Disclaimer

This blog is for entertainment purposes only, and is not meant to teach you how to build anything. The author is not responsible for any accident, injury, or loss that occurs as a result of reading this blog. Read this blog at your own risk.

Thursday, February 28, 2013

"Tooling up" - Mill's X & Y axis


Mini-Mill's digital X and Y axis conversion

Manufacturing most of the metal parts for my airplane, implicitly meant accepting a longer building experience, this however didn’t mean that I couldn't improve my situation whenever I could.

One of the issue that was affecting my milling ability and precision was the Z axis backlash, this was effectively dealt with by installing a DRO (digital read out) on the mill column (see the conversion here). I have been very pleased with the outcome since the upgrade, and my machine is much more precise, faster, and easier to work with.

I have been considering doing the same thing to the X and Y axis for sometime, and I finally broke down and bought two remote display DROs.


micromark.com items #85290 & 85291


Having DROs on all axis will turn my mini-mill into a “Cadillac”, so to speak. I will be able to precisely locate every feature of a new part quickly, accurately, and with excellent repeatability. The small backlash present on both axis will be irrelevant from now on, since I will be reading actual table movement, rather than hand-crank turns.

As always, nothing is ever “bolt on” at this level, and some minor adaptation had to be devised since there were no guidelines on how to mount them. 

I found a good location for the X axis DRO behind the table, where I could keep it away from harm and below the table top. Instead of using the standoff brackets that came with the kit, I decided to mount it directly to the table, so I had to improvise my own mounts. This enabled me to save precious space, and not reduce the Y travel. I opened the sending unit, removed the circuit board, and drilled and counterbored two holes for 4 mm bolts, then with the unit in place, I drilled into the table to locate the matching holes. 


Sending unit disassembled

One hole will be enlarged, the other will be where the + sign is.

Holes drilled and counterbored (unit upside down)

Counterboring removed enough material to keep the head of the bolts flush

Slide testing the sending unit


Happy with the result, I tapped the holes.


Holes center-punched, drilled, and tapped. 


I had to also drill two holes on the DRO casing to allow me to tighten the bolts. I later put some tape over them to prevent chips and cutting fluid to get in it.


Back case drilled, and board reassembled.

This is how I will tighten the bolts

Sending unit in position.


The Y axis proved a little easier, and did not require me to take the table down again. I just drilled and tapped two holes on the mill base, and one on the side of the sliding table. I adapted one of the brackets supplied with the kit and was back in business.


Tapping the first hole for the Y axis

Rear Y axis hole

X and Y axis sending units mounted on the mill

Ops check... Good!

I need to find a better place for the displays


I know, the column is not a very good place to mount the DROs, since the mill head has to move up and down, so I made a bracket from a piece of scrap aluminum that came with my mill.


New "instrument panel"?

The red box is plastic, but the instrument panel is very light weight

Drilling and tapping a few more holes

VoilĂ ! 3 axis DRO mill!

Close up of the panel


I was able to route the wires away from all moving parts, including myself, and I feel pretty confident that they should be safe where they are.


Wiring DIY




Friday, February 22, 2013

Retractable heated Pitot tube - part 1


After solving the temperature problem that had plagued my heated Pitot tube concept, I was finally able to devote a little more time to the retractable tube idea. I will work on that until the circuit board etching supplies arrives, then it will be R&D  time once again.

I’m not really sure how good an idea it is having a retractable Pitot tube, but that’s what my friend Wade really wanted, so I decided to do my best to deliver it. Personally, I like the fixed one better, it has very few parts made out of aluminum, they are pretty easy and relatively fast to make, and it will not have any issues associated with moving it back and forth over time.

Cognizant of the fact that anything that moves will eventually get loose and/or break, I wanted to devise the simplest and strongest solution I could think of, that involved the least amount of moving parts. 

The only other retractable tube I had ever seen online was pretty clever, it required a bolt floxed parallel to, and at a specific distance from the tube, a flange with a machined slot to engage the bolt, and a spring loaded ball thingy also floxed at a critical distance to match a hole in the flange. Very interesting setup, but cumbersome to operate, requiring the opening of a nose hatch, and the release of the spring loaded ball using a pen. On top of that, it was not a heated probe.

I suppose this might be a good time to talk about why making a retractable/heated pitot tube is so complicated.

The issue is really pretty simple to visualize, but the solution has been elusive. Let’s take a look at the tube I made for my friend Mike...


Mike's heated Pitot (serial #001)


In this arrangement, the G10 insulator is permanently attached to the nose structure, and the tube is held by friction to the insulator (Mike will also tie the tube down somehow for extra peace of mind).

Now, suppose I was to polish the inner diameter of the insulator to make the tube able to slide freely in it. Immediately you can realize a few of the issues. First, there needs to be a way to restrain the tube from sliding all the way into the nose in flight. Second, the wires are located in a position incompatible with the free sliding tube, if we are to avoid damage to the electrical components.

My tallest objective was being able to retract the tube with no tools, without needing to open any hatches, without having to disconnect anything, and being able to do it blindfolded, as in... without light (late night departure perhaps). If that was not enough, I wanted this operation to be super fast, with a maximum allotted time of around 10 seconds.

Good luck, you say? That's exactly the way I felt!

Many of the solutions Wade and I explored could have made this probable, but most required welding of some sort. Now, some of you might not be aware of the fact that welding 2024 aluminum is unreliable at best, 2024 cracks quickly when welded, and the part turns into junk in short order. One of our other crazy ideas was to make the tube out of stainless steel, then do the necessary welding. We later dismissed this idea as cheating, poor design, and heavy, but we saved it as a fallback option.

As the head scratching went on, it appeared that most of our solution were either easy to machine but difficult to work by the pilot, or difficult to machine but easier for the pilot to operate.

