Machining CNC mill conversion project begins

Some of the pneumatic manifolds, fittings and tube for the one-shot oil system came in today from Automation Direct. I got inexpensive Nitra stuff without the more costly check valves to use for prototyping and experiments. Here I've rigged up a 2-port manifold that will go on the back of the column to oil the Z axis.

The ports are 1/4 NPT on the ends and 1/8 NPT along the sides of the manifolds. I also verified (as reported elsewhere) that the 6mm tubing will fit both the 6mm and 1/4 inch fittings. Here's another view of the assembled Z manifold. It still needs a plug on the left end.


Setting up manifold(s) for the XY saddle is a bit more interesting. But by checking travel and interferences inside the saddle I found that you can hang a manifold off the rear part of the saddle in such a way that it will not hit the column at max rearward travel nor cause any problems for the X table travel. This photo shows the idea with a 7-port manifold. I only need six ports here, but the sixes were out of stock at AD. If I can't find a cheap 6-port manifold I'll probably cut down the 7 and re-tap the cut end for 1/4 NPT (easy since the thing is already bored end-to-end with the proper tap drill size).

The other really good thing I found is that there's plenty of clearance to run three 1/4" tubes through the saddle under the X table to feed the oil points on the right side.

I'm still considering what size to bore the oil points on the saddles but am now leaning toward going to 1/8 NPT (Q size tap drill). I got a few 6mm tube x M5 thread fittings, but the 0.8mm bore through the center is pretty restrictive, and installing threaded inserts for them would be a complication. M8 thread fittings exist and would allow a slightly smaller tap drill, but the passage is still only 1mm, and you have to go to McMaster instead of Automation Direct and thus they are a lot more expensive.

Whatever size I choose, I've got a few more fittings to order, and also need to get a couple of 1/8 NPT taps.

Meanwhile the 4th axis servo drive and motor that will be used for the rotary table arrived. This is a baby 100W NEMA-23 motor with the small "L" series DYN4 drive. The drive got bolted to the electronics tray already. After I wire it up, the tray is completely ready to go into the electronics cabinet...badly needed to reclaim some workspace. I wanted to install this drive now to save having to repeat the major work of RnR'ing the tray in a couple of months. I just noticed too that I need to label the drives since the 400W L model is indistinguishable from the 750W H models unless you look at the label on the side.

The A axis 4th servo is all wired up and working. The 100W motor is a wee baby compared to the hulking 750W behemoths that move the main axes. Been looking at 6" rotary tables, so far the South Bend (Taiwanese) SB1364 from Grizzly looks pretty good for a conversion baseline.

The electronics tray itself is all but complete and ready to go into the case. I lack exactly one ring terminal for the A axis ground...trip number eleventy-seven to Marshall's coming up early Monday morning. Tonight and tomorrow I'll do a little cable dressing and labeling.

In other action today, I ordered some 6mm tube x M8 thread fittings from Amazon, and a 1/8 NPT tap set and M8x1.0 tap and the corresponding 7.00 mm tap drills from Drills and Cutters. I made sure to get the drills in the shorter "machine" or "stub" length which will enable me to do all but two of the saddle holes in my drill press. Tomorrow I may get as far as cutting down the 7-port manifold.

Rotary table search ongoing...after reading discussion on CNC forums I'm now leaning toward a Vertex (Taiwan) VH-6 or VH-8 table. Reportedly good quality and more table for less money. I'm leaning toward the smaller 6" size since a lot of what I do with it will be putting radial holes and slots on 2-5" diameter work. If the table is too tall in the vertical position it will obstruct the head/spindle from reaching workpieces that are substantially smaller than the table. The units look easily convertible to CNC drive.

All the Vertex hardware I have ever used has been excellent. We have quite a bit of it in our prototyping shop at work that I have had the pleasure of working with. As you have already found, good quality for reasonable pricing.

BTW, where did you get your Dyn4 drives and servo motors please? You have reignited my thoughts about a CNC conversion on my full-size mill .

@OverTheTop I got my DMM gear (Dyn4 drives and servos) from CNC4PC in Florida. DMM is in Canada, for shipping to the land of koalas you might want to order direct since they will probably know more about getting stuff delivered to Melbourne. CNC4PC drop-ships anyway. For industrial conversions I would also have a look at Centroid controllers. WarnerR just got an AllIn1DC card for his Bridgie clone; it has lots of well organized IOs and they have a number of professionally done example schematics on their website.

