So yeah, it only took a few weeks for me to break down and buy a real milling machine. Like you didn't see that coming.
Actually I hadn't intended to get one right away, but I was unexpectedly lucky on eBay and got a very nice 12" Sherline mill to go with my 3.5×17" Sherline lathe. Judging from the few parts I've made with it so far, I can definitely see that a dedicated machine is a lot nicer and more capable than a lathe with a milling conversion attached.
The mill seems to have hardly been used at all, but since it's at least fifteen years old and doesn't seem to have gotten much TLC during its life, I decided to strip it down to its component parts so I could clean and lubricate everything. I also replaced the AC cord and switch, and added a power indicator safety lamp as I did with the lathe. After putting everything together, the mill worked a lot smoother and with less backlash, so it was worth the effort.
Then while I was at it, I tore down the lathe to individual pieces too, and gave it the same tune-up treatment. Then for some reason I decided to really get crazy and add DRO scales to it too. That was more challenging than I initially thought, and took a couple weeks to really get sorted out, but it did give me a chance to use the new mill to make real parts. And now that the lathe is back together, it too works better and has digital readouts to boot. (I may add DRO to the mill as well, but if I do I'll almost certainly take the easy way out and buy a kit!)
Hopefully this will all come in handy for the airplane project. Don't worry, I haven't totally forgotten that I have one of those too.
A while back (several years ago, in fact) I installed a manifold pressure fitting on the firewall, using the location referenced in the plans. Fast forward to today, through several iterations of equipment selection, and I no longer have a need to bring the manifold pressure plumbing through the firewall. I do, however, need to transition from the heavy Aeroquip hose to something lighter, and also to plug the now-unnecessary hole in the firewall.
I chose to tackle both at once by modifying the bulkhead tee fitting that I previously installed through the firewall. I cut off the right angle leg, filed it flat (this was before I had a milling machine, or else I'd have used that) and tapped the hole for 10-32 threads. Then I installed a similarly threaded quick connect air fitting using Permatex #2.
Here's the modified air fitting re-installed in the firewall. The aft side, not shown here, is blocked off with an AN929 cap. This effectively turns the former tee fitting into a firewall-mounted right-angle hose-to-tube adapter.
This is a wider shot of the area in question. Manifold pressure is conveyed from a port on the back on the #3 cylinder to the fitting/adapter on the firewall via a hose, which is strain-relieved via an adel clamp attached to the engine mount. This originally had two attach points; I may hook the other one up again someday if it appears necessary.
From the firewall connection, the manifold pressure hookup transitions to 1/4" nylon tube, seen here as a black stripe because the camera wouldn't focus where I wanted it.
I attached a quick-connect tee fitting to the firewall using a simple aluminum tab affixed to an existing bolt hole. The middle leg of this fitting will eventually connect to the manifold pressure input on my ignition system (more on that in a future update) but for now it's just plugged. A pair of adel clamps keeps the manifold pressure tube from rubbing on the oil pressure hose and vice versa.
Here you can vaguely see the entire route of the black plastic manifold pressure line, from the hose transition on the starboard side, across the firewall through a tee to the sensor manifold on the pilot's side. In retrospect it's not exactly how I might have chosen to hook it up had I known I wouldn't need to bring it through the firewall, but it's not a bad arrangement regardless.
The electronic transducers that sense engine oil pressure, fuel pressure, and manifold pressure are all mounted on a manifold on the left side of the firewall. I ran the wiring to all three, then used some brass air fittings and the compressor to test the oil and fuel pressure sensors. Manifold pressure is easier to test since you can just compare the readout to ambient air pressure.
