OK guys, It's time to get out the big guns.
I am developing plans for a moderately sized home DIY CNC lens grinding machine. It will be capable of grinding lenses of any shape up to about 12 inches in diameter. It will use captured air bearings for the ways, the turntable and the spindle motor and non contact bearingless torque motors. It will require an air compressor and a PC or old notebook computer to control it's two axis motion.
I figure it's about time we got serious with this hobby and besides I'm studing to be a machine designer so it is an interesting project. I have idea sketches for every part except how to impliment the ball screws. But I think that will come also as I am visualizing something right now even as I type this.
I will be posting pictures of the design from time to time as it developes. Several of the parts will require CNC machining to build but might be able to be duplicated on manual machines with some extra work.
Some of you out there might have better resources to machine tools but when I go back to school I will have some access myself and may try to make this into my senior project. If I ever get that far.
Hezz
I am developing plans for a moderately sized home DIY CNC lens grinding machine. It will be capable of grinding lenses of any shape up to about 12 inches in diameter. It will use captured air bearings for the ways, the turntable and the spindle motor and non contact bearingless torque motors. It will require an air compressor and a PC or old notebook computer to control it's two axis motion.
I figure it's about time we got serious with this hobby and besides I'm studing to be a machine designer so it is an interesting project. I have idea sketches for every part except how to impliment the ball screws. But I think that will come also as I am visualizing something right now even as I type this.
I will be posting pictures of the design from time to time as it developes. Several of the parts will require CNC machining to build but might be able to be duplicated on manual machines with some extra work.
Some of you out there might have better resources to machine tools but when I go back to school I will have some access myself and may try to make this into my senior project. If I ever get that far.
Hezz
If you need a hand hezz ill help ya in this, ive built alot of machines before mainly presses, mobile cranes, allsorts of robotic machines, you will need a controller card and sensors, possibly stepper motors too, whats the rpm on the grinding head and how fast you want it to go?
Trev
Trev
Trev,
The machine can be implimented with stepper motors but that would be the last resort. I am hoping to get some used small servos from ebay or something and some ground ball screws which are very expensive if you buy them new. Since the grinding head spindle will be turning on a bed of air it will have very low vibration and can easily go up to 60,000 rpm depending on how good our machining is. In reality to keep costs reasonable we won't be able to get the highest quality machining so say maybe 20,000 RPM for the grinding motor. The lens turntable will only need to go about 100 - 500 RPM.
One option is to make our own linear motors for the ways instead of using lead screws or ball screws. This would be the ideal thing as no touching parts. The linear motors would have to be embedded with some kind of positioning encoder or something.
To get the highest possible surface finishes on the lens we need to design out all vibrations and machinical slop. Even if our machine does not approach the best micro machining centers it will grind a lens better than a standard balanced CNC mill would be capable of.
I will definately need help with the electronics. The turntable where the lens blanks will be mounted is kind of like a phonograph turntable except that it has a radial linear motor around the outside edge. The easiest and cheapest way to impliment this motor is with flat flexible permanent magnets and an AC winding. The speed is varied by increasing the AC frequency.
I am thinking of using this technique for the motor spindle also since it is essentailly a grinding machine it does not need really accurate control over the motor such as a CNC mill would. It only needs to be able to hold a set speed with some precision.
Hezz
The machine can be implimented with stepper motors but that would be the last resort. I am hoping to get some used small servos from ebay or something and some ground ball screws which are very expensive if you buy them new. Since the grinding head spindle will be turning on a bed of air it will have very low vibration and can easily go up to 60,000 rpm depending on how good our machining is. In reality to keep costs reasonable we won't be able to get the highest quality machining so say maybe 20,000 RPM for the grinding motor. The lens turntable will only need to go about 100 - 500 RPM.
One option is to make our own linear motors for the ways instead of using lead screws or ball screws. This would be the ideal thing as no touching parts. The linear motors would have to be embedded with some kind of positioning encoder or something.
To get the highest possible surface finishes on the lens we need to design out all vibrations and machinical slop. Even if our machine does not approach the best micro machining centers it will grind a lens better than a standard balanced CNC mill would be capable of.
I will definately need help with the electronics. The turntable where the lens blanks will be mounted is kind of like a phonograph turntable except that it has a radial linear motor around the outside edge. The easiest and cheapest way to impliment this motor is with flat flexible permanent magnets and an AC winding. The speed is varied by increasing the AC frequency.
I am thinking of using this technique for the motor spindle also since it is essentailly a grinding machine it does not need really accurate control over the motor such as a CNC mill would. It only needs to be able to hold a set speed with some precision.
