Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Mountainman on April 25, 2013, 07:27:14 AM
-
I have an old Bridgeport V2XT that I retrofitted. I used almost all of the old mechanics in thsi case teh old Balluff limit switches. X and Y axis have a housing with a limit and homing switch in them. Each side of the axis has a dog that pushes down on teh plunger. One side of each axis has an extra dog for the limit switch. My problem is this, homing accuracy is at best .010". I am currently using the limit switches ( low on inputs). My homing procedure runs at 60IPM. The only reason I am reluctant to go with proximity switches is I am afraid that chips might throw everything off. I'm mostly cutting steel and aluminum. Anyways the existing switches are made in Germany, why are they not accurate?
I need to be able to rely on the repeatability of my home cause I run plenty of the same part over and over again. If every ten parts I take a break and let the machine home that would be wonderful. Say I lose .0005 with every part on every axis, ten parts later this could be a problem. I am very intersted in any suggestions anyone may have to improve my homing accuracy. My last option would be a proximity switch wired in series to the lower limit switch of my Z axis ( that is housed in a watertight quill housing) and two of those switches that a member here makes based on an optical system. This would help me minimize on inputs. But if anyone has a way to make what I have work, I am very interested.
-
Are you sure the home switches aren't repeating or is it skipping steps in the run somewhere creating a cumulative error?
I would slow the homing speed way down for sure.
Brett
-
My servo configurator program ( DG2S drives) and mach agree as to the number of steps that are being sent to the drive. So that doesn't concern me. MY NSK ballscrews have like .001 backlash in them.
What do you mean by the switch might be repeatiing?
I will definately be slowing down. I'm thinking about getting a card to utillize the 3rd port through my ESS and that will free up some inputs so I can run teh homing switches through my main BOB. I suspect the had additional homing switches because they are more accurate than the limit switches?
-
Can you not home in the drive using your index pulse?
Brett
-
No index pulse only A and B pulses
-
I been thinking, basically what you are talking about is inputting the index pulse into the break out board and rewriting the homing macro so that after the limit is hit, only when the index pulse is sensed will that position be deemed home. I'll have to splice something into the encoder cable. Any macro help out there?
-
You are not going to manage that successfully via a macro I am afraid. It will be far too slow. Ideally it needs done in the hardware and the SS/ESS does not support Index homing, in fact very few controllers available for Mach do support this.
On my machines with SmoothSteppers I do the index homing in my servo drives as they support this function. On my other machines I use the CSMIO controllers from CS-Lab and they support Index Homing.
Your best option would be enclosed optical switches, they are very accurate and being enclosed you dont have to worry about much or coolant. You cant buy them however they are relatively easy to make.
Hood
-
CS lab, I would have been more than happy to dish oput the extra dime for what they make. OK, I will more than likely be using the old homing switches. You think there is an advantage to using 5VDC for for homing over 24VDC? 5VDC more susceptible to VFD noise, 5vdc I believe needs better "closing" on the contacts to make a circuit.
-
24v has all the advantages rather than 5v.
Hood
-
Isn't 24V more likely to jump say a .005" gap than 5V?
-
You shouldnt have to worry about such a thing if you have decent snap action switches.
Hood
-
Hood, If I was to enclose 24 V proximity switches in my old balluf limit switch enclosures you think that would give me high accuracy? I also have to ask, will the angle of the dogs that trip the limit switches have any negative affect on accuracy? The old limit switch enclosures had these stainless looking plungers spring tensioned and sticking out of the enclosure. They seem to have been liquid tight and basically they would activate the switches when depressed by a "trip dog ( I think that is what they are called?)" One "trip dog" is mounted in the max and one in teh Min of an axis motion so only one switch is used. Is there a more appropriate angle on those trip dogs for homing and another for limit? I am trying to figure why the machine originally used 2 different switches for homing and limit? At this point I am thinking of chaining the all together.
-
I have never used prox's for homing switches so not sure what accuracy can be achieved.
They likely are used on some industrial machines however they will not be relying on the switches for the homing accuracy as they will just be using them as a trigger to start looking for the index pulse on the encoder.
That also is likely the reason your BP had separate homing switches as they would be used to signal the control to look for the index pulse. Likely it did not back off the switch before looking for the index so the home switch had to be separate so as not to trip a limit.
Hood
-
OK, so the angle of teh trip dog is not why they used one set of switches for homing and another for limit? So basically even if my swictes are top of the line, they were never made to achieve a precise location and never can achieve such a thing? Sure wish there was an auxiliary board that can combine limit switch inpuit and encoder index signal so mach can read a precise point for homing.
-
They may well be accurate, the honeywell switches I use are accurate enough although I always did index homing anyway. I could do that because my servo drives supported it but on the other machines that I have the CSMIO fitted then it does it so drives with homing capability are not needed.
Regarding an index homing board, there use to be such a thing, Ed Gilbert made various boards, one of which was just that. He doesnt make them any more as far as I know, in fact I dont think he makes anything now, his company was called CNC Building Blocks. I have a complete breakout board of his on the big lathe and it has the homing to index feature although I do not use that part of it now.
Hood