Matt's RV-7 Project

I went to fire up my drill press the other day, and absolutely nothing happened. I figured it must be the switch, which is a really low-quality plastic thing, but tested okay on my ohmmeter. So, I tore the thing apart trying to diagnose the problem.

Long story short, it actually was the switch – it was just making intermittent contact, and when I tested it the first time, it gave the false impression of being okay. But while I had the machine disassembled trying to pinpoint the problem, I managed to crack one of the cheaply-made pot-metal drive pulleys. Argh!

Since the cost for replacement parts was going to be about $50, and the whole drill press cost me less than $250 to begin with, I thought hard about just junking it – after salvaging the motor and chuck, of course – and buying a better-quality drill press to replace it. But after pondering it further, I decided I really want my next drill press to be a floor-stander, and I just don't have the space for one of those right now. So, fifty bucks and a few days later, I had a new switch and pulley in my mailbox.

Here's what the drive mechanism on this model of drill press (Delta DP350) looks like. The pulleys on the motor and spindle are actually split into two halves, and control of the spindle speed is effected by varying the spacing between the pulley halves, thereby changing the effective pulley radius as seen by the belt. It's partially clever and partially hokey. On the one hand, this is the only low-end variable-speed drill press that doesn't require you to stop the machine and change belts to alter the spindle RPM; on the other hand, the low-end speed is too fast to use with a fly cutter, the special belt is expensive to replace, and the speed-change mechanism sometimes gets bound up. By the way, getting the collar and snap ring back onto the motor shaft while compressing that big spring was a fun chore.

All back together and working again. One thing I will say for this drill press, it does have a very good depth stop (which I hardly ever use), the chuck is easy to adjust, and it generally seems to run true. It's just too bad about all the cheap internal parts. Hopefully by the time it breaks again, I'll have more space for a bigger and better model.

In other tool-related news, I bought a lathe! I've wanted to get one for a while, and when I saw that somebody in town was selling a Sherline 4400 on Craigslist, I couldn't resist.

I got a pretty good deal, if I do say so myself. In addition to the basic lathe itself, I got a three-jaw chuck for the headstock, both a 1/4" Jacobs chuck and a 3/16" Albrecht chuck for the tailstock, a dead center, a live center, a lathe dog, a steady rest, a milling vise, and a collection of carbide cutting tools. And probably some other stuff that I don't even know how to recognize yet!

I have scant experience using a lathe, so this will be fun to learn on. I doubt I'll do anything fancy with it, but it should be great for making tubular spacers and the like. However, one thing I didn't like about the lathe when I got it was the fact that with speed knob dialed all the way down to zero, you don't have any indication that the power is switched on. I figured it would be bad to start messing with the chuck while accidentally leaving the lathe in a state where it could unexpectedly start turning, so I took apart the motor control box to see if I could rig up some kind of reminder lamp.

I bought a 120VAC neon bulb from Radio Shack, mounted it in the plastic cover, and connected it downstream from the power switch. I used high strength loctite to secure the mounting nut on the lamp, so it can't work its way loose and fall down onto the speed control board and cause a short.

Voila, an obvious red light now reminds me to shut off the power before adjusting the lathe.

The lathe will get put to work soon to turn out some spacers for the oil cooler mounting bolts. Meanwhile, I have been working quite a bit on reinforcements to the baffles where the oil cooler will be attached, but I still have some work to do before I'm satisfied with the design.

As I was making antenna cables, I thought it would be useful to show my method for crimping BNC connectors. The usual disclaimer applies: I am not an expert, I'm just showing you what I do.

I get my connectors from Aircraft Spruce, although you can find them just about anywhere. These particular ones are made by Amphenol and sell for about three bucks apiece.

Here's what comes in the bag… connector body, ferrule, and center pin:

One of the secrets is to use the right kind of coax stripper. The tool I use has three spring-loaded blades in it (kind of like my razor, I suppose). If you see one at your local hardware emporium, be sure to check the model number – all the ones I found in the local stores were configured for television coax cable. For stripping the RG-58/RG-400 coax we use in airplanes, the model 1255 is the one you want. I couldn't find one locally so I had to mail-order it, but it was comparatively cheap.

Since I'm a tool junkie, I experimented with different kinds of crimping tools. On the top is the "economy" racheting crimper sold by B&C for $40. On the bottom is an Eclipse frame with a set of RG-58/RG-400 BNC connector dies in it – cost for this setup is about $25 for the frame and $15 for the die, so it's basically a wash (prices are from Terminal Town where I bought mine). The dies on the economy tool are removable, and might actually be interchangeable with the Eclipse, but I haven't tried.

Here's a closeup of the crimp dies on both tools. On the left are the RG-58/RG-400 dies in the Eclipse tool, and you can see that there is just one crimp position for the pin and one for the ferrule. On the right is the economy tool, which can crimp multiple connector sizes. Either one works, although I find myself mostly using the Eclipse tool since I'm always crimping the same size of connector and it's impossible to get the pin in the wrong hole if you only have one choice. The other reason I prefer the Eclipse is that it makes a cleaner crimp on the pin… with the economy tool, the pin seems to get squashed flat a little more, so it's sometimes harder to get seated.