Hezz
yeah i got ya, well your certainly on the right track working the motor on frequency rather then voltage, u dont loose power running them this way, you could run a 3000rpm motor on voltage at that speed and have maximum torque while you could run the same motor at 40rpm and maintain the same maximum torque if its run on frequency. Ive got a cnc router project comming up back in aust, it will be simple enough to make, but it will have to be made strong in all elements so it will be acurate under the given load. If you got any pics of your design post them up so i can have a better look at how this thing will work, im sure we could break it down to be simpler then what you say and yet have the same results in the end product maintaining acuracy.
Trev
Trev
Trev,
I am trying to design it as simple as possible. I havn't got any solid models yet except for a basic frame and I wanted to sketch out the bearing and motor elements before I figured out where all of the frame members need to go. My feeling is that a welded frame of thick walled tubular steel would be very easy to make both ridgid and reasonably light as opposed to castings. No this thing has to be able to be moved out into a garage or storage shed when not being used.
Hezz
I am trying to design it as simple as possible. I havn't got any solid models yet except for a basic frame and I wanted to sketch out the bearing and motor elements before I figured out where all of the frame members need to go. My feeling is that a welded frame of thick walled tubular steel would be very easy to make both ridgid and reasonably light as opposed to castings. No this thing has to be able to be moved out into a garage or storage shed when not being used.
Hezz
Hey yeah tubes are way strong if u brace it in the right places, id just use 20x20mm 2mm thick walled square tube for this, its cheap and very strong.
Wow i never knew u would need to run the machine at 20000rpm, thats kind of scary, i thought they ran slower then that, around the 1500rpm range lol, to tell you the truth hezz i doubt u need any cnc gear, if u make up the tooling to do the right angle of the lens then thats all you should need unless your making concave lenses.
I did see a lens machine on the net somwhere a while ago for making lenses and telescope reflectors, have you run up a search on this? one guy had a real buety of a machine just made out of wood lol, but it worked and i think its right up your aley, ill have a hunt for it, i think he used a washing machine motor for the turntable.
Trev
hezz here is one here, not the same as i saw before though but its a start for ya. Also some interesting reads.
http://www.carlb.de/Grinding.html
http://www.turbofast.com.au/astrotel/mirrorgrinding.html
http://www.canon.com/camera-museum/tech/l_plant/kakou/ka_1/kakou1.html
http://www.opticsexcellence.org/InfoAboutCom/Articles/BeatingTheGrind.pdf
Trev
http://www.carlb.de/Grinding.html
http://www.turbofast.com.au/astrotel/mirrorgrinding.html
http://www.canon.com/camera-museum/tech/l_plant/kakou/ka_1/kakou1.html
http://www.opticsexcellence.org/InfoAboutCom/Articles/BeatingTheGrind.pdf
Trev
Trev,
thanks for all the links. I have studied most of those designs based on the Waineo (or something like that) lens grinding design. These work well for telescope makers because they only need very long focal length lenses. As soon as the curve becomes more than a small amount those designs don't work well. I know my idea is a little high but with it we could make any kind of lens even aspheric lenses. As for speed grinders typically run at high rpm's but there is no reason we could not run slower for hogging operations.
Here is a quick sketch of the turntable bearing design. It is deceptively simple. It uses a large cylindrical round ended axle for a large surface area to resist larger axial side loads.
thanks for all the links. I have studied most of those designs based on the Waineo (or something like that) lens grinding design. These work well for telescope makers because they only need very long focal length lenses. As soon as the curve becomes more than a small amount those designs don't work well. I know my idea is a little high but with it we could make any kind of lens even aspheric lenses. As for speed grinders typically run at high rpm's but there is no reason we could not run slower for hogging operations.
Here is a quick sketch of the turntable bearing design. It is deceptively simple. It uses a large cylindrical round ended axle for a large surface area to resist larger axial side loads.
Attachments
Too true but you can modd the design, use shorter arms to give you a wider throw, the smaller the arms the better from what ive read as in acuracy. You can buy a lens grinding machine for $8k, not bad considering what it costs for lenses, CNC too.
Trev
Btw ive got you email too thank you very much for it, ive sent you a few over time and now i know why you didnt get them lol.
Nmd,
I think the parts in any hard drive would be too small. We have high loads and forces compared to a hard drive so everything must be enlarged and created to handle the kinds of forces that occur during machining.
The interesting thing about air bearings is that they are actually more simple that regular mechanical bearing but the system needs higher maching accuracy.
