Re: [Emc-users] DRO Input?
> Message: 9 > Date: Thu, 31 Jan 2008 13:36:51 -0800 > From: Dave Engvall <[EMAIL PROTECTED]> > Subject: Re: [Emc-users] DRO input? > > Hi Dave, > > It all depends on what you want to do. If you just want manual > position then the glass scales will do your job. > > However, if you ever intend to control the axes then encoders mounted > on the end of the ball screws would be my choice. USdigital or > Automation Direct (Koyo) encoders are reasonably priced and should do > the job. With respect to you all, I know about encoders, and I know any CNC control software is going to need positional input many, many times per second. but... sticking an encoder on an axis is not measuring the position of anything except the leadscrew, it takes no account whatsoever of backlash, uneven wear or pitch errors. To a certain extent you can map out backlash and areas of wear on a leadscrew, but leadscrew pitch errors are going to be tough. The DRO is, quite differently, measuring the *actual* X Y Z position, if the DRO says you are 100.000 mm from point A then you can take it to the bank. My leadscrews are ten turns to the inch, I can compensate for backlash etc manually, second nature, and even with less than theoretically perfect accuracy I can still trust feeding a leadscrew in a few thou and know and measure that the error on that movement is very small. We appear to be confusing two things. Encoding leadscrew rotational angles and feeding this back to the human brain or CNC software, and factoring in mental or electronic fudge tables for backlash and wear, can give us fairly good accuracy of MOVEMENT. A DRO with proper glass scales gives us fairly good accuracy of POSITION. 5 times a second is *plenty* for an accurate positional measurement system to update an accurate movement system, and produce a system with true accuracy. While leadscrew encoders and a copy of emc and no stepper or servo motors will indeed give me a system that will display X Y Z co-ordinates on a screen, the accuracy of these readings is going to be just as suspect as it is sans emc and encoders, only the DRO and pukka glass scales will give true positional accuracy. As someone who couldn't code "hello world" (except maybe in BASIC) writing the code isn't an option, so unless emc has this facility on the roadmap then being free as in beer isn't enough. Many thanks to all, hopefully there is some more meat in this subject yet. end - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO Input?
Dave. I would say all the advice given to you is sort of a shot in the dark- until you tell us your intentions with your equipment. Of course what you do with the advice is up to you. As far EMC and serial input from a DRO goes - there is no driver currently to do what you ask, but it could be written. One would need some information about the DRO first. Another idea is you could feed the scales output to EMC then Get EMC to output a signal to your DRO (assuming the scales output square wave signals) The question is why? Cheers Chris Morley > Date: Fri, 1 Feb 2008 10:07:32 + > From: [EMAIL PROTECTED] > To: emc-users@lists.sourceforge.net > Subject: Re: [Emc-users] DRO Input? > > >> Message: 9 >> Date: Thu, 31 Jan 2008 13:36:51 -0800 >> From: Dave Engvall >> Subject: Re: [Emc-users] DRO input? > >> >> Hi Dave, >> >> It all depends on what you want to do. If you just want manual >> position then the glass scales will do your job. >> >> However, if you ever intend to control the axes then encoders mounted >> on the end of the ball screws would be my choice. USdigital or >> Automation Direct (Koyo) encoders are reasonably priced and should do >> the job. > > > With respect to you all, I know about encoders, and I know any CNC control > software is going to need positional input many, many times per second. > > but... sticking an encoder on an axis is not measuring the position of > anything except the leadscrew, it takes no account whatsoever of backlash, > uneven wear or pitch errors. > > To a certain extent you can map out backlash and areas of wear on a > leadscrew, but leadscrew pitch errors are going to be tough. > > The DRO is, quite differently, measuring the *actual* X Y Z position, if > the DRO says you are 100.000 mm from point A then you can take it to the > bank. > > My leadscrews are ten turns to the inch, I can compensate for backlash etc > manually, second nature, and even with less than theoretically perfect > accuracy I can still trust feeding a leadscrew in a few thou and know and > measure that the error on that movement is very small. > > We appear to be confusing two things. > > Encoding leadscrew rotational angles and feeding this back to the human > brain or CNC software, and factoring in mental or electronic fudge tables > for backlash and wear, can give us fairly good accuracy of MOVEMENT. > > A DRO with proper glass scales gives us fairly good accuracy of POSITION. > > 5 times a second is *plenty* for an accurate positional measurement system > to update an accurate movement system, and produce a system with true > accuracy. > > While leadscrew encoders and a copy of emc and no stepper or servo motors > will indeed give me a system that will display X Y Z co-ordinates on a > screen, the accuracy of these readings is going to be just as suspect as > it is sans emc and encoders, only the DRO and pukka glass scales will give > true positional accuracy. > > As someone who couldn't code "hello world" (except maybe in BASIC) writing > the code isn't an option, so unless emc has this facility on the roadmap > then being free as in beer isn't enough. > > Many thanks to all, hopefully there is some more meat in this subject yet. > > end > > > > - > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ > ___ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users _ - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO Input?