We interrupt this build to bring you a bit of actual rocketry. Here's Marc McReynolds at this morning's DART Apollo 50th anniversary commemorative launch, flying a model that he first flew in Scale Altitude...at NARAM-11...50 years ago this summer. Yes - today's flight was successful!


OK, OK, back to the mill.

HOLY NEGATIVE PROGRESS BATMAN!! The electronics tray has been demolished! No power, no back panel, no servos hooked up, what were you thinking duck??

Actually everything has been labeled and all the external connections taken off - the tray is ready for installation into the enclosure. The eagle-eyed will notice that the FWD and REV labels on the breakout board have been swapped into their correct locations (in previous photos the wire labels were correct but the ones on the connectors were backwards...finally fixed it).

I still need to terminate the A drive ground wire, but it's all ready for a test run. It remains to be seen how many case fittings may have to get pulled off to install the tray. The inlet plug for sure, the breaker and E-stop are in the "maybe" category. After the tray is in, the back panel and other case fan will go on, various things will get hooked back up, and the final power wiring from the inlet plug to the terminal blocks and to the 120VAC aux outlets will get done.

Finally, as promised, I did truncate the 7-port manifold down to 6 with the trusty chopsaw. The near end will get re-tapped for 1/4 NPT and plugged. The cut went through the mounting holes but it's of no consequence except cosmetic.

Progress on the oil manifold assembly - tapped the cut end, found a couple of 1/4 NPT plugs at Marshall's, and put in all the 1/8 NPT right angle male elbows. Looks like a multi-legged creature from the dark recesses of the garage.


But how to mount it? I ended up just making a bracket from a piece of 3/4 x 3/4 aluminum angle stock. The next photo shows how it will sit on the end of the left rear rail of the XY saddle. The inlet feed at the left end will be the only tube that has to flex during operation as the Y axis moves. The other 6 tubes leading to the oil ports on the saddle will be completely stationary. This is more or less the standard strategy for one-shot oiler retrofits.


I also got the ring terminals and put one on the end of the A axis ground, and re-bundled the servo drive ground cables on the electrical tray.

The tray went into the box with surprisingly little trouble. Only the inlet plug (top where the wires pass through) and the breaker (bottom right) had to be pulled out for clearance. Everything fits - all the iterations on layout for the DC power supply paid off at last. There is really no room for anything else in there without going double-decker.

It would be possible in this box to do a 5 or 6 drive unit for something like a knee mill with tilt/rotary, but the digital boards and power would need to go into a separate, smaller box.


I did need to use a couple of socket extensions to tighten a couple of the nuts, especially the one at lower left next to the DYN4 boxes.

Mounted the manifolds on the XY and Z saddles last night - #6-32 drill-n-tap. Here's the XY using a couple of #6-32 cap screws through the bracket into the cast iron:


And here's the Z setup, where I used #6-32 x 1.25 bolts going through the built-in mounting holes in the manifold. The tube on the upper port will go over the top of the saddle to the oil point on the other side of the saddle. I checked and it won't interfere with anything or get pinched at the top of travel. The lower port will have a very short tube loop that goes to the oil point just visible behind the lower elbow.


Yesterday I also received the MettleAir M8 elbows and straight fittings. It turns out that they are the M8x1.25 coarse threads, so another order just went in to DrillsAndCutters for an M8x1.25 tap set...the one I already have is for the M8x1.0 fine pitch threads.

That also leaves me with a small adaptation problem since the outlet fittings on the oil pump are the M8 fine thread type. The fittings are not available with M8-1.0 threads, so I'm going to turn a pair of brass adapters that will have male M8-1.0 threads, and a female M8-1.25 to accept the fittings. Should be pretty easy on the lathe; I did add an M8-1.0 die to the order make doing the male threads easy. Tooling never ends...

Now I need to choose a mounting spot for the oil pump on the machine stand. It's going to be on the left side and needs to be positioned so that the cap can be taken off for filling, and the XY and Z tubes can be routed with enough-but-not-too-much flexibility to handle the saddle motions.