Ace,
I want to build something that is faster and more precise than those waineo type grinders. You have to be constantly adjusting things and rebuilding the lens grinding tool. If you don't have the thing set up exactly right it will not make the right curvature that you wanted. I want to make a single point grinding machine that uses a diamond grinding wheel or any other type of wheel that you might want to use.
I want to be able to grind a lens in two to four hours. The waineo type of machines take days to grind a lens and you have to constantly watch them. I will try and find a picture of a commercial machine that I am trying to mimic on a DIY level.
Hezz
I think the parts in any hard drive would be too small. We have high loads and forces compared to a hard drive so everything must be enlarged and created to handle the kinds of forces that occur during machining.
The interesting thing about air bearings is that they are actually more simple that regular mechanical bearing but the system needs higher maching accuracy.
Ace,
I want to build something that is faster and more precise than those waineo type grinders. You have to be constantly adjusting things and rebuilding the lens grinding tool. If you don't have the thing set up exactly right it will not make the right curvature that you wanted. I want to make a single point grinding machine that uses a diamond grinding wheel or any other type of wheel that you might want to use.
I want to be able to grind a lens in two to four hours. The waineo type of machines take days to grind a lens and you have to constantly watch them. I will try and find a picture of a commercial machine that I am trying to mimic on a DIY level.
Hezz
Just to let you guys know I'm very interested in this project, as a studying Mechanical Engineer, and a DIY projector - it's right in my area of interest. Just thought I'd mention it's a really interesting idea, though I'm not really sure what your design is.
I get the spinning part 😛 And I get those other designs - but Hezz, what do you propose to do with the tooling end of it? CNC tool on horizontal/vertical axes?
This sounds way cool and I might even try building one at some point depending on your progress.
I get the spinning part 😛 And I get those other designs - but Hezz, what do you propose to do with the tooling end of it? CNC tool on horizontal/vertical axes?
This sounds way cool and I might even try building one at some point depending on your progress.
Mr. Kh,
Ok, here is a picture of what I have designed so far and here is the concept. Try to replicate an inexpensive to manufacture table top CNC lens grinding machine for micro machining with high accuracy and very low vibration. It must be relatively light and ridgid. It will use a custom made spindle with an R-8 taper so that standard R8 collets can be used to mount any type of tool. It will be capable of using aluminum oxide or silicon carbide wheels or diamond single point grinding wheels or even a standard mill so you can make a lens blank out of metal for a molding process.
I know air bearings don't allow for as high of loads as liquid pressure bearings but since this is a grinder and not intended to take large amounts of material off at once with high feed rates I think we can get by by designing the air bearing surfaces with high surface areas. Also in keeping with the DIY philosophy it must be relatively inexpensive to run and air seemed to be the easiest way. Clean. Don't need an auxillary pump and reservoir, etc.
Ok, here is the model so far without the spindle motor attached.
Hezz
Ok, here is a picture of what I have designed so far and here is the concept. Try to replicate an inexpensive to manufacture table top CNC lens grinding machine for micro machining with high accuracy and very low vibration. It must be relatively light and ridgid. It will use a custom made spindle with an R-8 taper so that standard R8 collets can be used to mount any type of tool. It will be capable of using aluminum oxide or silicon carbide wheels or diamond single point grinding wheels or even a standard mill so you can make a lens blank out of metal for a molding process.
I know air bearings don't allow for as high of loads as liquid pressure bearings but since this is a grinder and not intended to take large amounts of material off at once with high feed rates I think we can get by by designing the air bearing surfaces with high surface areas. Also in keeping with the DIY philosophy it must be relatively inexpensive to run and air seemed to be the easiest way. Clean. Don't need an auxillary pump and reservoir, etc.
Ok, here is the model so far without the spindle motor attached.
Hezz
Attachments
Well I can see you guys are really getting serious with project. A cnc machine is not easy to build. I build one last year but it was a really simple. I used 4 drawer sliders as my tracks, two threaded rods, somes nuts, and a Dremel as my cutter. Instead of using servos or stepper motors I used to geared motors because they cheaper and easy to hookup to a simple circuit. My machine was not connected to a computer because I didnt no anything about the interface between the two machines. So I just made two switches that controlled each axis. It was like drawing with a etcha-sketch, couldnt make a simple circle only squares and rectangles.
