Symptoms of Contamination in the Carburetor Float Bowl

Things were going well and running fine until I decided to take my friend James for a ride. He had never been in a small plane before, so he brought his daughter, who is about the same age as our oldest, and we took off for a sightseeing hop. After about 15 minutes, the engine coughed really hard. This caught my attention and called for a calm reaction. The engine was running fine after the momentary power loss, so I selected the mixture to full rich and started heading for the airport. We landed uneventfully until it was time to taxi up the hill to get to the hangar. The engine ran fine in the idle range, but it wouldn’t run above 1500 RPM or so. Since we needed more than that to taxi up the hill, we shut down the engine and pushed the airplane back. This was certainly not the best impression for a first-time airplane ride.
I downloaded the diagnostic data from the D180 and uploaded it to Savvy Analysis. After much consideration and consulting with Bob Barrows on the phone, I decided that it must have been a carburetor ice event. The main symptom was that the carburetor’s main jet was plugged, but the problem had gone away. The carburetor air temperature indication did not support this theory, but then again I’ve heard other folks say that theirs indicated higher than actual temperatures. At the time, the ambient conditions weren’t exactly prime for carburetor ice, but it was cloudy and about 50 degrees. I ran the engine for half an hour on the ground, making passes up and down the runway to make sure it was operating correctly. I flew solo for a short flight, and then flew for almost an hour with another friend. The carb ice theory was starting to become the most plausible.
That is, until I went flying with another young couple that I had met recently. This time we were planning to fly for about an hour, but just as I was turning on what would have been a crosswind leg if we were staying in the pattern, we lost power again. Just as before, the power loss was very temporary. It was running again by the time I could do anything, without any input from me. This time we were within gliding distance of the airport, so I landed promptly, assuming it might quit temporarily or permanently at any time.
After landing, the symptom was the same. Any power application above the idle range would cause the engine to quit. This time I had the forethought to try and determine if the problem was a lean mixture or a rich mixture. I advanced the throttle into the quitting range, and then rapidly pumped it back and forth over about one inch of travel, to work the accelerator pump. This worked, indicating that the problem was fuel starvation. I used this technique to taxi up the same hill to the hangar, and sheepishly apologized to yet another set of passengers. They were more understanding than I expected.
This was all that I needed to stop flying the airplane until I was exactly sure about why these symptoms were popping up. I did lots more research, and found an old service bulletin about modifying the carburetor to deliver more fuel. The RV guys call this the “Mooney Mod” and it dates back to the 1960s. I asked Bob about this, and he said that was probably not related. He offered to swap carburetors with me, which was generous. I asked what he would do in such a situation, and he said he would take the float bowl off of the carb and see if anything looked suspicious in there. We knew that it had a brass float, but I wondered if it had begun to deteriorate somehow.
I felt comfortable with this, being armed with a detailed parts diagram and plenty of non-airplane carburetor experience. I didn’t really see how there could be anything in there though. After all, fuel flows from the tank, through a finger strainer, to the gascolator (which would separate out just about anything), and through a fine screen right where it enters the carb. The gascolator and carb screen were spotless, as usual.
I took the carb off, which was about a 30-minute job, and cleaned off a section of work bench for surgery.
MA4-5 Carburetor Float Bowl
I carefully collected the fuel from the float bowl, wanting to be sure to document any debris. There were a couple of little floaties, but nothing significant. Then, I looked into the float bowl and found the evidence I was looking for:
MA 4-5 Float Bowl
You might have to click on the big version of the picture to see the little piece of red RTV. I am pretty sure that this little bugger was what was causing all of the trouble. But how did it get in there? It’s way too big to get through any of the screens, and how could it be rattling around in there for 130 hours without any other symptoms? Well, regular readers may remember a time that I removed the float bowl drain under less than ideal circumstances. The best explanation that I have is that this was when the piece was introduced. Of course that was 60 hours ago… but maybe the piece just finally eroded to a point where it could cause trouble.
I put everything back together, though I misaligned the mixture parts, so the mixture didn’t work. I had to drop the float bowl down again and carefully ensure that the mixture adjusting piece sticking down from the throttle body was engaged in the tube at the bottom of the float bowl. After that, everything ran great. It has been running great for around 30 hours so far, with no more problems.
The takeaway is that we should be extremely careful when messing with the carburetor float bowl drain. There are so many defenses in place to keep contaminants out of there, but when the drain plug is out, those are all bypassed.

Posted on
Hours Logged This Session:
Total Hours: 1883.75

Symptoms of Losing the Idle Mixture Screw

We had some complicated plans for getting to Oshkosh Airventure 2014. The problem was that we had a wedding to go to in Boston on the Sunday before Oshkosh, and we had to be back at Oshkosh in time to set up the cookout on Tuesday. The plan we came up with was for me to position the airplane to Oshkosh in advance by myself, then to airline to Boston to join the girls. Then after the wedding, we would all airline back to Oshkosh and enjoy the show. I’m oversimplifying the plan here, because there is obviously no airline service to Oshkosh, but if I explain the whole thing here you’ll likely get bored and quit reading. So to get back to the Bearhawk part of the story, I loaded everything up that we would want for our trip but not need in Boston, and set out on Thursday morning for what should have been about 6 hours of flying. For some reason, the first trip to Oshkosh for several Bearhawks has historically been an unlucky one. If mine had been uneventful, I wouldn’t bother writing this entry, but as you will see, it was not.