On to drilling the saddles. The 1/8 NPT tap drill (size R) is *much* larger than the drill needed for the M8 fittings (an "I" drill is about right for M8-1.25), so I'm definitely going with the metric fittings. The internal bore is larger than expected so I don't think there will be problems with flow. Now that I have the fittings and know the threading for sure, I can actually start drilling.

Saturday night lathe work - made the adapter bodies for the oil pump fittings from some 0.5" round aluminum; couldn't find any solid brass in that size in my drawer o' metal. The small ends will get die-threaded M8-1.0 and the large end tapped M8-1.25 when the tap and die show up. Key dimensions: small diameter 0.313", large diameter 0.475", large end drilled with an "I" drill to depth 0.70", thru drilled with a #20 drill (0.161"), length of large cylinder 0.75". This took about half an hour.

It never ends! great write-up!

Hehe Warner I hope some of it ends soon...I'm planning to attend NARAM and LDRS and need to build that GoDaddy 54 for a 20k flight.

Making a push today to get the electrical box in shape.

Improv way of holding a batch of wire ends that need to be tinned that can't be removed completely. I discovered that those funky push-to-release connectors on the C62 BOB are *much* easier to manage if the wire end is tinned. This is the wire bundle from the spindle remote connector. The wires aren't actually clamped in the pliers, just bent slightly with the weight holding them in position.

Same wire bundle after trimming to length, dressing and routing with some Noble 3/8" self-adhesive wire guides. There are a LOT of zip ties inside this box.

...and it's off to HD...yet again...for some bolts to hold the case fan that goes in the round hole just to the right of the power inlet in the photo above.

Huge milestone this evening - the electronics box is complete!! All hard-wired connections are in, the aux outlets are wired in and tested with a newly made Y connector on the hot leg (also note white neutral wire added going up the center), case fans both running in the desired push-pull setup (HD had the 6-32 x 1.5" machine screws needed to mount the one in the lower right), grounds tidied up, and a couple of last things covered with heatshrink tubing.


And...the door actually closes. Here it is, finally upright with the computer finally hooked up properly via the back panel instead of via surgical intubation. The base and various mechanical parts can be seen in the background. The sheet metal door is a little warped but that will not be seen much since the box is going to sit on the ground on the right side of the machine.


The only things I can think of that I might still want to do the enclosure would be to cover the unused holes, and replace the noisier case fan (the stock one from CNC4PC) with a quieter one. The extra one I got for the back panel is a nice low profile Noctua unit that puts the other one to shame sound-wise. I'm also noodling over some possibilities for vinyl graphics on the case, and maybe some thick rubber feet.

Coffee has been essential to the project. Coffee purists please don't roast me over the green lady stuff.

Suddenly I have hit the point of having beaten down almost the entire backlog of things I can do until the M8 taps get delivered. Feels weird...hoping to be able to start operating the machine by the All-Star break.

Excellent work on the electronics box! That is quite well put together. I am looking forward to seeing everything running! The workmanship on all of your parts has been pretty excellent so far in fact!

It is just a little bit painful to me that all this work is going onto the Precision Mathews mill. That kit would go excellent on a nice bed mill. Though the several tons those weigh, are certainly not nearly as convenient to move around...

I'd start out real gentle on the acceleration with those 750W servos on your table. For reference we use some 2.2 kW ones to move a several ton, 20 foot antenna dish at a pretty high rate of speed. It does go through something around a 50:1 gearbox, which nets it a considerable amount of torque. But I don't think you will want for power in moving the table around; the flexing of the machine itself will definitely be the limiting factor.

Looking forward to more updates!

@Xrain Lol I'm pretty much working with the largest mill that will work in my available space. There's not enough vertical clearance nor floorspace to do anything that requires a hoist. At the outset I found that many knee mills would not fit through the garage door, and even getting a baby Tormach in there would be problematic. I might have been able to go up to a ~1/2 ton unit but they start getting pretty tall. The cool thing is that this control box can actually be applied to anything that can be driven with the 3x0.75 kVA motors; it could easily be used with a larger mill later. A fourth large channel could be added later via an outrigger box if needed for a knee etc. Tuning the movement will be interesting - the gearing is 1:1 but I already have plenty of torque.

