OK the dots really ARE multiple beams. Perhaps you have a very poor quality laser and the lens has facets like a cut diamond? If it really is "speckle" then it will change when you change anything and look like your other laser.
But even if the beam has 40 spots, it will still work just fine this purpose and maybe even better than a perfect laser would. A complex signal is helpful. As long as the dots don't have a pattern. What you care about is the displacement (shift) from one image to the next as you take images. So you shift one image until it lines up perfectly with the previous image and you remember how much shifting was required and figure that must be caused by the bent/worn/warped bed. With a complex image made of many random dots you not only can measure the X and Y offset but a rotational component too, if one exists. Yes, there just might be a rotation on the lathe bad. If you multiply two images together you get a huge spike when the alignment is perfect The Wiki describe the 1D case but 2D is just the same. https://en.wikipedia.org/wiki/Autocorrelation On Wed, Sep 26, 2018 at 9:31 PM Gene Heskett <ghesk...@shentel.net> wrote: > On Wednesday 26 September 2018 21:45:42 Chris Albertson wrote: > > > I think what you are seeing is a "speckle pattern" The laser is not > > making 40 small beams. it is making one nice uniform beam but > > remember the light is all in phase. After a reflection from a > > textured surface the light hitting the back your eye adds in phase and > > the spots you are are where the light adds in phase and the other > > places the wave self-cancels > > > > see the second paragraph on this page: > > https://en.m.wikipedia.org/wiki/Speckle_pattern. > > > If I am interpreting that correctly, and with the spindle stopped, I > should see this "interference" pattern change rapidly as the carriage to > target distance changes by a single microstep of the screws driving > motor. I don't see that effect, the individual dots remain stationary. > > Not discused is the diffraction effect of passing this coherent bean thru > a pinhole other than a passing statement that this speckle effect will > cause the spots to be larger if viewed thru a pinhole. One naturally > assumes the pinhole is in front of the eye, not the lasers exit aperture > from a statement like that. > > Also, all the patterns shown are circular. The patterns thrown on the > wall by a laser pointer I've had for years, a laser level, and both of > these devices are bars 3 or 4x longer than wide. The dot patterns I am > seeing rotate with the device, remaining stationary when the reference > is the rotation. IOW go put a magic marker dot on the paper for each dot > on a distant target many feet away, then turn the device on its axis by > some arbitrary number of degrees. The dots will rotate, and the dots can > be re-registered by rotating the target by the same number of degrees. > > One side of most alu wrapping foil has a frosted finish, so internal > reflection of that that light that does not pass thru the pinhole on its > first pass but bounces around should result in a larger more diffuse > pattern, The other side, being more highly polished should result in a > more easily seen circular diffraction, and it will be interesting to see > if in the real world, a difference can be seen. And the pinhole material > s/b thin. A 1/32" thick piece of brass will have far larger diffraction > than a .0005" piece of alu foil for a same sized hole. > > Interesting series of experiments ahead when the stuff gets here. And > someone mentioned using a lappy to view the results, but I'm not sure my > old hp with a 1.4ghz turion cpu is enough "iron" to get the job done. I > have 640x480 cameras that can focus down to 2 or 3" in order to focus on > the target, so it will be tried, and results reported in due time. > > > If the pattern of the dots stay the same as you move the laser around > > then the "speckling" is happening on the lens and some parts of the > > beam are flying in a shorter or longer path length. and add the all > > add in phase. > > > > Even if the laser were making 40 spots and the beam as not on-axis, > > don't bother to fix it. You don't need to. A 1 depress misprinted > > beam will give only tiny error for your purpose. Down in the 2% error > > range. and I bet you can aim the beam by eyeball to 1/2 a degree. > > That is good enough. > > > > Even if there are 40 beam hitting your target, so what? Be happy not > > you can take 4 measurements at once and drastically reduce the signal > > to noise ratio. If you are using software you are like tough an > > autocorrelation between images and the is a peak when yu have found a > > match. having a complex parter makes the peak of the function more > > steam and you have better and more accurate results. 