Weather for the first half of the trip was so-so; it was good enough to go, but it was less than ideal. My plan was to stop just northeast of Lexington KY at a small airport with cheap fuel. From there I could top off and fly all the way to Brodhead WI to camp for the night, then fly the short hop to Oshkosh on Friday morning to pick up a rental car. As I made my way over the mountains towards Kentucky, I was easily able to maintain VFR conditions above variable cloud covers below. At one point I climbed as high as 11,500 briefly, and the airplane was performing well in spite of a substantial load of “stuff.”

The forecast called for clearing conditions in the Lexington area when I was scheduled to get there, but as I got closer, I could see that those improvements were running a bit behind. I was on top of a flat but solid layer that I could see was breaking up to the north. I was showing a fuel quantity on board of around 8 gallons, and while that’s an hour of flying, it’s not much! One plan was to call Lexington approach and file a pop-up IFR to get through the thin layer. Once underneath, I could cancel and proceed VFR to my original destination. The only problem with this plan was that for some reason, my primary radio (the only panel-mounted radio) was not transmitting my voice. Are you starting to see how a superstitious person might be able to wonder if these glitches were related to my destination? The only way to use that plan would have been to revert to my handheld radio, which didn’t seem wise for IFR in Class C airspace. So I looked at the map, and saw that Frankfort was the next option past Lexington. It was farther than I wanted to go, but based on wing root sight gauges and Dynon fuel flow-based calculations, I had enough fuel to get there with legal reserves.

I was flying right over the top of the LEX Class C airspace as I saw for sure that the clouds were not going to be a problem. The clouds and the airspace led me to stay higher than normal, instead of making a power reduction to start descending towards FFT. Once clear of the outer airspace ring, I reduced the throttle to near idle to start what would need to be a circling descent. That’s when the engine started missing. It wasn’t entirely obvious at idle, because in that condition the relative wind is driving the prop and engine, rather than combustion. As I opened the throttle a little to see if it was indeed missing, I confirmed that the engine was running intermittently.

This was not cause for dire concern for a few reasons. One was that because of my high descent profile, I had enough energy to make it to FFT without any engine power. Another was that I had enough altitude to tinker with it while I was on the way. Keep in mind that the power-off glide in our Bearhawk is on the order of 500 feet per minute, allowing around one minute of glide per thousand feet of altitude. This meant I had around 8 minutes to sort it out. While these factors meant the power loss was not a dire concern, it was still a pretty serious concern, mostly because I didn’t know why it was happening. I assumed that the most likely cause was fuel delivery, because I had not yet ever flown the airplane with so little fuel on board. Maybe it was sloshing around somehow and not getting to the engine. Of course I realize now that I could have ruled this out if I had checked the fuel flow gauge, but at the time I didn’t think of that.

The situation became doubly undesirable when I finally had the FFT runway in sight. It was well within my gliding range and beautiful to see, except for the two big X marks. That’s right, it was closed for resurfacing. At that point, I had no intention of landing anywhere else. An off-airport landing is much less desirable than an off-runway airport landing, and while it was not exactly safe to land on a closed runway, I could see that the runway was clear of any personnel and equipment that I might have endangered by landing there, and since I was exercising my emergency authority, coloring outside of the usual lines was allowed to get the airplane on the ground safely. If the runway had not been clear, I could have made use of the taxiway, or even the surrounding grass.

I landed uneventfully and coasted off of the runway, and attempted to restart the engine to taxi onto the apron. It would run, but not well. The airport manager met me at the airplane to amicably express his discontent for me having just landed at his closed airport. I explained that it was an emergency landing, and that we’d sort it all out, and he agreed. We pushed the airplane to a tiedown spot and I started recalculating my complex travel plans. The airport manager said that he wouldn’t be able to let me depart from the closed airport, which was really going to complicate things. Fortunately, the construction project was nearing completion, and the airport might end up reopening on Saturday, or more likely Monday. The only viable option was for me to leave the airplane there and get to the wedding, then return after the wedding to fly the airplane to Oshkosh.

After an explanation to the friendly (and quite reasonable) local FSDO inspector who wanted to hear why it was that I felt like landing at the closed airport was the safest course of action, I took a taxi to the Lexington airport. My plan was to catch an airline flight to Chicago, with plans to catch a second flight to Appleton. From there I’d try and figure out how to get to OSH, then try to figure out a place to sleep there for one night, then wake up on Friday morning to pick up the rental car and resume the original plan. Notice that the details of that plan start getting sketchy at the end. That’s part of the adventure of flying a GA airplane on a long trip- one is never really sure that he’ll get there at all, much less according to the original plan.