Earth station antennas eh? My company makes those over in Atlanta. I don't interact much with that team right now but have definitely been around that game. Used to be involved professionally with optical design and telescope control, backlash is my personal enemy

Getting a knee mill in can be a pain. I have a crane company I know that have some neat equipment that put my mill (960kg) into my workshop. I just paid the not terribly expensive fee to move the mill in ( and the lathe years before). Effortless. I made a removable section of wall years before, so when I purchased the mill access was relatively easy through the garage.

Aha, no one expects the ol' removable wall section! Riggers are for sure a decent and economical enough idea for moving heavy equipment. They will probably come into play when I get the 500 kg+ 12x36 lathe.

Here's a schematic of the one-shot oiler system with components listed. All the parts are in hand, so unless there is some fit/routing problem through the XY saddle this should be accurate. Thinking tonight about whether I can mount the oil pump up on the side or back of the column. Underneath the chip tray on the base looks like it would be a pretty un-handy location.

Back to posting...dribs n drabs for the last few days but a productive weekend day yesterday. Decided to mount the oil pump on the back of the column; putting it low on the base seemed like it was going to be inconvenient, and pushing the oil uphill also seemed less favorable. There are four mounting ears but I only used two since the pump is a bit wide to allow all four. It's plenty sturdy. I didn't remove the unused ears in case I want to relocate it later. I drilled and tapped two 1/4-20 holes in the column and used 1/4-20 x 0.75" stainless socket head cap screws.


Continuing with oil pump related things, I die-threaded and tapped the little adapters shown above. They thread into the square brass fittings on either side of the oil pump. The outside threads are M8-1.0 and the inner tap is M8-1.25. They ended up fitting perfectly. Die threading using the lathe chuck as a holder:



Tapping, same way (by hand, not power-tapping):


Finished product:

Setting up the Z axis grease line extension. I had to argue with this to get grease to inject through it...turns out there was some blockage behind the zerk. I ended up drilling the ball and spring out of the right angle zerk to reduce the backpressure too. The extension drops down below the sled and can be fished out through the slot in the column. Ultimately it all works as intended.



The Z axis getting installed. To do this you first insert the aluminum bracket down through the top of the column. It will not go in through the front via the slot! Then put the ballscrew down through the top of the column with the nuts positioned 2/3 of the way down toward the end, but do not bolt it yet. Now slide the aluminum bracket into place over the end of the leadscrew and onto the ballnuts. Now slide everything back up toward the top of the column and thread in the 6 bolts that hold the bracket. You need to work through the top of the column here, which is why you need to have enough of the leadscrew exposed so that you can use your T-handle hex wrench on the bolts without the Z motor bracket getting in the way.



Looking in the end of the column, ready to put in the bolts:


Now we can bolt the Z motor bracket loosely to the top of the column and get ready to put the sled on. Oil up the ways and the gib strip, start the sled onto the ways in the orientation shown, and insert the gib strip followed by its restraining bolt.

Z sled attachment continues. Side the sled down until you can insert the two bolts through the holes in the sled and into the aluminum traveler block. Make sure you use the bolt holes with the slots, not the pair of holes closer to the ring which were for the gas strut.


Now test for free motion by driving the sled to both ends of travel. By far the easiest way to do this is by attaching the outer half of the shaft coupler to a piece of 1/2" rod and chucking that in a 1/2" electric drill. The drill should be able to move things easily, without any binding. When you are happy with the motion, you can tighten up the bolts shown above, as well as the bolts holding the Z bracket atop the column. Mine went perfectly on the first go, no fiddling required.

Some quick photos of the sled modifications for the oil tube fittings. I ended up threading all the oil ports for M8-1.25 fittings, using an "I" tap drill - officially supposed to be 6.8mm but those are a little hard to get, size "I" is close enough. You need to drill them about 3/4" deep in order for a plug tap to create enough parallel threads. I found that I needed to get about 11 turns from where the tap just started to engage.

The cast iron drills pretty easily. Use low speed and plenty of oil on the drill. I did it by hand and the drill had a slight tendency to grab; when that happens, stop immediately and you can easily reverse the drill back out. A "pecking" technique helps quite a bit. Once you get the feel of it, it goes really fast. There are 8 ports to do (6 on XY, 2 on Z) and I think the whole operation took about an hour.

