kerseykyle
Plastic
- Joined
- Jul 31, 2017
- Location
- Washington state
Recently I mounted a linear scale on the Z axis of a Myford MG12 cylindrical grinder for a DRO, and wanted to hear any feedback on the project or simpler approaches I had overlooked.
I could not find a good place to mount the scale on the outside of the table's casting, Also the table can be inadvertently lifted and needs to be periodically taken off for maintenance.
So I choose to mount the linear scale on the underside of the table and the encoder reader-head is hidden inside the machine.
What I used was a magnetic linear scale with an adhesive back from RLS (Distributed by Renishaw), this is the product I used https://www.rls.si/eng/hilin
Here is a photo of the machine after the project was completed.
coincidentally the table has a machined ledge for the rack gear, this surface extends the entire length of the table but the last 18.125 inches was empty providing a perfect place to stick the scale. It was parallel within 0.001 of the ways. The gray band is the magnetic scale. (They ship them in coiled a movie reel looking canister)
I created a CAD model to plan out the clearances under the table and where to mount the readhead bracket.
It was a tight fit. it clears the reduction gear for the cross feed screw by about a 1/16 inch. any further away and I would not have enough travel.
The bracket that holds the readerhead locates off of 3 balls 0.2500 diameter 440C. A single M6 button screw holds down the bracket.
I could not easily mill locating surfaces for the bracket, instead I drilled divots for the balls with a long spot drill using a drilling jig.
The features of this jig was modeled as a sub-assembly containing sketch geometry driven from the v-and-flat ways, also the mounting holes locations. I used a a piece of TG&P 4140 shaft for the jig and relived the center so it sits on 2 points.
I used a M6 button head screw for the 3rd point and secured the height with some loctite after getting a measurement from the CAD model I swept the top of the round bar with an indicator then raised the knee to that distance and swept the screw head and adjusted it until the indicator read zero.
Here the locating divots are being drilled. I discovered that by using two shaft collars as a depth stop I could incrementally adjust the depth using a gage pin.
I needed to be 0.025 deeper so I shifted the rear collar 0.125" using a gage pin to set the gap then clamped the front collar over a 0.100" pin, Worked very accurately
You can see how the drilling jig is constrained by the ways of the machine.
Using a depth micrometer I checked how deep the spheres seated and compared with the CAD model measurements.
Here are the kinematic locating features on the underside of the bracket made from 17-4 stainless.
Another reason for the kinematic mounting is to allow for it to be detached and mill the readerhead seating surfaces to their final size after checking on the machine.
I left 0.030" of extra material on those faces for final adjustment.
The drilling jig was repurposed into a sled for the indicator checking how much needed to be milled off to achieve parallelism with the ways.
A depth micrometer was used for the other shinny surface.
Here is a picture showing it assembled. On the bracket underside are M4 socket head cap screws for the readerhead.
I needed to solder a DB-9 connector onto the cord because the pinout was different, this was my first time assembling these type of connectors, I had trouble holding the small wires while soldering each pin.
RLS had very good datasheets showing the tolerances for the readahead alignment and detailing the RS-422 quadrature signal, also detailed CAD models for download.
The readerhead cost about $350 and $50 for the adhesive backed scale with a custom defined refence point and length, the only bad part is the lead time of about 2 months for the readerhead and 1 month for the scale.
The repeatability has been excellent, I chose the 0.1um resolution output. It has an accuracy of 10um per meter when using the adhesive backed scale, the DRO I have set to a resolution of 0.00001 (10-microinch).
This machine is in a an climate controlled shop and the flood coolant temperature is monitored during operation.
None of the parts I have done so far require a tolerance better than +- 0.0002, but that extra decimal of resolution helps sneek up on the final size when I take it off the grinder for measuring.
Having a Z axis readout makes it a lot better grinding the shoulders and faces of parts. I also have found plunge grinding works best, I step over about 1/3 the wheel width and punge to 0.0005 away from the final size and take two lengthwise finish passes. Wheel wear has been less of a problem with this strategy. Having a good flood coolant setup also made this process very enjoyable.
I enjoy using this small cylindrical grinder a lot when doing small batch production or for making tooling. It puts me in a good mood whenever I use it.
It is a Myford MG12 made in 1977, I have found very little information about the machines from this generation. All of the photos I have seen are from before the redesign in 1970 or the later hydraulic machines that were semi-automated. Never have I seen seen photos of machines that use the larger 12 inch wheel and are manual only, Myford must have sold very few of this configuration.
