With all the parts done I could finally start to install the ball screws, bearing blocks, servos etc.
In order to tension the z-axis belt during the installation I used two bolts with long nuts.
The quick connect fitting for lubrication of the Y-axis ball screw nut clears the casting by a few millimeters using the little adapter I made.
The use of feeler gauges simplified the initial alignment of the X-axis ball screw nut mount.
At this point it was time for another side project again as I realized that the nut on z-axis trapezoidal screw wasn't aligned properly when the mill was assembled at the factory.
The head of this machine is quite heavy so I instead of removing it altogether I rested it on the table of the machine using a few wooden blocks.
I then moved the table forward until I could reach the bolts for the nut.
With that done I had to go through the laborious task of the aligning the head which always takes a while.
Now that I'm going into the CNC era I removed the TouchDRO unit, optical scales and Android tablet from the machine and posted them for sale in this thread here at Hobby Machinist.
I also wanted to remove the stock electrical cabinet from the back of machine but this required moving the mill forward so that I could reach it as I have the mill standing against a wall.
In order to be able to move the mill without using the large engine hoist for tractors I've previously used I decided to build a little "dolly" for it.
I started by buying two dollies made for moving cars inside of garages.
I did this so that I could use the wheels from them as this was cheaper than buying the wheels separately.
I used two flat bars of steel underneath the stand of the machine and two L-shaped beams for the wheels.
I placed some wooden floorboards underneath the wheels to prevent them from digging into the rubber carpet I have around the mill.
Using this setup I lifted the mill about 10mm of the floor and carefully moved it forward.
I removed the stock electrical cabinet and all of the stock electronics from the machine. The only thing that was reused is the VFD. However I was surprised to some western electrical brand components in the stock electrical cabinet.
Apart from the VFD which I knew was a Emerson M100 unit I expected the rest to be no name China stuff. But there was a Delta 3-phase 24v 5A power supply and Schneider contactors.
With all the stock wiring removed I could redo the cable chain with the new cables, coolant hose, central lubrication lines and fog buster hoses.
Next I installed the VFD in the top left corner of the new control cabinet.
As VFD's can create a lot of electrical noise I used a shielded cable for the spindle motor and terminated the shield with a copper bracket right on the VFD itself to reduce the chance of any interference inside the cabinet.
I also used a sheilded cable for the control connections and terminated that shield by the Mesa card.
This is how the control cabinet looked at this stage.
With the major components in place I switched focus to the inductive Omron limit switches I've had laying on my bench waiting to be installed for several months.
I used aluminum brackets to mounts the switches and steel brackets to trigger them.
For the x-axis I was able to mount both switches and the center of the saddle in a very compact installation.
After installing all the limit switches I finished the last bits of wiring for the external estop and the probe.
Here is how the finished control cabinet looks. The oddly placed green cable at the bottom is for when I get a proper spindle encoder so so it's currently unused.
Did my first real test today with everything assembled.
Going to run the calibration software for the Delta B3 servos tomorrow. If everything looks good I'll probably try to mill some aluminium covers for the X and Y axis belts.
I've kept working on the machine and I'm getting close to be able to make some actual parts in CNC mode.
Ran the calibration software for the Delta B3 servos. The X and Y axis went very smoothly while the Z axis took some fiddling since I'm not using any counterweight or gas springs at the moment to compensate for the weight of the head. The calibration essentially eliminated any overshoot during quick moves which was nice to see.
Here is a video of the calibration running.
Next up I focused on the Slavetec Probe I bought a while back. It had to be aligned with the spindle which took 2-3 hours of fiddling with the adjustment screws.
After getting close with the hundreds of mm indicator I switched to a 1µm indicator and in the end I got a result that I'm very satisfied with.
Given how long time the alignment took I'm probably never going to remove the probe from that tool holder ever again so it now has a permanent spot in my cart with tool holders.
After that I was about to start numbering all my tool holders, measure the tool lengths and enter them into the tool table in Probe Basic (the LinuxCNC GUI I'm using). However I quickly realized that typing on the touchscreen was a bit suboptimal so I did a little side project and extended my control panel with a keyboard tray.
Since the last update I've been finishing the tooling in preparation for the first CNC chips.
Bought some more ER32 collets and BT40 ER32 tool holders. Made some room for them in my little tool card which lives beside the mill and measured all the tool offsets so they could be entered into Probe Basic. I'm planning to mirror the tool information in FreeCAD later on.
Cutting the Kaizen style tool foam takes a while but the end result is worth it in my opinion.
Ordered an Omron encoder for the spindle along with two HTD 3M pulleys and belts but I haven't installed it yet.
Here is a short clip showing the first actual CNC chips.
The part I'm milling is gonna be the belt and pulley cover for the Y-axis.
I've got everything up and running now. However when making the first parts it became clear that the belt covers I had neglected to make during the build process were a must so I've made covers for both the X and Y-axis belts now.
I used aluminum extrusions as the base for both covers with end plates made from billet aluminum.
Here is the Y-axis belt cover being made
The X-axis belt cover is designed a bit differently as one of the belt pulleys protrude a few millimeters below the servo mounting bracket.
The next items on the "Todo list" now are to replace the spindle bearings and get jog controls in Probe Basic to play ball with my physical jog buttons. However first I'm planing to make a few batches of parts as I'm itching to do some proper CNC milling
I've started milling the first proper parts after the conversion and I have so say that I'm very pleased with the precision and performance of the Delta B3 servos along with the extra preloaded Bosch-Rexroth ball screws.
Posted a short video on YouTube showing the results of milling 6082 aluminium using a 12mm DLC coated carbide end mill. The adaptive tool path is generated using FreeCAD.
The only thing I really have to look at now is the spindle which is completely untouched with the stock bearing, motor and VFD. Even at a measly 1500 rpm it gets too hot and the gearbox is way to noisy for long term use.
I have new bearings for the spindle on the shelf but I'm leaning more and more towards replacing the whole thing with a proper belt driven ATC spindle and Delta 2KW 6000rpm servo. That should bring down the noise and give much better control.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