Here I'm using a straight fitting for testing the threads; these will actually all be male elbows.



On one end of the XY saddle there is a small bit of interference from the casting. I threatened it with the Dremel tool, but that wasn't enough and I actually had to take the fitting out and use the Dremel to grind away a bit of the casting.


The MettleAir fittings don't have pre-applied thread seal compound, so you need to install them with Teflon pipe tape. I used the yellow tape with the high gas/petroleum rating. Two wraps and trim the excess.




The XY saddle with all fittings installed:

Exactimator came over Sunday morning and we lifted the column (with Z sled) onto the base. With two people it was pretty easy; we carefully planted the column end on the base and lifted the other end, kinda like the famous Iwo Jima picture except we were not getting fired at.



There was a minor gotcha, and the photo above holds the clue - with the column in place you can't install the Y screw! Fortunately there was an easy solution - we removed 3 of the column base bolts, leaving the left near corner one, and rotated the column about 45 degrees around the bolt, which gave enough room to get the Y screw lined up with the hole out in front of the base.

I put the XY saddle on the base and plumbed up the oil tubes for it. This required playing with the tube lengths and locations so they would sit in reasonable locations when the X screw is installed.



It actually worked out pretty well and I didn't have to mess with drilling holes for hold-down clips. Interestingly, I did find out that hot glue will not stick to cast iron. Kind of amazing, usually it sticks to anything and refuses to come off. The aluminum X bracket actually helps keep the oil tubes in place since it doesn't move relative to the saddle.

Next the X table went on. BTW I followed the assembly procedures given several posts back, they turned out to be pretty accurate. Having a second person to lift the free end of the table while you install the gib is a *real* help. If I was doing it alone I'd need to find something to prop up the free end of the table (Jawstand + something for it to sit on probably). For the X axis, the shaft coupler is not threaded and there is no Belleville washer on the free end - the shaft is fully retained and preloaded at the other (pre-built) end. Once again, the electric drill ersatz drive helped confirm that the table moved full range without binding.



The ultimate X travel did not increase as much as I'd hoped. Basically the brackets run into the XY saddle casting after only about 1/2" of overtravel. I've got just about 22.0" of absolute X travel to work with.

Finally time to put the head back on. This was the biggest finagle of the weekend - I had to figure out how to do it single-handedly.



The key turned out to be to have the outside part (nearer in the above picture) tilted up as you get the 3 fiddly bolts lined up into the holes. I used the board to do that. In the picture I've already taken the board back out from under the front of the head. Once the bolts are in, you can lift the head and get the big round rotation axis cylinder lined up.

Next the motor went back on, threading the cables through the holes in the head and then re-terminating the motor power cable which won't go through the hole.


Finally I bolted the spindle electrical box back on the left side of the head and re-connected the motor cables to the controller board. This required dismounting the Z oil manifold...maybe I will move it to the other side.



You also need to re-connect the spindle tach cable to the display on the front panel (not shown). Finally I buttoned everything back up and fired up the spindle with the transfer switch in manual mode. Success!



Suddenly it's pretty much a mill! It just needs to have the servo motors bolted on and cabling arranged into the control box.

Hi Dave,

Looks like I have shown up late for the party and have a lot of catching up to do. I found your thread searching for a spindle control solution for my PM-30MV conversion. I'm very interested to see how yours turns out.

Now off to catch up on all your post!

Mark

Hi Mark,

Glad you found it! I've hidden a CNC project in a rocketry forum, because rocket guy The as-built spindle control diagram is in post #23 here: https://www.rocketryforum.com/threads/cnc-mill-conversion-project-begins.150928/#post-1884754 (Note that it updates the earlier diagram). It works great and the transfer switch lets you revert to manual spindle operation as needed for drill press type work etc. if you have any specific questions about what I did, feel free to post here or PM me.

Update covering the last few days - waiting on a variant NEMA-34 mounting plate from ArizonaVideo99. The one I have is a few mm too thick and the motor shaft won't engage into the coupler.

I can't just mill down the one I have because the counterbore shoulders in the inner ring of mounting holes would be lost.

Reinstalled the quill spindle lock to keep it in place when using the main Z drive for CNC operations.

...and hooked up the servo power and encoder cables using some Noble 3/4" split loom to protect the cables where they exit the electronics enclosure.

