilling?
Is there an increase in tensions in the glass by polishing?
And rustling these thoughts, Is this affecting the accuracy?
And in minute 8, you show the sphere, and an interference pattern is visible. Do you know the origin?
I really like your "wild" idea (optical contact bonding) to connect both sides.
Maybe some surface treatment with acid or bases could help. Or using acrylic :D
Sorry If my questions are going too much into detail. And thank you for making the video.
rustling = resulting (autocorrect, my bad)
Tension due to drilling can only be completely removed by fine grinding and polishing all machined surfaces. I did not do that, so basically I did not worry too much about it. These tensions can deform the parts a few microns, but generally not much more. The interference pattern you mentioned is due to the fact that the part is wet, it is interference in the water film. Optical contact bonding will not work in combination with contact with fluids. Best way is to do thermal bonding of the parts by first contact bonding and then heating to 500-600 deg C to create a permanent bond.
@@HuygensOptics Thank you, for the answer.
Okay, I guess the temperature thing might also work. The glass blower told me for removing the tension after drilling, tempering might be a solution, that's why I was curious.
Keep up with the great work and videos. Highly informative :)
9:51 the first time in my life that I hear “Van der Waals” pronounced correctly. 😄
How have you heard it pronounced before?
@@arcturuslight_ "Van (the vehicle) der walls" is definitely the most heard, as I've always pronounced. Will correct myself from now on. :D
Amazing work!
What you need for ultimate level is an EDS Precision Systems water level setup. I've used one before to level a 4' x 8' gantry router at a local hackerspace, HackPGH in Pittsburgh, PA. This is pretty awesome too though! I love your dedication to flatten pool tables
damn imagine taking this to your friend's and relatives' houses and pointing out to them how all their surfaces are off by 0.1 degree.
With deference to the premise of "what's sufficiently level" ... there's also a philosophy of thought that -- a good player can (within reason) adjust their style to minimize the TIME (as I believe the "roll off" is exclusively an issue of how long the ball is rolling before hitting the the 3-rails and the 2nd ball for a billiard) ...
Also -- as I'm sure you know -- you can level the table (at the legs) but you can also level areas of the table by drilling through the table and pulling it down in spots.
Great video, tell ya dad that last shot was a peach. Both of you keep up the great work.
Fantastic project, thanks for all the explanations. I don't understand how you sealed the back side whilst keeping it level though.
Dutch man does the entire video in english
End the video with "daaag". Lol.
Great vid, really enjoyed the whole thing.
Can't the phone be calibrated by checking the measurements at various orientations, or at least 2 opposite orientations?
looks like why not
My question to you is did you perhaps make a mistake in saying microns when you meant "hundreds". At 5:05 drawing you specify 0.005 parallelity. 1 micron is 1/1000 ("thousand") = 0.001 mm. But at 6:18 you said you have 5 microns or less at the edge when your micrometer can only measure up to 0.01 mm of accuracy which is 1 "hundred" and not a micron.
Anyways your videos are really awesome. So educational and concise. Thank you for your hard work.
Wow, very sharp. You are right about this type of micrometer, the accuracy is only 0.01 mm (on the scale). I actually also measured the height of the edge and the flatness using a flat granite plate in combination with a vertical 0.0005mm (0.5um) scale micrometer, but unfortunately I did not record any video on that. So, I cheated a bit with the video, showing a shot of the raw thickness measurement. I thought nobody would notice. Damn... ;-).
you said to calibrate a phone you need a horizontal plane that deviates less than 0.01 degree, well, that is probably could be achieved on any reasonably leveled plane, just rotate the phone 180 degrees and you will get amount you need to compensate phone leveling
Keep posting excellent content like this... Soon you will get subscribers and views will go up. Eventually you can afford to purchase some pockets for your dads pool table
Wouldn't the bubble act as sort of an expansion vessel for the expansion of the liquid? What kind of strength/pressure is the bond?