Would it be difficult to update the actual position with the position from the linear encoders? The DRO position would just stomp on the actual position. Hopefully most of the time the error would be small and it would require real time DRO position... or perhaps only assign the DRO value when the axis is moving very slow or stopped. - Original Message - From: <[EMAIL PROTECTED]> To: Sent: Friday, February 01, 2008 4:07 AM Subject: Re: [Emc-users] DRO Input? > >> Message: 9 >> Date: Thu, 31 Jan 2008 13:36:51 -0800 >> From: Dave Engvall <[EMAIL PROTECTED]> >> Subject: Re: [Emc-users] DRO input? > >> >> Hi Dave, >> >> It all depends on what you want to do. If you just want manual >> position then the glass scales will do your job. >> >> However, if you ever intend to control the axes then encoders mounted >> on the end of the ball screws would be my choice. USdigital or >> Automation Direct (Koyo) encoders are reasonably priced and should do >> the job. > > > With respect to you all, I know about encoders, and I know any CNC > control > software is going to need positional input many, many times per > second. > > but... sticking an encoder on an axis is not measuring the position of > anything except the leadscrew, it takes no account whatsoever of > backlash, > uneven wear or pitch errors. > > To a certain extent you can map out backlash and areas of wear on a > leadscrew, but leadscrew pitch errors are going to be tough. > > The DRO is, quite differently, measuring the *actual* X Y Z position, > if > the DRO says you are 100.000 mm from point A then you can take it to > the > bank. > > My leadscrews are ten turns to the inch, I can compensate for backlash > etc > manually, second nature, and even with less than theoretically perfect > accuracy I can still trust feeding a leadscrew in a few thou and know > and > measure that the error on that movement is very small. > > We appear to be confusing two things. > > Encoding leadscrew rotational angles and feeding this back to the > human > brain or CNC software, and factoring in mental or electronic fudge > tables > for backlash and wear, can give us fairly good accuracy of MOVEMENT. > > A DRO with proper glass scales gives us fairly good accuracy of > POSITION. > > 5 times a second is *plenty* for an accurate positional measurement > system > to update an accurate movement system, and produce a system with true > accuracy. > > While leadscrew encoders and a copy of emc and no stepper or servo > motors > will indeed give me a system that will display X Y Z co-ordinates on a > screen, the accuracy of these readings is going to be just as suspect > as > it is sans emc and encoders, only the DRO and pukka glass scales will > give > true positional accuracy. > > As someone who couldn't code "hello world" (except maybe in BASIC) > writing > the code isn't an option, so unless emc has this facility on the > roadmap > then being free as in beer isn't enough. > > Many thanks to all, hopefully there is some more meat in this subject > yet. > > end > > > > - > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ > ___ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Emc-users DRO input?
That is how my DRO's work that I built for my mill and lathe. http://www.shumatech.com/ John On 31 Jan 2008 at 18:04, [EMAIL PROTECTED] wrote: > Another thought, is there any value in having an RS-232 connection > with serial calipers or micrometers? - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO input?
That would depend on the accuracy of your scales... .4" precision (.001mm)is not cheap... In my case mine are not nearly that accurate... I guess it depends on how much accuracy is needed and the future of the machine. EMC does have screw mapping... John On 1 Feb 2008 at 4:04, [EMAIL PROTECTED] wrote: > The DRO is, quite differently, measuring the *actual* X Y Z position, > if the DRO says you are 100.000 mm from point A then you can take it > to the bank. - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Emc-users DRO input?