The flight to ORD was uneventful, though a little bit delayed. Fortunately the crew that worked that flight was also the crew for the Appleton flight, so it was also delayed. While in the gate area I noticed one of my coworkers, Mark; this is one of the few perks of having to all wear the same flammable work clothes. We had not met before, but we talked for a little while and he explained that he was picking up a rental car in Appleton to drive to Oshkosh upon arrival, and offered me a ride. This turned out to be most useful. I figured if I could get to the show grounds, I could find a place to sleep, even if it was under Chris Owens’s Camper RV, which he had pre-positioned in his usual campsite. As we flew from Chicago, I looked out the window to see the beautiful green countryside, blue sky, and calm winds. I really wish I had been able to fly the Bearhawk instead! Conditions were just perfect. I had a great visit with Mark as we drove from Appleton, and he dropped me off in the campground at OSH. It was somewhat surreal to be on the grounds before the start of the show. Things were certainly ramping up, but the airplane parking areas were primarily empty and there weren’t any crowds.

After sizing up the grass under Chris’s camper, the scant clothing that I was wearing/carrying, and the forecast overnight low, the option of sleeping there was becoming less appealing. I had brought a tent for sleeping at Brodhead, but I was having to travel light, so it was still with the airplane. Then it occurred to me that it was Thursday, and while the local hotels are historically unobtainable during the show, I thought it would be worth asking what the rate would be. When I called the Super 8 the lady told me that they had plenty of room, and the rate would be $70. I didn’t need to take a second glance at the grass to make my decision! I started walking in that direction, enjoying the memories from the special place, the beautiful sunset, and the mild temperature. I was not enjoying the airshow that the mosquitos were putting on around my head, but they only served to remind me of 100+ good reasons for an indoor sleeping space.

After checking in I walked over to Friar Tuck’s for a hot crock of french onion soup and a roast beef sandwich, my first consequential food for the day. The next morning I woke up and continued with the travel plans, including procuring a used child seat for the rental car. The seat I had planned to use remained in the airplane, a casualty of our fragmented plan. The logistics of all of this trip would have been complicated without the need for carseats! It’s fun to have little kids, but they step up the travel complexity significantly. I drove the car down to ORD and parked it as planned, and took a flight to Boston. From there I took a bus to Cape Cod to catch up with the rest of my family tree, arriving in the middle of the night. We had a great time at the wedding on Sunday, and well before the sun came up on Monday, Tabitha’s parents took us to the airport so that we could get back to the Midwest. The girls flew to ORD to pick up the rental car, and I flew to ORD to catch a flight to LEX. After another expensive car ride to FFT, I was finally back in position.

We added some fuel to the airplane, though not a top-off due to the high fuel price. The airport was not open yet, but it was expected to be open that afternoon. This gave me some time to troubleshoot the voice problem with the Garmin 430. My research over the weekend suggested that removing and reseating the radio in the tray might fix the problem, and fortunately it did. I prepared the airplane so that I could hop in and go as soon as the airport opened, and after a few hours of waiting, the manager finally gave me the thumbs up. I hopped in and started the engine, and it wouldn’t run. This was most frustrating.

It was acting like it wasn’t getting fuel. If I pumped the throttle it would stumble along, and if I got the RPM above 1500 it would run reasonably well. A local airport patron named Gene was hanging around, and he helped me with some troubleshooting. He was a long-time owner, pilot, instructor, and home builder, and was immensely useful. One problem that I found was with standing water in the airbox. The FAB instructions say to drill a hole in the back to allow water to drain, and somehow I had omitted that step. Shortly after Oshkosh, I saw a mention in one of the magazines that said this is a very common problem to pop up at Oshkosh, since so many of us keep our airplanes outside for the first time while there. I drilled a 3/16″ hole and allowed a cup or so of water to run out onto the ramp, hopeful that I had fixed the problem. I hopped back in and started the engine, only to find no change.

I’ll save you a lengthy review of everything that we checked, including draining the float bowl, which required removing the airbox, which required removing the cowl. With all of that stuff out of the way, I happened to look at the back of the carb to see a hole where the idle mixture adjustment screw was supposed to be! This explained the problem. The engine would run fine at high power settings, when the idle circuit was not in use. At low power settings, the idle circuit was running very lean, since there was a 1/4″ air hole where the needle was supposed to be. The friendly local mechanic loaned me a needle off of one of his grounded airplanes, and after reassembly, everything ran great. It turned out to be harder than I expected to replace his needle. I started the process of ordering a replacement at the Aircraft Spruce booth, but it ended up taking several weeks, in part due to the complicated nature of ordering carburetor parts from ACS. This ended up creating a most-confusing situation in the subsequent weeks, where when I saw the shipment finally leave ACS, I sent the mechanic his original needle back, but at the same time he called ACS and redirected the new needle back to FFT, presumably assuming that I had run away with his needle, etc. Our carb is the ubiquitous 10-3878, but it’s not readily clear which variants of the MA4-5 use the same part number for the needle.