I see you dabble in precision too, good sir... :D
Very nice level! I once took a small glass tube, filled it up with water and sealed it up. The bubble is relatively large and sluggish, but if you would replace it with a lower viscosity fluid and a bigger tube, couldn't you make a simple, infinitely resolving bubble level? Because there is no curvature, the bubble would move with the slightest tilt. If it wouldn't move at all, you would be perfectly level, right?!
This assumes that the inside of the tube is perfectly straight and nothing physical is perfectly straight. Light paths in a perfect vacuum come very close, but between diffraction and the uncertainty principle, how would you measure it to be exactly straight?
Another problem is that any slight tilt would drive the bubble to one end of the level, which would make it hard to estimate how much you would need to adjust the tilt. This would make it harder to adjust.
Finally, gravity is not perfectly uniform on the surface of the earth. There is a curvature equal to the radius of the earth and then distortions due to the local mass distribution. The closest concrete wall would introduce an error.
@@bretcannon3826 Okay, you're right. I shouldn't have used infinity. Perfectness and infinity always makes it difficult ;-)
The bubble level I made is relatively short, but If the surface is roughly level, the bubble still needs quite a bit of time to reach one end. You can estimate the error by the speed the bubble is moving. It isn't actually too bad to use. Still, a traditional level is more convenient to use.
Yes, finding ideal parts would be impossible and even if you would have a ideal setup, you couldn't simply touch it, because of heat expansion etc. Even your body would be emitting infrared and heat the bubble level unevenly, as well as the surface to be measured..
But roughly speaking for the practical case of leveling the billiard table, this could make it a much easier and precise approach. But admittedly, it wouldn't look as nice.
Fantastic stuff as always!
Nice! I got the chance to play with a spirit level used to align CNC plasma cutters. I could tell if someone farted across the city. It was impossible to align it perfectly.
Impossible to align likely because of thermal expansion. When you get really sensitive with things, your own body heat makes the machine change shape while you're moving around.
This sort of raises the question what is the true shape of a thing or if the true shape of thing actually exists! The shape of an object at that level is more like a probability wave telling us the probability of any given point being on the surface of the part or not. Just a thought.
Can't wait for the next next level level
incredible work to play billiards😁
The problem with your dad's billiard table isn't the level, it's the lack of pockets.
Dude I'm looking at that table thinking wtf no pockets, is this a different type of pool game lol
@@SlowerIsFaster139 It's Carom Billiards
Maybe international
Pool is one version of a game that uses a table like this.
Theres obviously American 8 and 9 ball pool, but there's also 8 ball pool on a smaller table (I want to say its british, maybe it isn't)
There's snooker, which Americans often call billiards. Its bigger in europe, australia, and some south east asian countries. It uses a much bigger table, but has very small pockets compared to American pool tables.
And then theres billiards, like this, where you dont get points for pocketing balls, you get them for how balls contact eachother or the cushions. Each player can hit their ball (either red or yellow, decided before the game), or the cue ball, and it must hit another ball. Its probably more complicated than I'm making out, but I only know the basics
Leveled good : pockets
Leveled bad: no pockets
🤣🤣🤣🤣
I would love to see you making a gemstone to the optical specs of a telescope mirror
I play pool in Australia and found this level fascinating. Players are always complaining because they lost a game due to the table not being level. I so wish your level was in production. Or is it?
So, when and where can we buy our custom billiard/pool table levels? 🤣
Seriously, if these were available at a price that was sufficient for you to make some profit, and for some of us to be able to afford, I think you might have a winner!
Beautiful work and really interesting. Do you squash the adhesive between the two parts or is there an adhesive of low enough viscosity to have capillary action draw an adhesive in from the sides/ outer edge / seam ? I'm just wondering how that all levels out ?
I first tried the capillary method but my UV glue was way too viscous for that. So I ended up applying a very minimal layer of glue and use a lot of pressure to make the layer as thin as possible before curing.
I was thinking about create/polish another outside little ring, to accommodate the epoxy
A ball bearing might be better than a bubble. However I suppose if you use an electrical insulating type like aluminium oxide you have to worry about electrical charges and electrostatic force, rather than surface tension forces. Some kind of pendulum device might be easier to free from errors. Maybe with a laser diode at the end of the pendulum pointing at a webcam sensor.