I used Permatex #2 to seal the sensors, plugs, and fittings as I screwed them into the manifold. The oil pressure sensor is hung off the side of the manifold, since the oil pressure switch (which drives the Hobbs meter and the low oil pressure warning light) is much too big to be placed anywhere else other than cantilevered straight out from the firewall. In fact, I discovered to my chagrin that the oil pressure sensor doesn't quite fit either – the radius is too big by an eighth of an inch – but I have a plan for that…
In classic style, I used this little challenge as an opportunity to buy a new tool. Seen here is a vertical milling table attached my lathe, which turns it into a very small milling machine of sorts:
I cut a piece of 3/16" aluminum bar stock, match drilled it to the transducer manifold, and then bolted/clamped it to the new milling table and milled out some lightening holes:
It's not perfect, but it's not bad for my first real part made using a milling machine. It's quite light, since most of the material in the heavy bar of aluminum has been milled out.
Now the manifold can be spaced out from the firewall, allowing the oil pressure sensor to fit:
…just like this:
I secured all the sensor wiring with tie wraps. Later I'll connect the three wires for the oil pressure switch.
The milling conversion attachment for the lathe wasn't all that expensive, but I have a feeling that it is going to end up costing me a more in the long run. Now that I've successfully made a useful part on a mill – something I never thought I'd be able to do – I'm already having impure thoughts about upgrading to a larger and more capable machine. I wonder if Mary would notice if I put a Bridgeport in the guest bedroom?
I hereby deem the cowl and baffles to be as finished as they're going to get, at least for right now. To celebrate I removed the forward top skin – which had become quite dusty! – in order to get caught up on some wiring tasks.
I connected and secured the standby alternator field wire and B-lead… note adel clamps and strain relief:
I had previously mounted the current sensor for the main alternator, but I never got around to wiring it or installing its twin that measures current from the standby alternator. The second sensor I installed with an adel clamp from the engine mount, right above the fuse holder where the standby alternator B-lead connects. Since these are 100-amp sensors and the standby alternator is only capable of 20 amps, I looped the wire through three times in order to achieve a little better resolution on the display. A calibration step in the G3X software allows you to apply a scale factor of 0.33 to account for this trick.
Here's a wider shot showing both current sensor hookups. To make them serviceable I used mini molex connectors, which are shown here prior to being secured in the wire bundles.
I wrapped the connectors with silicone tape in order to make them somewhat waterproof:
Then I powered up the avionics and calibrated both current measurements to zero. The machine is starting to wake up…
One of the many places on the airplane where builders are given the opportunity to customize is the little door on the engine cowl that lets you stick your hand in and check the oil level. The standard RV oil door is a piece of fiberglass with a piano hinge and a couple of quarter-turn fasteners – nothing wrong with that, but I wanted something a little different. I had the fiberglass door on my last airplane, and it had a tendency to flex during high-speed flight, and would even pop open unexpectedly on occasion. The solution, of course, is to either add stiffness to the fiberglass door, or build the door out of something stronger. I had some 0.063" aluminum laying around, so I decided to make the oil door for this airplane out of metal instead of glass. I also decided that I wanted a different hinge and latching mechanism, for purely personal reasons.
I started by squaring up the edges of the cowl cutout and making the corners nice and round:
Then lots of iterative fitting, trimming, and bending to get the aluminum door to fit the opening. The cowl has a gentle conic section at this point, a shape which the aluminum doesn't naturally want to follow without some smashing persuasion.
Here's the aluminum door, shaped and trimmed, sitting next to the stock fiberglass door that came with the kit:
Instead of the standard piano hinge, I bought a "hidden hinge" from Nonstop Aviation (whose name always makes me think of an old Brian Aldiss novel, which by the way was not great). Lots of folks install this type of hinge, including plenty of certified airplane factories. It's just a piece of regular hinge with a sort of gooseneck extension riveted on, thus allowing the pivot axis to be tucked cleanly out of sight underneath the cowl – I could probably make my own if I had a decent bending brake. It works the same as a piece of piano hinge, of course, but it looks a little nicer. It's spring-loaded too, although I left the spring out while I was fitting the oil door so it wouldn't launch itself across the garage.
Here's what it looks like with the door closed – no visible hinge line:
It takes some adjusting to get the door to open properly without binding, and without the inside edge crashing into the cowl and scratching the paint.