To cut the coax to length, be sure to use a real cable cutter, not a pair of dykes. The cable cutter has curved jaws that won't squash the coax insulation nearly as much, and it's cheap. These should be available wherever electrician's supplies are sold.

To strip the cable, clamp the stripper on the cut end using the illustration on the tool as a guide. It takes several practice runs to get the height-changing setscrews adjusted to give the proper depth of cut, so don't be afraid to use up a foot or two of coax getting your tool configured. To use the tool, use your finger to twirl it clockwise around the end of the cable (if oriented as it is in this picture) four times, then one turn counterclockwise.

If the blades are set correctly, you'll end up with (from left to right) a shallow cut through the outer insulation, a deeper cut through the braided shield, and a third cut that goes through the inner insulation.

With gentle persuasion, the cut segments should come right off and leave you looking at the untouched material underneath the cut. Take a minute to closely examine the cable for stray shield strands that might float around and bridge your two conductors together. You can leave the center conductor slightly long, since you'll probably be trimming it a bit anyway.

Slide the pin over the exposed center conductor. You want the base of the pin to be almost, but not quite, touching the white insulation. Trim the center conductor a bit at a time until it's the right length.

Squash the base of the pin with your crimper, and give it a tug to make sure it's secure.

Now slide the ferrule over the cable. If you're going to label it, now is also the time to slide on your heatshrink label.

Push the connector body down over the end of the cable until you feel a click as the pin is seated. The knurled bit on the back will dig under the braided shield. If you encounter strong resistance, can't get the pin all the way in, or don't feel the click, the most likely problem is that the crimper has deformed your pin enough to make it hang up in the connector body (see above).

When properly seated, the end of the pin should be flush with the end of the plastic insert inside the connector:

Without unseating the center pin, slide the ferrule down over the exposed shield. With the neato coax stripper I use, you should find that the shield has been trimmed to the perfect length and you don't have any stray shield strands poking out.

Now crimp the ferrule, and you're done.

I like to do a quick continuity check just to make sure the center pin isn't shorted to the shield anywhere. Antennas don't work very well with a shorted cable.

I got tired of the fuselage swaying and rolling back and forth while I was leaning over the gunwales working on something, so I took out the tailwheel fork and built this rig… it's just a plastic bucket filled with concrete, with a dowel sunk into it.

In the middle of the concrete glob is the business end of a toggle bolt, and the dowel is drilled to accept the bolt and act as a guide. After letting it cure for a day, I ended up with the world's heaviest nutplate.

To dress it up a bit, I cut a disc out of some scrap particle board.

Another piece of drilled dowel goes on top, and the tailwheel fork slips over it.

I made a wooden washer and bolted the tailwheel fork into the bucket of cement. It's now so secure that it might as well be anchored to the floor. I don't know why I didn't do this a couple years ago.

I can't afford a real two-inch standby attitude gyro (they cost a mint… and I'm talking this kind of mint, not this one) so I have a TruTrak ADI instead. It's not a real attitude indicator, but it should be good enough for a backup instrument. And since it's a safety-of-flight item, I wanted to make sure it was fed by dual power paths, as are the other most important avionics in my panel. Unfortunately TruTrak didn't provide dual diode-isolated power inputs like Garmin did, so I had to use a pair of 1N4001s to create my own:

Heatshrink tubing protects against shorts – after I took this photo I added another piece, then crammed the whole assembly into the D-sub backshell. One side is fed by Bus 2, which powers my important avionics, and the other side is fed by the standby battery. This arrangement will also keep the gyro alive while cranking the engine, which might be a good thing considering some of the funny behavior Matthew reported when he had his ADI powered on while starting.

Here it is mounted in the panel… I verified that it powers on, figures out which way is up, and receives GPS data.

Port-side view of the ADI wiring harness. The D-sub connector comes pretty close to the panel, but there's enough room to avoid kinking any wires. However, it would not have fit if I'd mounted it in the center standby instrument hole as I'd planned, which is why I moved it to the top.

"But I thought the ADI wouldn't fit in the top hole!" you say, or at least that's what you might say if I allowed comments on this blog, which thanks to the efforts of spammers I don't, thank you very much spammers. "You even posted a photo of this exact thing!" you might also exclaim if I let you. Well, yes, but that was before I figured out that a UMA light bezel inserted ahead of the instrument spaces it back far enough to clear the panel frame:

So thanks to the light bezel the ADI now lives at the top of the stack of standby instruments.

This is actually great news, because I was already planning to use UMA light bezels for my standby altimeter and airspeed indicator, and I was bothered by the fact that the ADI's internal lighting didn't match them in color or intensity. So, I broke out my Fluke and did some probing around… It turns out that I was able to defeat the ADI's internal lighting by connecting the 5V output pin on the dimmer unit to the ADI pin that's designed to be connected to a 24V lighting bus. What happens is that when I turn on the nav light switch, thus energizing the dimmer, the ADI sees five volts on its 24V lighting input (equivalent to 2.5V with a 12V lighting bus) and reduces the brightness of its orange LED display… but 5V isn't high enough to turn on its internal lighting, so I'm free to use my light bezel to light the instrument face instead. The dimmer potentiometers don't care about the extra milliamp or two that the ADI sinks on its lighting input pin. Excellent.

If you're still awake after reading the above paragraph… yes, I can be a huge geek sometimes.