I curious if wouldn’t be better thin CA glue to join them together or to avoid any divination “ring” them at the beginning but follow with epoxy around afterwards
Would love to purchase one of these.
That's a lovely lovely product (and maybe something you could sell to billiard players if you could manufacture it more easily)...but you never mentioned the flatness of the billiard table. How did you determine this was the right diameter for the tool?
Actually there's way more to the complexity of the mobile levelling margin of error.
Most of it comes from the phone butts themselves being bevelled in all sorts of weird ways and having camera bumps and all the stuff on them.
Sides have buttons and aren't always flat too.
In addition to that the gyro sensor may not be soldered flush to the PCB.
And the butt often is angled against the plane of the frame and the PCB bolted to it.
Most chances are that the PCB is parallel to the screen (thanks to the fact that it's usually glued to the same frame) making it the only trustworthy surface, but using the level by plopping the phone on the screen seems impractical to say the least.
Can you teach us how dial indicator gauges are made work? Does this mean machinists spirit level is not as accurate as yours?
Awesome. Your craftsmanship is off the charts.
I’m curious about an observation I made during the video. Do the rules of this game differ in Germany? American pool tables have pockets in each corner and halfway along the length. I don’t see those in the example featured in the video. If the goal isn’t to sink ones own billiards into the pockets before the opponent does so with theirs, what is it?
It is the difference between pool and billiards. Billiards: different game play, no pockets required
I always appreciate replies from the channels I’m subscribed to. You speak very good English. One could barely tell it’s a second language, and you speak so eloquently.
Very nicely made.
Did you have to make any design considerations on how big the contact area between the level and the table needed to be, to average out the (non precision, yielding) surface of the cloth?
Interesting, thank you
Too bad it's gone now. Would have really liked to see a comparison to the digital protractors, like the ubiquitous DXL360S.
If you wanted a tube style spirit level, how long would it have to be to achieve 1/100th of a degree? I assume the same length as the diamiter of your level?
Great video, really enjoyed it.
Wondering if maybe it would be possible to use holographic non-destructive testing techniques to make a purely optical leveling system. An exercise for the student.
I played 3-cushion carom (with diamond calculations) for nearly 10 years, or more. Oh my god how would I LOVE to have a table in my apartment hehe.
Awesome 👍
Nice man. Did you show any part of the video with it on the pool table?
This is so awesome :D
If you want a seriously accurate level you need to look at machinist's levels. Starrett and/or Mitutoyo are some of the major reputable brands still manufacturing them but you better be ready to throw down some serious cash if you want one of them.
I believe professional snooker/pool referees just use conventional engineers' precision levels (Starrett etc). Some of the super-accurate ones have to have settling times measured in 10s of minutes!
@Andrew H Precisely (pun intended).
As the accuracy goes up, the magnification of the angular position to bubble position goes up. This is achieved by making the radius of the curve greater; thus the gravitational force acting on the bubble gets less and less.
@Andrew H Bubble levels use alcohol, or "spirits". That's why they're called "spirit" levels.
Everyone in the comments is like wow that molecular bonding is amazing and yes I agree but I really need one of these to level the pool tables at my local pool hall. Any idea if I can buy one this accurate from somewhere?
But you don't need a perfectly level surface to calibrate a level such as a phones one. You just need a solid unmoving surface and you calibrate by rotating the phone, a accurate level would show same tilt for surface no matter level orientation so that's your source of calibration. No bets on how accurate you can get the results this way, but I suspect it's surprisingly accurate.
That's correct, but that is the easy way. Sometimes the hard way is more fun.
you should just project a fringe pattern on the surface and adjust it until the fringes disappear. the fringe pattern would have to be somehow connected to gravity of course. like a grating hanging from a wire or something like that. or even better: make it so the fringes are connected to a water surface, so that if properly aligned, the table surface is exactly parallel to the water surface.
Flatness =/= level.
Fringe pattern will give you the flatness, but not the level.
Correcting for level can be done by adjusting the legs until a satisfactory result is obtained, provided the table has sufficient rigidity.