The good news was that I had a running airplane and another hour or two of daylight, so I finally got out of FFT and headed north. OSH closes at night during the show (and I don’t fly the Bearhawk at night anyway), so I knew I wasn’t going to be able to make it all the way, but I also knew that closer was better. I stopped in Indiana for the night, topping off on cheap fuel and sleeping like a king in my little tent. I wasn’t sure if camping was allowed at the airport, so I asked the tower controller when the first folks started coming around in the morning, and set my alarm for a few minutes prior. The next day I was cranking the engine just as the first car was arriving, and just as the dawn sky was lightening.

I flew from there to southern Wisconsin, where I topped off on fuel again. I didn’t want to have to worry about fuel while at OSH. I arrived at about 9:30 in the morning, and taxied to the Bearhawk Aircraft display booth. Mark’s other options for display airplanes fell through, so he put ours to use. My late arrival meant that we wouldn’t have time to clean out the airplane, or otherwise make it any more presentable. While at the show, I stopped by the Precision booth to ask if they’d ever heard of the idle mixture screw falling out. They said it can happen if the needle is backed out to a point where the spring is no longer applying good pressure to immobilize the needle. This was likely the case on our carb, because I had it backed out pretty far. The guidance about how to set the idle mixture says to look for a 50 RPM rise during shutdown. I incrementally backed the needle out looking for that rise, but never saw it. If you find that yours does not seem to respond to input, be wary of backing it out too far.

Using the GoPro on the Bearhawk

I’ve always been fascinated with the GoPro cameras. I had one of the early versions, and it was cool, but not very functional. It required AAA batteries, and could only handle a 2gb SD card. There was no preview to tell what the image frame looked like, and the on-camera buttons were the only way to control it.

Enter the later generation of cameras, as pictured here. I use the “white” edition to take all sorts of interesting pictures and videos. The version that I have is wide angle only, and having the option to go to a narrower angle would also be nice.

This little guy solves so many of the problems of the original. First, it uses a proprietary lithium battery. While I’m not usually keen on proprietary batteries, the lithium technology lasts so much longer than the aaa cells did, and it is fast to recharge. And now days with the folks in China not having any regard for intellectual property, it is easy to get a knock-off extra. I have ordered a three-pack and a two-bay USB-powered charger, though I haven’t had enough test time yet to be sure that they are worth recommending.

This new version uses better SD card technology that allows for much larger capacities. I use this one, but if you are shopping for one as of today I’d do a little bit of searching. The prices on these seem to fluctuate quite a bit from one packaging variant to another. I’ve never come close to filling up one of these. I suppose if I recorded a few hours of flying at one time then I might, but honestly the video would probably be pretty boring.

The real deal-sealer of features for me on the new generation of GoPro cameras is the wifi control. With a camera mounted on the exterior of the airplane, it’s nice to be able to start and stop the video (or take still shots) from the iPad or iPhone. Gone are the days of having to turn on the camera and record from start up to shut down (or more likely, battery death or card exhaustion).


The Bearhawk presents a mixed blessing of mounting options. On the wing, it’s very easy. I have a mount that is on the inspection panel where the front, outboard aileron pulley is. I applied the standard sticky mount to the inspection panel with the sticky adhesive, after making sure that I liked the angle and position. Then I removed the inspection plate and drilled two small holes through the black plastic and the inspection panel, and countersunk the holes in the plastic. I used two #6 stainless screws to double up the adhesive, since I’d rather not have the camera fall off in flight. Don’t get me wrong- the video would be spectacular, but I’d have to find the camera in order to get the video, and that seems unlikely to happen. As you can see in the picture below, this seems to be a very nice distance from the cabin, so that there is a little bit of background showing in front of and behind the airplane, but the fuselage mostly fills the frame. I point the camera fairly low so that the wing doesn’t dominate the top of the frame.

Cabin Mounting
Another easy situation is mounting the camera inside of the cabin. This time it’s easy because the risk of having the camera fall off are low. There is a little bit of concern for making sure that the camera doesn’t get in the way of something important or pose a risk of injury. I found that I really liked having it just aft of the aft carry-through tube, which in our case is also just aft of where the skylight ends. Here is what the frame looks like there:

Go pro in the cockpit

Go pro in the cockpit

Tail Mounting
This is where it starts to get much more difficult. Mounting the camera on the aft end of a Bearhawk is complicated because there’s not a good way to mount to the fabric. I’ve tried mounting on the aft handle like this with a handlebar mount:

Aircraft Go-Pro Mount

This has worked out fairly well, and here is what the pictures look like:

Tail View Wide Angle

Tail View Wide Angle

This view demonstrates how the wide angle view has its drawbacks. At first glance this may look like the horizontal stabilizer, but it’s actually the wing. With the wide angle view, the camera probably needs to be situated a little bit further away from the fuselage to make a meaningful image.

