Well good news folks. After this weekends LinuxCNC meeting and fabfest here at my shop we managed to get the spindle encoder finally installed. I machined a beautiful mount on top of a round boss that sits on top of the motor. It looks factory almost and it fit perfectly. Then we were able to solder and double heat shrink the leads from the new encoder to the original resolver cable and it had the right amount of leads inside it to allow full differential encoder feedback to the 7i77 card and linuxCNC. Yesterday I went with a friend from the LinuxCNC IRC to Kentucky to look at a nice used Cincinatti Sabre machine and he may be buying it for himself to use in his shop. Most likely he will be following my lead with components and boards and drives/motors from what he says apparently.
Yesterday when we got back from Kentucky to my shop he offered to help me get this encoder wired up and configured in the software. He is very capable with the programming end of stuff and it appeared to be pretty easy for him. We were able to get the encoder wired into the 7i77 card very easily and then powered up the machine to configure it and look for the signals in HAL. Once we found the signals it was a simple matter of wiring together the signals and the proper linuxCNC pins to get it to output to the control. It sure was nice to see the numbers moving when we turned the spindle by hand. Then once we were sure the encoder was working complete with index pulse and everything we commenced to setup the scale to match the commanded and actual speed. That was just some number juggling and then we were getting accurate feedback to commanded speed.
The next problem was that the encoder, since it is mounted to the spindle motors shaft with a 2-1 belt drive ratio to the spindle actually outputs TWO index pulses to the control for every one turn of the spindle. That would not do and the original control used a 180 degree cam on the spindle top and a proximity sensor to control when the resolver was at index position apparently. So we needed a way to mask one of the index pulses for the spindle so we can use the single other index pulse for rigid tapping cycle as well as for Spindle orient. Pete from MESANET to the rescue!!! He was able to write a new custom 5i25 firmware upgrade for my card that allowed the program to mask the other index during the evolution of the 180 out proximity sensors high time. It is all very technical and Pete is a freaking genius with this stuff so all we had to do is enact a firmware update for the 5i25 and make some changes in the configuration and lo and behold it worked!! I cannot thank him and Connor and everyone who helped with this issue enough.
Once we got the input from the encoder working properly we decided to then turn our attention to trying to setup a rigid tapping cycle. After reading a bit on the forum we realized how to properly program the cycle and were able to enact a rigid tap cycle for a 3/8 16 thread .75 deep in a simulated air cut.. Then we played with several other cuts to test to make sure the spindle and head movement looked right. I was so pleased to see it actually do a controlled rigid tap with spindle feedback. I have not actually tried to cut a tap in material yet as it got late in the shop and he had to leave for GA but I was so pleased that it actually worked and my efforts on the spindle motor encoder were working.
After I play with rigid tapping on the control and learn to use that properly which is going to be AWESOME to have on the VMC the next step is spindle orient. Connor and several others seem to think this is just a matter of some HAL changes and we can test that soon. That will be a landmark because once we have accurate spindle orient the next step is enacting a toolchange. We are taking baby steps here I know but so far everything has gone rather smoothly and carefully and the results are that I have been able to build a very accurate wonderful machine here that makes some beautiful parts. Once the toolchanger is working properly that will be the end of it until I get my hands on a suitable fourth axis rotary unit. I fully intend to have fourth axis and hopefully a renishaw probe on this machine as soon as possible. This will really bring it into it's own and rival any of the HAAS machines I have run at places I have worked or other similar machines. I could not be happier with the Cincinatti Arrow 500 I have here, the MESANET cards I chose to use, LinuxCNC which is dead reliable and amazingly configurable as well as my decision to gut the machine and start over with brand new motors, drives, cables, etc. etc. I cannot thank enough all of the kind folks who have helped me along the way. Now I am actively looking for a suitable turning center to work on and do the same basic thing.
I just got some more orders for my custom designed droop compensating scope riser rails and some other parts I need to make so I am going to get back out there and work. I will update here more as the machine takes shape and we get spindle orient working as well as the toolchanger setup. Connor has apparently got a full blown working simulation of the toolchanger setup ready to test once the orient is reliably working. This is really getting exciting around here now. Peace
Pete