Please let me know any feedback you have about the process I used to install the linear scale. Some have told me I have made it too complex, but those who offered critiques were unable to describe an alternative.
I could not find a good place to mount the scale on the outside of the table's casting, Also the table can be inadvertently lifted and needs to be periodically taken off for maintenance.
So I choose to mount the linear scale on the underside of the table and the encoder reader-head is hidden inside the machine.
What I used was a magnetic linear scale with an adhesive back from RLS (Distributed by Renishaw), this is the product I used https://www.rls.si/eng/hilin
Here is a photo of the machine after the project was completed.
coincidentally the table has a machined ledge for the rack gear, this surface extends the entire length of the table but the last 18.125 inches was empty providing a perfect place to stick the scale. It was parallel within 0.001 of the ways. The gray band is the magnetic scale. (They ship them in coiled a movie reel looking canister)
I created a CAD model to plan out the clearances under the table and where to mount the readhead bracket.
It was a tight fit. it clears the reduction gear for the cross feed screw by about a 1/16 inch. any further away and I would not have enough travel.
The bracket that holds the readerhead locates off of 3 balls 0.2500 diameter 440C. A single M6 button screw holds down the bracket.
I could not easily mill locating surfaces for the bracket, instead I drilled divots for the balls with a long spot drill using a drilling jig.
The features of this jig was modeled as a sub-assembly containing sketch geometry driven from the v-and-flat ways, also the mounting holes locations. I used a a piece of TG&P 4140 shaft for the jig and relived the center so it sits on 2 points.
I used a M6 button head screw for the 3rd point and secured the height with some loctite after getting a measurement from the CAD model I swept the top of the round bar with an indicator then raised the knee to that distance and swept the screw head and adjusted it until the indicator read zero.
Here the locating divots are being drilled. I discovered that by using two shaft collars as a depth stop I could incrementally adjust the depth using a gage pin.
I needed to be 0.025 deeper so I shifted the rear collar 0.125" using a gage pin to set the gap then clamped the front collar over a 0.100" pin, Worked very accurately
You can see how the drilling jig is constrained by the ways of the machine.
Using a depth micrometer I checked how deep the spheres seated and compared with the CAD model measurements.
Here are the kinematic locating features on the underside of the bracket made from 17-4 stainless.
Another reason for the kinematic mounting is to allow for it to be detached and mill the readerhead seating surfaces to their final size after checking on the machine.
I left 0.030" of extra material on those faces for final adjustment.
The drilling jig was repurposed into a sled for the indicator checking how much needed to be milled off to achieve parallelism with the ways.
A depth micrometer was used for the other shinny surface.
Here is a picture showing it assembled. On the bracket underside are M4 socket head cap screws for the readerhead.
I needed to solder a DB-9 connector onto the cord because the pinout was different, this was my first time assembling these type of connectors, I had trouble holding the small wires while soldering each pin.
RLS had very good datasheets showing the tolerances for the readahead alignment and detailing the RS-422 quadrature signal, also detailed CAD models for download.
The readerhead cost about $350 and $50 for the adhesive backed scale with a custom defined refence point and length, the only bad part is the lead time of about 2 months for the readerhead and 1 month for the scale.
The repeatability has been excellent, I chose the 0.1um resolution output. It has an accuracy of 10um per meter when using the adhesive backed scale, the DRO I have set to a resolution of 0.00001 (10-microinch).
This machine is in a an climate controlled shop and the flood coolant temperature is monitored during operation.
None of the parts I have done so far require a tolerance better than +- 0.0002, but that extra decimal of resolution helps sneek up on the final size when I take it off the grinder for measuring.
Having a Z axis readout makes it a lot better grinding the shoulders and faces of parts. I also have found plunge grinding works best, I step over about 1/3 the wheel width and punge to 0.0005 away from the final size and take two lengthwise finish passes. Wheel wear has been less of a problem with this strategy. Having a good flood coolant setup also made this process very enjoyable.
I enjoy using this small cylindrical grinder a lot when doing small batch production or for making tooling. It puts me in a good mood whenever I use it.
It is a Myford MG12 made in 1977, I have found very little information about the machines from this generation. All of the photos I have seen are from before the redesign in 1970 or the later hydraulic machines that were semi-automated. Never have I seen seen photos of machines that use the larger 12 inch wheel and are manual only, Myford must have sold very few of this configuration.
Please let me know any feedback you have about the process I used to install the linear scale. Some have told me I have made it too complex, but those who offered critiques were unable to describe an alternative.