On Fri, 2008-02-01 at 06:03 -0600, John Thornton wrote: > That is how my DRO's work that I built for my mill and lathe. > > http://www.shumatech.com/ > > John > On 31 Jan 2008 at 18:04, [EMAIL PROTECTED] wrote: > > > Another thought, is there any value in having an RS-232 connection > > with serial calipers or micrometers? That's a nice setup DRO system and website you have. I was fishing for a reason to have a serial micrometer connected to EMC for a reason other than as motion feed back or axis position. For instance, to automate offset adjustments, or something like Will Ferrell said in Old School, "Maybe she's wearing something I haven't even thought of yet." -- Kirk Wallace (California, USA http://www.wallacecompany.com/machine_shop/ Hardinge HNC lathe, Bridgeport mill conversion, doing XY now, Zubal lathe conversion pending) - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO input?
Dave Engvall wrote: > > Yep! Wilson machine bought a 50' x 4' CMM at a Boeing auction. Got it > cheap and spent a million moving it and > getting it set up again. And that was only a few miles. Holy COW! a 50 FOOT CMM? There must only be 2 or 3 in the entire US that big. (I can understand why Boeing would need such a machine.) Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO Input?
[EMAIL PROTECTED] wrote: > but... sticking an encoder on an axis is not measuring the position of > anything except the leadscrew, it takes no account whatsoever of backlash, > uneven wear or pitch errors. > > To a certain extent you can map out backlash and areas of wear on a > leadscrew, but leadscrew pitch errors are going to be tough. > You can "map" backlash, but you can't FIX it in software. Backlash means the table position is NOT fully constrained. EMC2 already has the capability to fix pitch irregularity, you just need to do the tedious measurement with gage blocks once and enter the data. > 5 times a second is *plenty* for an accurate positional measurement system > to update an accurate movement system, and produce a system with true > accuracy. > Not when the machine is moving at 60 IPM, doing a climb-mill cut. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO Input?
RogerN wrote: > Would it be difficult to update the actual position with the position > from the linear encoders? The DRO position would just stomp on the > actual position. Hopefully most of the time the error would be small > and it would require real time DRO position... or perhaps only assign > the DRO value when the axis is moving very slow or stopped. How about when moving at 60 IPM? What is the time delay in the DRO software? I've seen a number of good DROs that have SIGNIFICANT delays in the update, and settle once you slow down. If the serial output from the DRO box has a 1/4 second delay, and the servo is running at 1000 updates a second, you can only re-sync when you KNOW the machine is dead still. A true servo is NEVER dead-still, it is always bouncing between adjacent encoder counts. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO Input?
RogerN wrote: > Would it be difficult to update the actual position with the position > from the linear encoders? The DRO position would just stomp on the > actual position. Hopefully most of the time the error would be small > and it would require real time DRO position... or perhaps only assign > the DRO value when the axis is moving very slow or stopped. How about when moving at 60 IPM? What is the time delay in the DRO software? I've seen a number of good DROs that have SIGNIFICANT delays in the update, and settle once you slow down. If the serial output from the DRO box has a 1/4 second delay, and the servo is running at 1000 updates a second, you can only re-sync when you KNOW the machine is dead still. A true servo is NEVER dead-still, it is always bouncing between adjacent encoder counts. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO input?
>> >> Yep! Wilson machine bought a 50' x 4' CMM at a Boeing auction. Got it >> cheap and spent a million moving it and >> getting it set up again. And that was only a few miles. >Holy COW! a 50 FOOT CMM? There must only be 2 or 3 in the >entire US that big. (I can understand why Boeing would need >such a machine.) 50 foot at .0001 / Inch resolution would almost be at the limit for a 32 bit unsigned integer! I could see some really strange software bugs begin to show up. (Signed vs unsigned integers). Would a double precision float handle that? Yikes! 0xD693A400 - or something like that. Jim Combs - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO Input?