Flatness would require a surface finishing effort to correct any deviations, which would be substantially more time consuming.
I have a Fell All-Way bullseye level that is accurate to about 8.6 arc seconds over 300mm. It has a much larger bubble than your design. What are the design considerations for bubble size?
That is pretty good, about 4 times better than I designed this one for. About the bubble size: No design parameters here, it was purely trial and error. I quickly found that small bubbles can actually stop moving because of irregularities such as grease or dust particles especially when using water. In my experience the bubble needs to be at least 10-15mm in diameter when using alcohol as liquid. By the way, the bubble size is not constant when varying the temperature.
I see a market for this for those that set up the world class level snooker tables.
12:56 Or that you forgot to remove the case of the phone.
Awesome! Was there any concern about the epoxy thickness messing with your tolerances?
If the pool table isn't level, put a few beer coasters under one of the legs.
Nice informative video!
i think not using glue would have been very cool. maybe in a somewhat reduced pressure while closing the hole? i think of a transparent acrylic "bucket" glued on, and 2 syringes glued into the bucket (one for decompressing, the other to inject epoxy)... just a sudden idea, i guess there are options.
The real challenge is the leveling of the three slates to each other. We use a 12 inch Starrett level with 0.005 inch resolution per division.
The old timers would use a ball placed directly on the slates. The level is much faster. I have done one with the ball when I forgot the level.
For the smartphone inaccuracies, can't it come from the case
Would you be willing to do a tour of your workshop?
Of course, but probably not in the very near future. I think that a tour would be more appropriate when I have discussed more subjects / equipment in detail.
++ a tour would be great!
Can anyone explain to me why the table has not pockets?
I wish i could tool up to make my own telescope like this.
It doesn't actually require that many tools... if you look around on YT, you will eventually find plenty of solid gold videos on making your own telescope (Newtonian, I assume).
why does the table not have pockets?
omg. i'm learning so much from your channel. yet the videos are surprisingly well made and satisfying too! thank you and please keep on with the great work.
I’m an astrophotographer (hobbyist) and when I travel I have to make sure my mounting setup is as level as possible. I have been doing this with a carpenter’s level but this is error-prone. Something like this would be amazing Do you know where I could get a sprit level like this?
Starrett makes an inexpensive bulls eye spirit level. I’m sure you can pay more if you need better accuracy. www.starrett.com/metrology/product-detail/BE-1
Question: how does absence of centricity may affect accuracy if design consists of two parallel shapes put one on top of another?
If the spherical shape in the top disk is not exactly centered to the edges of the disk, it's highes point will not be in the center of the disk. Hence the bubble will also not be exactly centered when the bubble level is placed in the horizontal position. Hope this clarifies it.
@@HuygensOptics Ah, i see, so your goal wasn't just make sure top part is not tilted against bottom part, but also to make sure you can use outer edges of your level as a reference surface for marking) Nice!
Where did you learned/got all this knowledge of relationship of metrology and geometry - especially the one you used to pick specific tolerances to meet requirements?
That is actually a nice side effect. You are right about the bubble always being in the center of the sphere if you have a parallel shape as the bottom disk. But in this case, I also wanted the bubble exactly in the center of the disk, not just in the center of the sphere.
I guess I acquired the knowledge by making a lot of optics. I'm actually a chemist by education.
No errors or excuses!
That table has no pockets?
Why are there no pockets in that billiard table?
How do you make contact lens
can you buy this somewhere
12:41 It could also be the phone case
Did you ever update your Dell
you guys were so worried about leveling the table you didn't notice the pocket holes are missing from the table!!
Yeah, we only recently discovered this. It's leveled now but still completely useless...
@@HuygensOptics 😂😂
But, now the question arises WHY did the table shift? Did the floor shift? Did the foundation of the house shift? Did the earth under the house shift? I'm from an area where sink holes are common, we see something shift? We go looking for why ....and hope we don't get swallered up when we do it.
i wonder if one could use OPTICAL CONTACT BONDING to make a 3D printer that prints GLASS parts ?