Here are some video links that I’ll add to as I get more edited and published.

37 Hours – Oil Cooler

Here’s an update from about 37 hours since the first flight. I found a few hours ago that I was getting high oil temperatures. This turned out to be an indication of several minor factors that were working together. First, the ambient temperatures started rising with the arriving spring season. Second, I didn’t really come up with an optimal mount for the oil cooler. I had it right next to the cylinder, but it’s really better for there to be some space between the cylinder and cooler. Also, I had the cooler mounted in such a way that it could trap air. A better arrangement would have been to have the inlet down low and the outlet at the highest point, which would naturally purge any air from the cooler. Another factor is that the 7-row cooler that I started with is probably a little bit too small. Bob seems to think that it should be adequate, but I read about many other installations where the 7-row proves to be too small for the angle-valve 360s.

To help resolve the problem, I’ve made several changes to the oil cooler arrangement. Note from the future- I’ve subsequently moved the oil cooler to the firewall. See this post. First, I found a 9-row unit. It is the same width as the 7-row, just a little bit taller. I removed the back left baffle and fabricated a new mounting system. The new mount scoots the oil cooler back about an inch, and also rotates it closer to horizontal. One thing that I haven’t yet done, but plan to do, is to flip the cooler over so that the outlets are on the top. A better arrangement would be to have the oil cooler sitting with the long dimension running vertically, but I’d have to make more extensive changes to mount it that way.

I have completed almost all of my data gathering. A few days ago, I flew down to Shelby to get some fuel. When I had topped off the left tank, I tried to install the fuel cap but found that the threads had seized. I had not lubricated the threads with anything, but Bob suggests a regular application of fuel lube to keep the threads from galling. I borrowed a vice from the friendly folks at the maintenance shop, and was able to use my onboard crescent wrench to apply enough force to snap the center male threaded part in half at the o-ring gap. This allowed me to separate the parts of the fuel cap. I borrowed the airport courtesy car and drove a few miles to the hardware store, where I picked up a 3″ long 6-32 screw and nuts. My plan was to drill the center vent hole and cut threads in the lower portion, allowing me to hold the cap together with the 6-32 bolt. The hardware store didn’t carry thread cutting taps, so instead, I drilled the vent hole up to the major diameter of the number 6 threads, and just used a nut on the bottom.

This arrangement worked to seal the tank, but not to allow for any venting. While flying on one tank is not an optimal arrangement (indeed, it is even advised against by Bob because of concerns about unporting the feeding tank during uncoordinated flight), it was sufficient to get me home for a 10-minute flight. The right tank was completely full, so unporting the inboard side was highly improbable.

Once I got back home, i was able to salvage the bottom piece of the cap, and of course all of the o-rings. I chucked the bottom piece of the cap into the drill press vice so that the vent hole was in line with the drill chuck, then I started increasing the size of the vent hole. I used about four sizes to get up to the minor diameter of the 1/2″ threads, which left just a few small pieces of the old male threads in place. I was able to pick out most of those pieces with a scribe, at least enough of them to allow me to get a thread cutting tap started. I worked it back and forth slowly and the threads cleaned up just fine. I used a steel bolt from the hardware store with coarse 1/2-13 threads to make a temporary center post. I drilled a 1/8″ hole through the middle of the bolt, which will allow the cap to vent. Since the bolt has a hex head instead of the airfoil-shaped vertical blade, cap removal requires a wrench. This gets me back to a flying condition while I wait for a replacement cap to arrive. You can be sure that I have since added some lube to the threads on the other cap too! I’ll use the repaired old cap as a spare, which I’ll carry in my onboard tool kit.

Posted on
Hours Logged This Session:
Total Hours: 1883.75

Mobile Phone-Controlled Preheat

Now that I’m flying with some regularity, I’ve found all sorts of inconveniences that are wanting solutions.  One example is the engine pre-heat.  This is a little bit of a princess complaint, since the airport is only a mile from my house.  Still, I don’t want to have to get in the car, drive to the airport, open the gate, drive down the bumpy road, and all of that, just to plug in the heater and come back home.  Many of my data-collection flight tests require smooth air, so I’ve been making some very early morning flights.  The reality is that I end up rushing the preheat, which isn’t doing the engine any good.

The low-hanging fruit that I could pick to solve this problem is The Switchbox.  This fellow sells them ready to use, with short extension cords in place and a decorative logo.  He advertises on many of the type-specific message boards, which tells me that he knows how to find customers.  The only downside of his unit is the price.  As you may know if you have read much of my building log, I’m quite frequently able to spend way too much of my life trying to not spend a dollar.  This happens even when I make a conscious effort not to! A little bit of Googling pointed me to another option for a cell phone-based switch, and an excellent opportunity to see if I could save money by doing it myself.

