Check twice, cut once

For the lower receiver on my M16A1 build I'm using a Cerro forged 80% from Right To Bear.  I got it on sale as a "blem" for $30, and for the life of me, I can't find anything wrong with it that would make it a blem.  I'm not going to reprofile it into a "correct" A1 style lower(maybe a future project), but since the rest of this is all Colt M16A1, I decided to engrave the side of my lower to match.  Colt used several styles of markings on their lowers through the years.  With none being more "correct" than others, I decided to just make what I liked based on what I'd seen.  I couldn't find a good CAD drawing of the Colt horse logo that they used, so I ended up drawing my own from scratch.  I got it all drawn up and programmed, and headed to the mill...and that's when the trouble started.

This engraving is the whole reason I upgraded to ball screws on my CNC.  I actually started this project back in January, and only now with the mill done was able to get back to it.  I almost always run my programs on a test piece first, and it's a good thing I did because it didn't work...at all.  I started with a chunk of blue foam, and you can see the results on this test piece here(painted silver for more contrast).

It's a darn good thing I ran a test piece first, or I would have effed up my lower because both the lettering and the logo were very messed up.  The first two attempts were just...wrong...and I couldn't figure out why.  The CAD file was fine, the programming was fine, the toolpath preview on Mach3 was fine, but it wasn't machining correctly. The only think I could figure was that the stepper motors were skipping steps.  I upgraded to ball screws thinking that the more coarse screw pitch would need fewer steps, so it would be less likely to skip.  I was wrong on that too, it wasn't any better with the new screws(although still definitely worth the upgrade for many other reasons).  After much head scratching and internet searching, it turns out all I had to do was increase the pulse width going from my PC to the stepper driver.  My CNC PC is pretty old, and borderline fast enough to run my mill.  It just couldn't output a clean pulse with Mach3's default pulse width, causing the motors to miss steps.  Increasing the pulse width allowed the stepper driver to read the signals better.  Think of it like the difference between yelling "Ah!" and "Ahhhhhh!"  Same message but your brain can process the second one better because it's longer.

With the mill upgraded and test pieces finally looking good, I could do the real thing.  I had to get a little creative to hold it in my vice.  I have a solid block of iron where the trigger would be so that the lower is firmly held in place without the vice touching the thin areas of the mag well and trigger guard.  It's hard to see here, but the engraving turned out perfect.  I left off the Colt manufacturer info block by the selector since this isn't, in fact, a Colt made part and I'm not really trying to fool anyone who takes a close look.  After engraving and while on it's side and all indicated in, I CNC drilled the FCG holes too.

The fire control pocket was next.  I had to get creative with the clamping again, because these things are a weird shape.  As much as I like "hit go and let it run" machining, I had to babysit this thing the whole time.  My coolant mist system doesn't have the air pressure needed to blow the chips out of a deep pocket like this, so I had to pause it and clear the chips out frequently.

It took a while, but that's a pretty nice looking pocket if I do say so myself.  I also left a bridge of material between the the new FCG pocket and existing takedown pin pocket, so even if I hadn't crushed my autosear in a vice, it wouldn't physically fit into the lower.  It's not really necessary and most ARs have this whole area open, but when building a former machine gun with a lower receiver marked like a machine gun, I want my intent to be very clear.

New project time: a retro M16A1 build

Black rifles are like belly buttons, everybody's got one.  It seems like practically everyone has an AR-15 of some sort, everyone but me anyway(unless you count my 715T, which is just a 22LR 702 Plinkster in an AR shaped plastic shell).  So, I decided I should join the club.  Most ARs are lego builds, just buy the parts and slap it together, and this one won't be much different.  The big difference here is that instead of buying new parts I'm starting with a genuine Vietnam era Colt M16A1 parts kit.

My goal for this one is the "battlefield pickup" look.  These are Colt parts, and they're only original once, so I'm going to leave it looking well used.  I got my black rifle parts kit from Royal Tiger Imports on Black Friday when they were on sale.  The packaging from RTI was just plain awful.  All the parts were just thrown in the box with a single piece of bubble wrap around the buttstock.  The gas tube wore a hole in the corner of the box and the bag with all the little bits was broken open.  Consequently, a lot of the pins and little bits rattled out of the box, so I'll have to rob them from a lower parts kit I bought for another project.  After cleaning all the gunk off the parts, here's what I ended up with:

This is a Grade B kit and all in all, not too terribly rough for what it cost.  The forging codes on the upper indicate that it was made some time between '74-'82.  The parts are very well worn on the outside, but the internals look great, like the gun was carried a whole lot but seldom fired. The handguards are very rough with many cracks and missing teeth, but that's not uncommon with these early style handguards and this was a Grade B kit.  I plan to try to fix them, so we'll see how it goes.

