On 8/24/2016 1:02 PM, Gene Heskett wrote: > On Wednesday 24 August 2016 12:39:39 Jon Elson wrote: > >> On 08/23/2016 10:22 PM, Gene Heskett wrote: >>> On Tuesday 23 August 2016 22:13:55 Jon Elson wrote: >>>> On 08/23/2016 04:09 PM, Gene Heskett wrote: >>>>> On Tuesday 23 August 2016 06:55:03 andy pugh wrote: >>>>>> For gear tooth detection you probably want >>>>>> http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2050601.m570.l1 >>>>>> 31 3.T R0.TRC0.H0.XATS667.TRS0&_nkw=ATS667&_sacat=0 >>>>> 10 ordered, I'll figure out something to use the others for. 3 for >>>>> gear sensors and 6 for home & limits, leaves a spare. I'll need to >>>>> cobble up an index somehow. >>>> The ATS667 is a sensor to detect the passage of a gear >>>> tooth. It really is not designed as a proximity sensor. It >>>> will detect a ferrous item passing across the face of the >>>> sensor, but then stays in that state. If you check the >>>> datasheet, it has two sensors close together, compares the >>>> signal from them and sets a flip-flop for which one has more >>>> signal. So, it is designed to sense an object moving ACROSS >>>> the face of the sensor. >>>> >>>> Jon >>> So if we want to use it as home or limit, we should trip on the far >>> side of what we have passed? >>> >>> What occurs when it has come back from there? Or is at a great >>> distance from the trip generator? Or am I back to 59 cent roller >>> microswitches? I must admit, they do seem to be doing a quite >>> repeatable job where I've used them in essentially dry circuits. >> OK, so when a ferrous object passes left-to-right across the >> sensor, the output goes high, and then holds. >> When the object passes right-to-left, then the output goes >> low, and holds. > That would work for either limits or homeing I think. I have my > microswitches set so whatever trips them can pass by, but it quite a > ways before the lever would open the switch again. > >> The FF holds the state until the object >> comes back within range. >> >> There most certainly are Hall sensors that detect the >> presence of an object, without the latching or directional >> features. I think that is really what you want. Also, >> sensors that have big magnets in them are not ideal in metal >> cutting machines. They are fine inside the head of my mill >> where chips shouldn't be able to get. > There is that too. I have enough trouble with residual magnetism in my > hand tools as is, those that can approach a motor in particular. The > motor on my toy mill? I could remove it from the mill and stick it to a > tin plate decorative ceiling and not worry about its falling unless the > Richter scale is above 6.5. And tool that approaches the nearby drawbar > bolt is instantly grabbed. PITA. >> Jon > I found the parts list for the Sheldon and that gear I want to drive the > spindle encoder with has 60 teeth. Thats 240 edges, similar to my > optical's I've made before. But while I can envision a 1/2" wide alu bar > cut in the curve of the tooth OD, with pockets for the 667's, I've not > decided where I drill holes to mount the alu piece. Its a tight fit. > > Cheers, Gene Heskett
Gene, My experience with standard prox sensors is that they are plenty accurate enough for homing and limits especially when you are using a the two approach method which is a rapid approach and then a second slowdown approach. By doing that I can get single digit thousandths accuracy for homing which is fine for most automation jobs. Combine that with Index homing and you are all set for CNC machine work. Prox sensors are great in that they don't have a magnet in them so they don't attract chips and most are oil and coolant proof. There was a company selling oil proof proxes for less than $10 each in the US. Here they are: http://www.ezautomation.net/ezvizprox/ezvizprox_products.php?cat=290 I have never used their sensors so YMMV. There is a big difference between a shielded prox and an unshielded prox. In most cases you will want to use a shielded prox. (the plastic doesn't stick out the end of it.) Automation Direct is a good source of prox sensors. They also have a generous return policy. Dave ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