I’m willing to report that I do think I saved money, but I’m not ready to be sure that someone else in the same position would agree.  The short answer, in case you don’t want to read about all of the details, is that I believe that I’ve created a functionally-similar unit for about half the price, with the investment of about one hour of shop time and half a day of research.  With any luck, I’ll be able to save you most of that research, depending on how long it takes you to read all of this, and depending on whether you fall asleep in the middle of reading it.  For someone who is not comfortable stripping wires and going to the t-Mobile store, the ready-to-use Switchbox will probably be a superior value.  There may also be some functionality that the Switchbox has that my unit does not have, so I’ll stop short of saying that mine is exactly the same- not because I have reason to believe it isn’t, but rather just because I don’t know for sure that it is.

So here are the details, about how I did it, and about how you could too.  First, I ordered the circuit box from China.  In the days of the internet and the global economy, I find myself ordering things from China with some regularity.  The company that markets the box is called “Wafer Star” and they call it the GSM-AUTO-AC.  GSM is of course the name of the type of cell phone network (don’t get the box for CDMA if you are going to use the tMobile service).  The AC at the end is also very important, because it is designed to be powered by AC.  They also have a unit that is powered by DC, and lacks an on-board transformer to power the circuitry.  You don’t want that one for this application.

The best deal that I could find was on eBay, though you can also find similar products on Amazon.  If you go that route, be sure that you are getting the AC version and the GSM version.  The total was $140 with shipping.  I was ordering right in the midst of the Chinese New Year, and the eBay arrival date estimate was around 3-4 weeks.  I was willing to wait this long, but I was already starting to see how The Switchbox was able to score some value.  A few days later, my wife mentioned that when she checked the mail, she had a box from China that our cat was very interested in.  I didn’t remember ordering anything else, but figured there was no way that the GSM-AUTO would have arrived so quickly.  When I got back home, I was glad to see that it wasn’t a kilo of high-end Cantonese Catnip, but rather a very-quickly shipped package that promised to meet all of my preheating needs.

The unit, quite cleverly, ships without any electrical cords.  This probably poses at least two big benefits to the manufacturer.  One, he doesn’t have to try and supply area-specific plugs.  Two, he knows that the person hooking the box up is going to at least need to know which end of a screwdriver to use.  I went to the hardware store and purchased an 80-foot extension cord for $16.  (Total cost now $156).  This cord came with a male and female end, as they usually do.  I could have made use of just those ends, but the GSM-AUTO has two different relays.  I figured I might as well make both of those relays available, so I also purchased a female plug. (now $160)

I measured the distance from my electrical outlet to the area of the hangar over the engine compartment.  I applied an extra yard or so of error margin, then cut the same length off of the male end of the extension cord.  Then, I measured the distance from that ceiling location down to where I would want to plug in the heater.  I cut that same distance off of the female end of the cord.  Then, figuring that I wouldn’t need to have the second relay attached to something very far away from the engine, I cut the remaining bit of cord to match the female end and applied the female plug to one end.

Next, I stopped by the local tMobile store.  Keep in mind that the store-bought Switchbox comes with a SIM card already in place.  I gather that the other store-bought option (from Reif) does not.  Reif suggests calling tMobile to have them send a card, but I was fortunate enough to have a local store at my disposal.  The card costs $10 to purchase, and I also purchased the minimum $10 of prepaid credit.  The GSM-AUTO can operate in a few different modes.  One mode turns on the relays when you call the unit.  This method doesn’t consume any of the prepaid credit, since the unit never answers the call.  The other method is to operate by text, which does consume the credit.  I figured that since the credit is only good for 90 days, I might as well operate in text mode and use it before it expires.  The total cost was now up to $182 with tax.

Back at home, I called T-Mobile to have them disable the voice mail option.  One thing that I didn’t do, but probably should have, was to pop that SIM card into a phone, and register/activate an account online with t-Mobile.  If you are in a position to do this, now is a convenient place to do so.  The online account should in theory allow for automatic renewal of the expiring balance, among other things.  If you don’t have an unlocked or t-Mobile phone to do this with, you could probably have the folks at the store help.  Once the sim card is in a phone, go to the website to register a new account.  That site will ask for your mobile number, then it will send a secret code to your phone by text.  Enter that code, and it will know that you are authorized to create the account.  As of my writing this post, I haven’t completed these steps, because with the box installed in the hangar ceiling, I have to move the airplane out of the way and climb up there to pull the card back out.

With that step done (or disregarded), I put the card into the card holder on the circuit board.  Next, I connected the incoming AC to the appropriate threaded terminals on the board, in my case, the far left two.  Don’t forget to thread the incoming wire through the little compression nut first, then the box cover second.  Once you have the wire attached to the terminals, this won’t be very easy.  From that point, I wanted to test the operation of the unit before I went any further.