Hezz, you may want to check out these sites
www.cnczone.com
www.crankorgan.com/index.html
http://members.lycos.nl/majopi/index.html
www.luberth.com/cstep/
Hezz, you may want to check out these sites
www.cnczone.com
www.crankorgan.com/index.html
http://members.lycos.nl/majopi/index.html
www.luberth.com/cstep/
Rmccoll,
Thanks for the links. Right now I have compiled a rather large collection of links but what I need is a textbook on linear motor design.
Hezz
Thanks for the links. Right now I have compiled a rather large collection of links but what I need is a textbook on linear motor design.
Hezz
Try the liburary at the local university, motors arent hard to make bud, u just got to learn the winding and power formulas in your case, they should be on the net. I wouldnt go making a motor either, its cheap enough and easier to just modd a mains motor speed via its windings or a speed controller, reduction box ect.
Trev
Trev
Hezz,
Don't take any of this as negative, as I am intending it to be constructive. The light bulb wasn't invented in a day. . . and I laud your efforts for pursuing a goal that is difficult to achieve.
In my experience (paper machines and metal forming) the mechanical challenge is less than a successful control challenge. I suggest finding a zealous controls guy to help with your ultimate goal.
Also, especially in paper, the natural frequency of things like frames etc made a huge design difference. My company took all the records for many years (first to over 6000 feet per minute, largest daily tonnage etc) and focused on the natural frequency of the resultant structure. I say this as I imagine optics are even more critical. If NASA botched the Hubbell telescope, there are errors to be made. I feel that even in designing a 'perfect' drive system, you should seriously consider the effects of a frame system. Each and every beam (tube) in your frame will have a natural frequency and even a slight excitation could indeed be the nail in the coffin for the project.
I suggest you take advantage of the physics available to you, and chose to over design components which have a similar harmonic (1st thru 3rd at least) of resonance of each beam. These items may cost a bit more, but finding a harmonic frequency induced error will cost mega-bucks more. I have never calculated harmonics for a grinding wheel vs. a ground piece, but have run equipment which developed a harmonic vibration when turning in contact with each other (valve grinder vs. grinding wheel made an audible change in harmonic).
Good, no, great luck in tackling this beast. I'm sure there are other challenges to running a 24/7 machine of this sort, but for a tinkering DIY effort, hopefully you can go for it. I still think control is the biggest challenge. . .
The last disparaging remark and I'll leave you to your success. My senior project desire dealt with real-time flow bench results for analyzing pipe flows. The cheapest I could do it (consulting various vendors) was 68k USD and obviously my faculty advisor laughed when I approached them with that price tag. I ended up doing a data acquisition system for a racecar, which was fun, but remedial. Years later, I regret not reaching for the brass ring. Today, I imagine I could do the system with favors and 10k USD, but the mortgage just doesn't all quite that for DIY. . .
Good luck and try to think more about the things which shouldn't go wrong than those that should. That is the life of a machine designer, as far as my years can tell. . .
Sandy.
Don't take any of this as negative, as I am intending it to be constructive. The light bulb wasn't invented in a day. . . and I laud your efforts for pursuing a goal that is difficult to achieve.
In my experience (paper machines and metal forming) the mechanical challenge is less than a successful control challenge. I suggest finding a zealous controls guy to help with your ultimate goal.
Also, especially in paper, the natural frequency of things like frames etc made a huge design difference. My company took all the records for many years (first to over 6000 feet per minute, largest daily tonnage etc) and focused on the natural frequency of the resultant structure. I say this as I imagine optics are even more critical. If NASA botched the Hubbell telescope, there are errors to be made. I feel that even in designing a 'perfect' drive system, you should seriously consider the effects of a frame system. Each and every beam (tube) in your frame will have a natural frequency and even a slight excitation could indeed be the nail in the coffin for the project.
I suggest you take advantage of the physics available to you, and chose to over design components which have a similar harmonic (1st thru 3rd at least) of resonance of each beam. These items may cost a bit more, but finding a harmonic frequency induced error will cost mega-bucks more. I have never calculated harmonics for a grinding wheel vs. a ground piece, but have run equipment which developed a harmonic vibration when turning in contact with each other (valve grinder vs. grinding wheel made an audible change in harmonic).
Good, no, great luck in tackling this beast. I'm sure there are other challenges to running a 24/7 machine of this sort, but for a tinkering DIY effort, hopefully you can go for it. I still think control is the biggest challenge. . .
The last disparaging remark and I'll leave you to your success. My senior project desire dealt with real-time flow bench results for analyzing pipe flows. The cheapest I could do it (consulting various vendors) was 68k USD and obviously my faculty advisor laughed when I approached them with that price tag. I ended up doing a data acquisition system for a racecar, which was fun, but remedial. Years later, I regret not reaching for the brass ring. Today, I imagine I could do the system with favors and 10k USD, but the mortgage just doesn't all quite that for DIY. . .