I have a different idea. Why not tap in to the signal from the encoders themselves? If this is glass scales with proper quadrature output it should be simple to read this signals by a cheap card? There is nothing that says the quadrature signals can't be read by both EMC and the DRO main unit. I plan to do this myself. I also bought a far eastern DRO with glass scales and hope to be able to tap in to the signals being transmitted by the scales. I have a Universal Stepper Card with quadrature inputs. At the same time it is quite nice to have the DRO for manual milling. Does anyone have a good reason for not doing this? I should possibly use optoisolators though, right? Anders > -Original Message- > From: [EMAIL PROTECTED] [mailto:emc-users- > [EMAIL PROTECTED] On Behalf Of Chris Morley > Sent: 1. februar 2008 12:02 > To: Enhanced Machine Controller (EMC) > Subject: Re: [Emc-users] DRO Input? > > > Dave. > > I would say all the advice given to you is sort of a shot in the dark- > until you tell us your intentions with your equipment. Of course what you > do with the advice is up to you. > As far EMC and serial input from a DRO goes - there is no driver > currently to do what you ask, but it could be written. One would need some > information about the DRO first. > Another idea is you could feed the scales output to EMC then Get EMC to > output a signal to your DRO (assuming the scales output square wave > signals) > The question is why? > > Cheers > Chris Morley > > > Date: Fri, 1 Feb 2008 10:07:32 + > > From: [EMAIL PROTECTED] > > To: emc-users@lists.sourceforge.net > > Subject: Re: [Emc-users] DRO Input? > > > > > >> Message: 9 > >> Date: Thu, 31 Jan 2008 13:36:51 -0800 > >> From: Dave Engvall > >> Subject: Re: [Emc-users] DRO input? > > > >> > >> Hi Dave, > >> > >> It all depends on what you want to do. If you just want manual > >> position then the glass scales will do your job. > >> > >> However, if you ever intend to control the axes then encoders mounted > >> on the end of the ball screws would be my choice. USdigital or > >> Automation Direct (Koyo) encoders are reasonably priced and should do > >> the job. > > > > > > With respect to you all, I know about encoders, and I know any CNC > control > > software is going to need positional input many, many times per second. > > > > but... sticking an encoder on an axis is not measuring the position of > > anything except the leadscrew, it takes no account whatsoever of > backlash, > > uneven wear or pitch errors. > > > > To a certain extent you can map out backlash and areas of wear on a > > leadscrew, but leadscrew pitch errors are going to be tough. > > > > The DRO is, quite differently, measuring the *actual* X Y Z position, if > > the DRO says you are 100.000 mm from point A then you can take it to the > > bank. > > > > My leadscrews are ten turns to the inch, I can compensate for backlash > etc > > manually, second nature, and even with less than theoretically perfect > > accuracy I can still trust feeding a leadscrew in a few thou and know > and > > measure that the error on that movement is very small. > > > > We appear to be confusing two things. > > > > Encoding leadscrew rotational angles and feeding this back to the human > > brain or CNC software, and factoring in mental or electronic fudge > tables > > for backlash and wear, can give us fairly good accuracy of MOVEMENT. > > > > A DRO with proper glass scales gives us fairly good accuracy of > POSITION. > > > > 5 times a second is *plenty* for an accurate positional measurement > system > > to update an accurate movement system, and produce a system with true > > accuracy. > > > > While leadscrew encoders and a copy of emc and no stepper or servo > motors > > will indeed give me a system that will display X Y Z co-ordinates on a > > screen, the accuracy of these readings is going to be just as suspect as > > it is sans emc and encoders, only the DRO and pukka glass scales will > give > > true positional accuracy. > > > > As someone who couldn't code "hello world" (except maybe in BASIC) > writing > > the code isn't an option, so unless emc has this facility on the roadmap > > then being free as in beer isn't enough. > > > > Many thanks to all, hopefully there is some more meat in this subject > yet. > > > > end > > > > > > > > > - > > This SF.net email is sponsored by: Microsoft > > Defy all challenges. Microsoft(R) Visual Studio 2008. > > http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ > > ___ > > Emc-users mailing list > > Emc-users@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/emc-users > > _ > > > - > This SF.net email is sponsored by: Mi
Re: [Emc-users] DRO Input?