Upon powering it up, I used the instructions on the Wafer Star website to start communicating with it.  One communicates with the box by sending texts.  It came with a serial port option for programming by computer, but I haven’t had a serial port available for many years.  I didn’t try using the Dynon serial converter that I use for communicating with the 100-series boxes, because I didn’t have it on hand.  I sent a command to the box to have it change the password first.  Then I had it establish that my number was the only one on the white list, so that an errant wrong number wouldn’t make any relay changes.  The Wafer Star instructions are pretty good about learning how to make these commands.  From the iPhone, I used the copy and paste functions to save from having to type each command repeatedly.  I did not find that their iPhone app worked, which might be related to the iOS7 upgrade issues.

Next I sent a command to have it turn on relay 1.  A second or so later, I heard a click.  Another second later, I got a text back that confirmed that the relay was on.  How cool!  I unplugged the incoming power, and took a cookie break.  I’m always a little bit nervous about working on AC-powered devices that have been recently powered.  I don’t know what kind of capacitors there may be in there waiting to shock me, or worse, some poor component on the circuit board that I wouldn’t be able to fix.  After all, I an heal, but as of yet, electronics can’t.

Next came the hardest part of the whole venture, which really wasn’t that hard- connecting the wires.  After I ran the two bare ends of the female wires through the little nuts, then the top cover, I stripped them to expose about 2″ of inner black and white wires.  My extension cord had a ground conductor, which I cut off as close as I could to the end of the outer shell as to limit the possibility of shorts.  Then I cut two pieces of the scrap wire to about 4″ each, and removed the conductors from the outer green shell.  That  yielded 8 ends that needed to be stripped.  I happened to have a nice stripper that made quick work of those ends- she had them ready to go before the first song finished and I ran out of singles.

Now is the time for concentration.  Starting with the left-most terminal, put the black incoming power conductor and two of the black 4″ pieces under the first screw.  Under the second screw, put the incoming power neutral (white) with a 4″ piece of white scrap.  The way these relays work is that they are only switches.  The terminals are what you use to interrupt the power on its way to the heater.  So that means the outgoing white wires will join together with the 4″ stub that just went under the second screw on the left.  Another option would be to eliminate the 4″ stub of white wire and just put all three white wires under the left-most screw.

Now, run one of the 4″ black wires to the “COM2” terminal and run the other one to the “COM1” terminal.  If you want for your relays to turn on when you send a text (as opposed to having them turn off when you send a text), disregard the “NC” terminals.  You would use those if the default/fail-safe position was going to be on, or “closed” as the relay would say, if relays could talk.  In the case of a preheater, I definitely want the default position to be “off.”  So, if you are following along, run the two outgoing black conductors to the “NO2” and “NO1” terminals.  You can follow the electrons in your mind- you want for them to start at the wall outlet, then swim along to the left-most screw, then from there to the “COM2” terminal, then if the relay is on, to the “NO2” terminal, and to the heater.  If you object to my oversimplified metaphor for swimming electrons, then you probably know enough to hook the box up without needing to read my article, so go be an electrical engineer someplace else.

After making those connections, I did one final test before closing up the box.  Everything was good, so I attached it to the wooden trusses in the roof structure.  I suspended the incoming power and routed it to an outlet, and left the female outgoing plugs dangling.  I tossed a string over a part of the door hinge on the side of the hangar, and tied the other end to the female plugs.  I use that string to raise and lower the outlet cords.  This keeps them up and off of the floor, and also out of the way of the airplane as it comes and goes from the hangar.

So in all, I would say that I saved on the order of $100 by going with the direct-from-China option.  If you figure that I would have had to still buy extension cords to go with the store-bought Switchbox, then the cost is right around half.  If you are in the market for a similar setup and are willing to do a little work, you might consider going the route that I went.  If the above description for how to connect the wires is at all confusing, you’ll probably be better off going with the pre-made, plug and play, hassle-free option.  Keep in mind too that the Reif option is also available.  While my box appears to be identical on the outside to the Switchbox, it is different from the box that Reif shows on his website.  His price is only slightly more than what I’ve spent, unless you happen to be buying one of his heaters at the same time, in which case his price is a no-brainer.

Gross Weight Testing

Today I continued adding ballast to the Bearhawk to test how it performs at heavier weights.  My ballast is 80 pound bags of concrete, so I started with 160 pounds in the back seat.  The weight wasn’t really noticable during a few stalls and some phugoid tests.  Since I was flying with close to half tanks on that flight, this weight gain was not much above my solo full fuel weight.  On the next hop I added 160 pounds to the front right seat.  As with before, this weight didn’t seem to matter very much. 

For the next flight I added two more bags to the back seat.  By now, the pitch stability was still pretty good, though the phugoid definitely started to show a higher amplitude.  The takeoff roll was noticably longer (perhaps 700 feet or so) and the uphill back-taxi required much more engine power.  The latter indicator might have also been related to the somewhat soggy runway.  For the first time, I was able to land tailwheel-first.  The landing flare was much easier.