Good luck and try to think more about the things which shouldn't go wrong than those that should. That is the life of a machine designer, as far as my years can tell. . .
Sandy.
Sandy H,
Your observations are all salient. Vibrations are particularly a serious problem in machine tools unfortunately I don't yet have the engineering expertise to calculate the frame structures frequencies. The best I will be able to do is over build the frame and try to make it as assymetrical as possible so that frequencies that are emphasized will not be compounded too much. The higher I can drive up the Q of the system the easier it will be to damp out some of the vibrations and the more accurate the machine will be. For instance using extensive over bracing and filling the tube members with lead shot and mineral oil I can damp out a lot of the higher frequency vibrations. Overall. I expect that this machine is not intended to compete with thoroughly engineered designs but my goal is just being able to grind a glass lens that be be finished by hand lapping if necessary.
I also recognize that the sucess of this machines will relate to the drive electronics. Fortunately these problems have mostly been solved by others as I am doing nothing new. The beauty of permanent magnet linear motors is their simplicity but of course all of the intelligence needs to be in the control system.
I look at this as a long term project. Not one that I will be doing in a weekend.
Hezz
Your observations are all salient. Vibrations are particularly a serious problem in machine tools unfortunately I don't yet have the engineering expertise to calculate the frame structures frequencies. The best I will be able to do is over build the frame and try to make it as assymetrical as possible so that frequencies that are emphasized will not be compounded too much. The higher I can drive up the Q of the system the easier it will be to damp out some of the vibrations and the more accurate the machine will be. For instance using extensive over bracing and filling the tube members with lead shot and mineral oil I can damp out a lot of the higher frequency vibrations. Overall. I expect that this machine is not intended to compete with thoroughly engineered designs but my goal is just being able to grind a glass lens that be be finished by hand lapping if necessary.
I also recognize that the sucess of this machines will relate to the drive electronics. Fortunately these problems have mostly been solved by others as I am doing nothing new. The beauty of permanent magnet linear motors is their simplicity but of course all of the intelligence needs to be in the control system.
I look at this as a long term project. Not one that I will be doing in a weekend.
Hezz
Sandy has some points about controls. I do controls consulting and diagnostics for robotics, etc. Some suggestions based on my experience:
- Don't count on asymetical structures helping out too much. If you have an object that is mounted on two sides, you will get a first resonant frequency (among others) off of each mounting. If your structure is symetrical, you will get a single resonance that is 6 dB higher than either by itself. Sounds bad, but the alternative is two resonances. Now you have a problem over a wider frequency range. Pick your poison.
- Home made linear motors sound pretty sketchy if you aren't already comfortable in controls. A ball screw (or even a decent lead screw) based system is much easier to do the controls on. Remember that with linear motors, your control system is the only thing providing stiffness for your system. You better be confident in it. It is very possible, but I wouldn't recommend it for a first project.
- You may consider that a machine that is stiff and relies on its controls system to acurately place a grinder may not be ideal. Think about how people figure mirrors in the diy telescope world (by hand). People, are not accurate, stiff, nor repeatable, but they can make great mirrors. Consider making a relatively stiff machine and having a comparativley flexible mount for your grinder. If your grinder is mounded stiffly, all your machine imperfections will be translated to the lens.
- Don't count on asymetical structures helping out too much. If you have an object that is mounted on two sides, you will get a first resonant frequency (among others) off of each mounting. If your structure is symetrical, you will get a single resonance that is 6 dB higher than either by itself. Sounds bad, but the alternative is two resonances. Now you have a problem over a wider frequency range. Pick your poison.
- Home made linear motors sound pretty sketchy if you aren't already comfortable in controls. A ball screw (or even a decent lead screw) based system is much easier to do the controls on. Remember that with linear motors, your control system is the only thing providing stiffness for your system. You better be confident in it. It is very possible, but I wouldn't recommend it for a first project.
- You may consider that a machine that is stiff and relies on its controls system to acurately place a grinder may not be ideal. Think about how people figure mirrors in the diy telescope world (by hand). People, are not accurate, stiff, nor repeatable, but they can make great mirrors. Consider making a relatively stiff machine and having a comparativley flexible mount for your grinder. If your grinder is mounded stiffly, all your machine imperfections will be translated to the lens.
- Status
- Not open for further replies.