Here’s an example of one I liked, for a while...


Testing a quick disconnect idea


This is the grenade approach, where you first have to pull the pin


"Exploded" view


This was pretty simple to make for me, but required that the pilot opened a hatch, lined up the holes, and inserted the pin. Unfortunately all of these seemed designed to make the pilot’s life more miserable, and when faced with the very confined space in the nose of the plane, this idea was dropped to the bottom of the trash bin.

Finally, while on a work trip, I sketched what eventually became the core of the final design solution.


Brain farting on the road!


In the sketch you might notice that some of the G10 is missing from the right side, that’s because I was still hanging on to the “wires coming out of the side” scheme, a fact that only worked to complicate anything else I tried to do. The next leap forward didn’t occur until I finally divorced myself from this concept, and machined the slot into a solid piece of G10. 

One last unknown was the strength such a tab would have, but this was settled after milling the first prototype. The G10 tab was super tough, and I couldn't even begin to bend it using both hands.


Milling the retaining tab in the G10


Testing the locking logic... Retracted Pitot.


Extended Pitot


90˚ clockwise twist


Extended and locked Pitot


Nevertheless, I'll try to leave a little more "meat" on the tab for extra insurance in the final version.

With the current setup, the heating element wires would have no choice but to come out the rear end through the back cap, and while I had briefly thought about it in the past, I just could not imagine how it would be possible, given the gigantic hole the final vacuum fitting already required. 

There was just not enough metal to put 2 more holes for wires in the back cap!

Or, was there?

How about increasing the diameter of the tube to 1” (2.54 cm)?

Man, the last thing I wanted to do was placing a small cannon in the nose of the EZ! No way! 3/4” (1.91 cm) was it! We would just have to try harder.

Thinking out loud with Wade, the idea came up of some kind of adapter that could move the giant vacuum hole to a separate piece of aluminum, while the back cap could endure two more very precisely drilled holes. It sounded possible in theory, but I wasn’t so sure I could pull it off, so I started modeling this adapter in CAD.

Designing in CAD lead me to realize that, with a wall thickness of only 0.035” (0.89 mm), the adapter would have to be made out of steel, if it was to last. Did I mention that machining aluminum is so much nicer than steel? Oh well!


Cutting a piece of 4130 steel out of a 3/4" bar


Future steel adapter


Milling flats on the lathe! (the mill was down for maintenance)


Adding a thread to the adapter


Finished adapter


Vacuum fitting screwed into the steel adapter


Aluminum back cap, steel adapter, and vacuum fitting


Basically, I replaced a 0.355" (9.017 mm) hole in the aluminum back end cap, with a 0.250" (6.35 mm) one, giving me an extra 0.105" (2.67 mm) to work with. 

This was a good starting point, but I still did not have enough space to get the wires out yet. I thought of chamfering the steel adapter to create some room.


Just enough aluminum to drill the two holes


Creating some space to get the wires out of the back by chamfering the adapter


It was pretty obvious that there wasn't much room for inaccuracies while machining these parts. Additionally, in order to create even more exit room for the wires I decided to chamfer the aluminum back cap as well.


Design getting even more complicated


A lot going on in a very small space



Trying to wrap my head around this weird assembly (please excuse the dusty screen)



As if this arrangement was not complicated enough, there was still one more piece that needed to fit, I had to press a steel pin into the back cap. This pin would ride in the groove I milled into the G10, and become the star of the whole retraction mechanism.


Ok, this is starting to look like a space ship!


I sure hope this is the last modification


But enough virtual machining, let’s make some chips!


Setting up for chamfering the adapter's shoulder



Creating the rounded shoulder



My lovely "assistant" landing a hand to help Wade


I needed to eliminate the sharp edge produced by the mill, so that the heating element's wires would not chafe on it. I used my lathe to produce a smooth contour.


More tight confines machining


The wires ought to enjoy that smooth ramp


Next, I very carefully positioned the mill head in order to produce the wire exit holes, while leaving enough material on either side of them. The holes were positioned exactly 60˚ apart, so that they would align with the "beefier" side of the hexagonal bolt pattern, in correspondence of two of the vertices. 


Rotary table on its back once again


Drilling the wire holes


Adding a small chamfer to help guide the wires


Then, it was back to chamfering...


1/4" mill at work on the back cap (note the rotary table is vertical again).


So, here’s what we’ve got so far (minus the hexagonal bolt pattern, and the pressed steel pin)...


First successful rear exit of the heating element wires


Strangely, the bottom of my end mill produced an unsightly line along the 6 faces. That ticked me off a bit, but it’s pure cosmetics. Next time I’ll machine the faces with the side of the end mill.


Milling the hexagonal 15mm bolt pattern



milling the last flat face


15 mm bolt pattern? Check!


Test fitting the components


The final manufacturing of the back end of the tube involved the steel pin. 

Critically important was preventing the drill bit to cut too deeply, through to the center air passage. I calculated a 0.200” ( 5.08mm) depth, and I was easily able to stop at this preprogrammed value using my Z axis DRO (see my digital read out conversion).


Spot drilling the back cap


"Blooming aluminum"


Reaching the final depth of cut 


Ready for pin insertion

Ultimately the pin was cut to size on the lathe, and press fit into the aluminum piece.




Here’s what the locking mechanism looks like...


Final rear end testing... Retracted


Extended and locked


And, here’s how the wires exit from the back...


Complicated perhaps, but successful.


Note: The wires seem a little short now because I have not properly sized the outer tube yet, I’m just using a left over piece of tube here that is probably too long.

Next time I'll try to finish the front end of the heated Pitot.



This is what it should end up looking like (outer tube semitransparent)