They are secured in place using a 6" piece of 1" heatshrink, plus a cable tie.

Next I did some work on cable dressing, using some 3/8" and 1/2" plastic cable clamps fastened to the column with #8-32 x 5/16" socket screws, some cable ties, and a couple of pieces of OneWrap velcro.

I spent a bit of time playing with various arrangements of the cords on the right side of the column to keep them from touching too much as the X and Y axes move.

The next and last major thing will be the limit switches, but I really need to get the Y axis in jog operation to do the positioning. Finally today I got a few things on order including a Kurt vise, some parallels, and some Vactra 2 way oil. I'd call the build 90% complete now.

Tooling up - the Kurt DX6 machine vise and a parallels set arrived. The DX6 is every bit as righteous as its reputation...perhaps a little big for this mill, but I felt a DX4 was going to be too small and would not carry on well to the eventual larger mill (and they are also out of stock). The Y travel modification will let me reach the fixed back jaw of the vise with the vise fastened to the middle T-slot. Ordered from allindustrial.com in Huntington Beach; quick service and decent prices. Poor UPS driver - 70 lb on the doorstep The jug of Vactra 2 way oil came yesterday. As soon as the thinner Y motor bracket plate arrives I'm gonna be makin' chips almost immediately.

The only actual work I got done on the unit in the last couple of days was to cut 3 pieces of angle stock for the Y and Z limit switch brackets, and install a few short M8 bolts in the slide lock holes to trip the inductive sensors.

Here's a slightly worrisome thing I found that had a surprisingly easy solution. I started mounting some R8 tooling in the spindle taper, and quickly found out that some things would fit and others would not. The chuck's keyway would admit the alignment pin inside the spindle, but the collets wouldn't. After looking at them, I could see that the keyway on the chuck was both wider and deeper than that on the collet. Hmmm...for a little while this didn't look good for the home team.

But after a slight bit of research, it became apparent that a) the keyways on R8 shafts are nowhere near consistent, and b) the alignment pin in most of the PM-xxxx mills don't do anything useful and can be trivially removed.

The pin is actually just a small setscrew in the side of the spindle. You can get access just by lowering the quill all the way (lock the drillpress handles and turn the knob on the front panel so the quill will stay down). In the next vidcap the offending item is visible in the lower part of the vertical oval slot. You can just unscrew it with a hex driver. Be *very* careful not to drop it down inside the spindle or you will probably need to go find a magnetic pickup probe to get it back out.


Here is the little rascal after removal. As somebody observed on another forum, it's far too wimpy to actually transmit much torque, and you really don't need it.

So another part hits the reject bin, and suddenly all my R8 gadgets fit perfectly. A couple of posters in CNC forums mentioned that the PM-727 drawbars are sometimes too long and need washers added at the top, but the one in my PM-30 seems fine.

First chips!! On to some (kludgey) milling...I'm waiting for a thinner Y axis NEMA mount plate from ArizonaVideo99, but in the meantime I thought I'd try to modify the one I have. Of course it's fitting that the first thing the mill will do is help modify itself.

The mount plate would optimally be about 8mm thinner to allow the motor shaft to fully engage in the coupler, but there is a ~7 mm shoulder for the bolt head, so I figured I could safely remove only about 4mm, leaving a 3mm shoulder. That would get me to about 5mm of engagement on the motor shaft, which seemed minimally adequate.

To do this operation I had to move the Y axis manually, using the usual trick of mounting the outer coupler half on a piece of 1/2" drill rod and chucking that in a cordless drill. For the X motion I used some G-code to program the distance and feed rate, and jogged around through the gaps with the pendant. I turned the spindle on/off manually via Mach4 after setting the speed with an S1500 M3 command. I just went back and forth across the back of the plate in lawnmower fashion, stopping after each pass to move the Y axis by hand. Not full CNC by any means, but exercised quite a bit of functionality including G-code execution, pendant jog in X and Z (and man the position readout on the pendant is the bee's knees), spindle control, and constant-rate X feed.

It came out great, and the thinned plate is now in service. Here's the first photo of the fully put together mill, lacking only the limit switches.


Next work will be squaring up and slotting the limit switch brackets, and to make a cover plate for the banished Z crank handle opening at the top of the column.