The hinge is flat and the inside of the cowl is curved, so I made it flat by laying up an epoxy/flox mixture and letting it cure with the hinge clecoed in place:
I wanted to use a push-button latch instead of the quarter-turn fasteners called for by the plans, so the oil door can be opened without tools and without anything sticking out into the breeze. There are basically two choices here: the Hartwell H5000, which holds very securely but is hard to make look nice, and the Camloc KM610, which is easier to install but trickier to make latch properly. The tie-breaker for me is that the Hartwell latch when operated tends to spring open like a demented mousetrap, and I've bruised my knuckles on them too many times to want to go to the trouble of having one on my airplane.
Also, pay no attention to the prices in the preceding links – those are new-certified prices, which are ridiculous. There's a surplus place in the neighborhood that sells new-condition Hartwells for ten or fifteen bucks and Camloc pushbuttons for five, so I picked up a handful of the latter in various sizes to play with:
Since they were so cheap, I removed the spring from one and drilled a hole in the exact center of the button, thus turning it into a drill jig for properly locating the mounting holes relative to the main hole:
It was then no sweat to drill the holes to mount it, although you do have to be super careful about where you position it relative to the edge of the cowl cutout. Too far forward, and it won't latch; too far back, and it will latch easily but it won't hold properly. Patience counts here.
So does test-fitting:
Here's a view from the underside. Note how the latch tongue is pretty short as well as pronouncedly rounded, which is what causes the difficulty.
To get the clecoes out of the way and make sure the whole works was properly rigid, I riveted the latch and hinge to the door:
As pictured, the latch tongue bears on an un-reinforced fiberglass surface, which of course won't do at all. To fix this I fabbed a little striker plate out of some thin stainless steel material I bought from the K&S display down at the local hardware emporium:
A single flush rivet holds the striker plate to the cowl:
Once I had the metal parts built and the overall mechanism working (which was fun) it was time to make the fiberglass look cosmetically acceptable (which is never fun).
I protected the door with packing tape, then latched it in place and squeegeed a flox/microballoons/cabosil mixture into the gaps between the door and the cowl.
After a couple iterations of sand-fill-sand, the cowl was a good match to the shape and contour of the oil door. But I had a problem – there wasn't enough gap around the door to let it actually open. Hmm, obviously zero-clearance is no good here.
What I wanted was something I could put around the circumference of the door to provide a uniform separation between the metal and fiberglass for a cast-in-place operation. I experimented a bit with some of this nylon grommet edging material I had laying around, but it wasn't quite right. Kind of an interesting idea to keep in mind for the future, though.
In the end I hit upon the idea of using some silicone fusion tape instead. This is close to ideal, since it's fairly thick, resists epoxy, and follows the curved corners without wrinkling. I dremeled out a gap around the cowl opening, then wrapped the door with two layers of silicone and latched it in place.
Then I spackled in a new layer of filler:
The result, after puling the door out and sanding down the high spots, was a set of nice straight edges and round corners. I forgot to take a photo of it, but this technique gave me a fairly uniform 1/16" gap all around the door opening.
As I expected, the clearance was a bit less around the corners where the rubber tape stretched, but I got that cleaned up pretty well with a file.
I attached the hinge to the cowl with stainless screws and tinnerman washers, the better to keep the fasteners from pulling through the fiberglass. The grey splotch here is just a misting of primer I sprayed on as a guide coat to help me sand down the filler.
Inside, I made a little hinge pin retainer out of some angle stock and a leftover hinge eye I found in the scrap pile. It picks up one of the mounting screws and keeps the hinge pin from getting away.
Here you can see how the gap looks – it's a bit deceptive in a photo because the coloring is uneven where I sanded, but in person it looks pretty nice. I'll let the painter fix up the last 10%.
Here's a video I shot showing how the whole thing operates. Dig that solid latching action!
And now I need to clean up the garage before I do anything else. What a complete mess.