I'm going to build it on a standard AR-15 80% lower, and I have to make it a semi-auto only rifle.  The parts kit was a machine gun, and it included all the naughty bits.  You could just throw them away and buy a standard lower parts kit, but I'm cheap, and I like the idea of using as many Colt parts as possible.  Before I even started machining the lower I had to take care of them.  Most of the naughty bits can be converted to semi-only, but one part can't be used at all.  The very first thing I did was crush the auto sear to pieces in my vice, making it completely unusable.

Next up was the hammer.  The M16 hammer has a tail on it that needs to be ground off, making it practically identical to a standard AR hammer.  Original on the left, modified on the right.

Next was the disconnector.  Like the hammer, it's got a tail on it that needs to be removed.  A few seconds with a grinder fixes it right up.

The selector lever is up next.  It's got a fin sticking up in the middle of it that needs to be ground down.  Modified in this way the selector still spins all the way around but the positions give you safe-semi-semi.

Last, the trigger needs some welding.  The AR trigger has a closed back, the M16 has an open back because of the tail on the FA selector.  The end of the trigger needs to be welded closed so that a FA disconnector will not fit.

Just changing the selector or disconnector would make it work as a semi-only rifle, but for legal reasons you need to modify all the parts so that there is no question about your intent.

X2 part 4, Z end is here

Today's episode of Making It Up As I Go Along is the Z axis on my Sieg X2D.  When I initially did my CNC conversion, I used the "make it work, and upgrade later" method.  Originally, I modified the mill's fine Z feed system to be motor driven.  Inside the original fine feed assembly, it uses a worm gear to turn the sprocket to move the quill up and down on the rack.

All I can say about it is that it worked.  It didn't work well though.  It was slow.  It had almost .060" of backlash and gravity was the main backlash reduction system, the weight of the head kept it tensioned in one direction.  There was also a ton of lead error in the rack, meaning if I told the mill to move .100", depending on where in the travel it was sometimes it would move .095", sometimes .103", it was just all over the place with no consistency.  TBH, it would probably work fine for most people's machining projects, but it was a pain to work with when trying to actually hold a tolerance.

So, just like the X and Y, a ball screw was the solution.  There are a lot of ways people have added Z screws to their X2s.  The easiest is a side mounted screw with the ball nut attached to the head.  Since I've already had problems with the Z axis binding, I didn't want to side mount it because lifting from one side would only make it worse.  So I decided on a center mounted screw.  In order to keep the screw out of the chip zone, the only practical way to do it is to fix the screw to the head, and drive the ball nut with the motor.  There are a couple of designs using this method out there, but I didn't really like any of them.

So I came up with my own.  I, of course, didn't take any pictures of the finished part, so you get a CAD model instead.  It's very thick so there shouldn't be any flex in it like some other designs have.

Actually mounting it to the mill was another sticky point for me.  The easiest way would be to have brackets on the sides that bolt into the column, but I am always hesitant to modify original parts when I make things up like this in case I change my mind half way through.  So what I did instead was use some eye hooks cut down to fit inside the column, then a pair of bolts goes through the back of the column, and the hooks get tensioned against them which pulls the Z mount tight against the top of the column.  It is not the strongest mounting system, but for the amount of load the rest of the mill is capable of handling, it's more than enough.

Making the ball nut the part that spins was a bit of a challenge too, much more complex than turning the screw would be.  Not only do you need to turn the ball nut, you need a way to take up any axial thrust in the system.  I ended up with a stack of parts that has an annular bearing on the bottom to handle the side load and the axial load pushing down, and a standard flat thrust bearing on top(for reference the bearings are an NTN SF0725 35mm x 56.4mm x 14mm annular bearing, and a 35mm x 52mm x 2mm thrust bearing).  I designed the mount to use shims under the top cap to set the bearing preload.  Here's how the ball nut/bearing stack looks:

With all that sorted out, I could finish the assembly and test everything out.  The screw is fixed to the head using the bolt holes for the gas spring that the mill used to use to lift the head.  All in all, this ball screw conversion was a fair amount of work but definitely worth it.  I've got about $160 in ball screw conversion parts, and my backlash, which started at .006"/.008" X&Y, and .060" in Z, is now down to .001" in all axis(and even that I could get rid of by ordering slightly oversized balls for my ball nuts if I really wanted).