On Fri, 2008-02-01 at 23:19 +0100, Anders Blix wrote: > I have a different idea. > > Why not tap in to the signal from the encoders themselves? If this is glass > scales with proper quadrature output it should be simple to read this > signals by a cheap card? There is nothing that says the quadrature signals > can't be read by both EMC and the DRO main unit. > > I plan to do this myself. I also bought a far eastern DRO with glass scales > and hope to be able to tap in to the signals being transmitted by the > scales. I have a Universal Stepper Card with quadrature inputs. > At the same time it is quite nice to have the DRO for manual milling. > > Does anyone have a good reason for not doing this? I should possibly use > optoisolators though, right? > > Anders I don't see why you need the DRO unit. EMC makes for a good DRO as well as a motion controller and desert topping. See here: http://www.wallacecompany.com/machine_shop/EMC2/dro_vfd/ You can change between EMC and EMC/DRO by running the appropriate file. -- Kirk Wallace (California, USA http://www.wallacecompany.com/machine_shop/ Hardinge HNC lathe, Bridgeport mill conversion, doing XY now, Zubal lathe conversion pending) - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
[Emc-users] Serial DAC Done (Almost)
I have just successfully tested my serial DAC with my VFD's and KBIC. It's kind of strange to see it work with the KBIC, since it is floating at about 70 Volts AC, but it works fine. See here: http://www.wallacecompany.com/machine_shop/EMC2/serial_dac/ If anyone is interested, I can make a parts list, layout a real board and have boards etched and silk-screened. At least three parallel port pins are needed. The Clock and Data pins can run to multiple serial devices. A Load pin is needed for each device. I did a breadboard of the serial ADC, which is working, and an H-bridge to be used as a DC motor reverser on the KBIC, and so far it's also working well. -- Kirk Wallace (California, USA http://www.wallacecompany.com/machine_shop/ Hardinge HNC lathe, Bridgeport mill conversion, doing XY now, Zubal lathe conversion pending) - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO input?
[EMAIL PROTECTED] wrote: >>>Yep! Wilson machine bought a 50' x 4' CMM at a Boeing auction. Got it >>>cheap and spent a million moving it and >>>getting it set up again. And that was only a few miles. >> >>Holy COW! a 50 FOOT CMM? There must only be 2 or 3 in the >>entire US that big. (I can understand why Boeing would need >>such a machine.) > > > > 50 foot at .0001 / Inch resolution would almost be at the limit for a 32 > bit unsigned integer! I could see some > really strange software bugs begin to show up. (Signed vs unsigned > integers). Would a double precision > float handle that? That's a good point. I vaguely seem to recall the OLD EMC(1) overflowed the 32-bit raw encoder count into a double float. The current version just goes to a 32-bit signed value (unless this has been changed recently). I was wondering if that would ever be a problem. Hmm, calculating it, I get 50 * 12 * 1 = 6 million, which is not anywhere near 4 billion, or even +/- 2 billion, which is what the numerical limits of a 32-bit integer are. I calculate that 2^31 / ( 1 * 12) = 17896.xxx FEET, which sounds like it is a lot less of a problem. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] DRO input?
On Feb 1, 2008, at 8:12 PM, Jon Elson wrote: [EMAIL PROTECTED] wrote: Yep! Wilson machine bought a 50' x 4' CMM at a Boeing auction. Got it cheap and spent a million moving it and getting it set up again. And that was only a few miles. Holy COW! a 50 FOOT CMM? There must only be 2 or 3 in the entire US that big. (I can understand why Boeing would need such a machine.) 50 foot at .0001 / Inch resolution would almost be at the limit for a 32 bit unsigned integer! I could see some really strange software bugs begin to show up. (Signed vs unsigned integers). Would a double precision float handle that? That's a good point. I vaguely seem to recall the OLD EMC(1) overflowed the 32-bit raw encoder count into a double float. The current version just goes to a 32-bit signed value (unless this has been changed recently). I was wondering if that would ever be a problem. Hmm, calculating it, I get 50 * 12 * 1 = 6 million, which is not anywhere near 4 billion, or even +/- 2 billion, which is what the numerical limits of a 32-bit integer are. I calculate that 2^31 / ( 1 * 12) = 17896.xxx FEET, which sounds like it is a lot less of a problem. Putting a face on it with a bit of arithmetic always helps. Dave Jon -- --- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