40 points is a > > Good Thing, as long as they don't change. > > > > Again, you beam is fine, it is just that most people forget the laser > > is 'coherent" light and all in phase. > > > > On Wed, Sep 26, 2018 at 4:24 PM Gene Heskett <ghesk...@shentel.net> > wrote: > > > On Wednesday 26 September 2018 17:22:01 Chris Albertson wrote: > > > > ---------- Forwarded message --------- > > > > From: Gene Heskett <ghesk...@shentel.net> > > > > Date: Wed, Sep 26, 2018 at 12:08 AM > > > > Subject: Re: [Emc-users] More on bed wear fix > > > > To: <emc-users@lists.sourceforge.net> > > > > > > > > > > > > But finding a camera chip big enough to shine these laser's on so > > > > that > > > > > > > > > the laser stays on the imaging chip cannot be found in an > > > > > integrated package that will plug into the 15 pin connector on a > > > > > pi. The laser beams are 10x the size of the imaging chips even > > > > > if they're perfectly aligned. > > > > > > > > What? If I buy a normal green laser pointer the green spot is > > > > about 1mm in dimmer. If I place a 0.01 pinhole in the path of > > > > the laser the spot is pretty close to 0.01. As for a camera that > > > > plugs into a Raspberry Pi, any USB web cam will do. But I'd use a > > > > notebook PC rather then the Pi. What you probably want is not > > > > to shine the beam on the sensor but have a glass or plastic target > > > > and use a $20 USB microscope to focus on the target. Glass is good > > > > because it will reflect 99% of the laser light away and not blind > > > > the camera. > > > > > > > > As you move the carriage the laser spot should move on the target > > > > and the USB microscope will image the back side of the target and > > > > you can measure the movement by counting pixels, Software can do > > > > better by doing some curve fitting. Even if the bed is very bad > > > > I doubt the laser spot move even 0.5mm > > > > > > > > If the laser spot looks huge, perhaps what you have is a small > > > > spot that overloads the sensor. > > > > > > I am looking at the spot with my eyes. I can see the individual > > > dots, probably 40 of them. If I could restrict it with a pinhole on > > > the laser to block most of the dots, that would be great, but even > > > with edm I don't think I could make a pinhole that small. > > > > > > > The target may need to do some serious > > > > amount of attenuation. > > > > > > I have a polarizing variable filter with a range from ND-2 to ND-400 > > > coming. > > > > > > > And while I said "USB "microscope" I bet a > > > > > > > > magnifier hot glued to a web cam works. But I did buy a $10 > > > > microscope and the image is poor but I can inspect the solder > > > > joints on SMT parts, A medium size SMT resister pretty much fills > > > > frame, these things are on eBay for $9 to $35 and they all appear > > > > to the the same. > > > > > > Humm, more experimentation seems in order. I have drills down to > > > #80, and yards of Reynolds wrap, which might get me down to a usable > > > spot size. > > > > > > > Pretty much everyone who wants to measure displacement from a line > > > > uses a laser. It is hard to find a better reference line > > > > > > > > If your bed where really bad Just put a steel rule in the tool > > > > holder and eye-ball the laser spot in the rule, maybe use a loupe > > > > to read it. > > > > > > > > Something is wrong if the laser spot is large. > > > > > > Its way too big at the source, and 2 of them (different brands) are > > > doing it. The pinhole at the src seems like a good starting point. > > > > > > [...] > > > > > > -- > > > Cheers, Gene Heskett > > > -- > > > "There are four boxes to be used in defense of liberty: > > > soap, ballot, jury, and ammo. Please use in that order." > > > -Ed Howdershelt (Author) > > > Genes Web page <http://geneslinuxbox.net:6309/gene> > > > > > > > > > _______________________________________________ > > > Emc-users mailing list > > > Emc-users@lists.sourceforge.net > > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > -- > Cheers, Gene Heskett > -- > "There are four boxes to be used in defense of liberty: > soap, ballot, jury, and ammo. Please use in that order." > -Ed Howdershelt (Author) > Genes Web page <http://geneslinuxbox.net:6309/gene> > > > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > -- Chris Albertson Redondo Beach, California _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