For my final configuration, I added 105 pounds to the baggage area and another 80 pounds to the front right seat.  To recap, that’s 240 pounds in the front right seat, 320 pounds of concrete (plus another 20 pounds of fly-away-kit), plus 105 in the baggage area.  I was at about 1/3 fuel, which would mean that when I flew up to HKY to top up the tanks, that would put me about 8 pounds short of the max gross 2500.  I should mention that with each additional weight increase, I also measured the main landing gear tread.  As the weight increases, the gear spreads out.  Bob advises that the center-to-center distance of the main landing gear tires must not exceed 72″ at max gross, an fortunately mine did not.  

That heavy configuration had my CG at 21.88 inches by my calculations.  Bob specifies a limit of 22.5 inches, but I didn’t really care for flying it at 21.88.  If I encountered a little updraft, I found myself working the elevator control back and forth to maintain level flight.  If I trimmed for level flight at around 80 knots, bumped the stick aft about 2″, then let go, the resulting pitch up was 1.6g all the way up to around 25 degrees.  While I thought all of this through over lunch at HKY, I decided not to test any further aft.  Instead I made a note that 21.5 will be my operational limit.  To get there for the next flight, I moved the 80 pound bag from the baggage area to the back seat, and left the 25 pounds of kitty litter in the back.  Then I topped off the tanks and took off.  As before, the most noticable difference was the longer takeoff roll and climb rate.  

I had intended to fly around the pattern a few times, but all of this extra weight made it apparent that my tailwheel tire was a little bit low on air.  On the second landing (which was not one to be especially proud of anyway), the tailwheel started to shimmy as I slowed through about 25 knots.  I attributed the shimmy to the low tire pressure and departed for home.  On the way I tried to validate my calculated high-altitude performance with a climb to 10,000 feet.  I had to level off several times along the way due to high CHT and/or oil temperature.  This tells me that I need to improve the effectiveness of my cooling, since the OAT was right around freezing.  I have made a few adjustments and am going to try again on another day.  

In summary, I would say that the fully-loaded Bearhawk, particularly at a fully-aft CG, is a whole different animal than an almost-empty one.  This is the case with every other airplane that I’ve flown in both conditions, so I’m not surprised.  The good news is that from the family-hauling standpoint, I won’t be that heavy or that far aft.  My concrete family is at least 300 pounds heavier than my regular family, at least as of now.

Russ Erb Visit

I had a long overnight in LAX today, so I checked in with Russ Erb.  His schedule was available for me to stop by for a visit, so I rented a car and drove up to the desert.  I was excited about getting to catch up with Russ and discuss a few flight testing strategies, and I was also eager to see his plane up close again.  I’ve seen it several times over the past few years, but I’ve noticed that each time I see a completed Bearhawk, I have different things to look at.  When I was doing fabric work, I was looking at fabric work on other Bearhawks.  Now that I’m seeing how mine is working out in service, I look for similar information on others. 

Just yesterday I had a long overnight in IAD, and I spent the afternoon at the Air and Space Museum.  Similarly, I can go to a place like that and see the same airplanes that I saw there 9 years ago, yet my life experience has changed enough to give me something totally new to look at.

While I was exited about visiting with Russ, I was also very excited about his gracious offer to take me flying in his Bearhawk.  Upon arrival in Rosemond, Russ drove us out to where his Bearhawk lives.  It’s a very nice airpark with a paved runway and everything.  I picked his brain with all sorts of questions about how he has best found to operate his Bearhawk in the past 300+ hours spread over 4 years while he prepared a few last-minute things.  We got in and he started the engine for a brief warm up.  

There are several unmistakeable differences in Russ’s cockpit and mine.  First, his is much more festively decorated in yellow and red.  Second, he has quite a few more gadgets at his disposal.  He commented about this, being aware of the very full pilot toolbox.  This was my first experience in a 540-powered Bearhawk, so I wasn’t quite sure of what to expect from a performance standpoint.  The takeoff roll was comparable to what I’ve experienced with my Bearhawk, but there are several things to consider in that comparison.  First, we were operating at a heavier weight than I have been operating in my solo flight tests.  I suspect that we were around 400 pounds heavier than my usual weight, because of the extra passenger weight and higher empty weight.  Second, and perhaps most importantly, we were at a much higher density altitude.  The airport elevation is much higher than I’m used to, and so was the temperature.  Those factors combine to explain why the performance seemed only somewhat better.  I’m interested in doing more comparison work someday where I can better quantify the difference, using tools other than intuition and the seat of my pants.

Seeing all of these factors in play, it makes much more sense to me why Russ chose to equip his Bearhawk as he did.  After all, it was January!  In the summer, I can imagine that he sees much higher density altitudes.  The departure density altitude is an issue, but so is the need to climb high to clear the surrounding terrain.  In the short hop over to his glider airport, I saw why he explained that any cross-country trip of length is going to require a cruise altitude in the 10,000 foot realm.  I’m especially interested to get our Bearhawk into this kind of environment to see how it does, but I haven’t yet.

I had a great time riding with Russ, and I learned a bunch about how he operates his Bearhawk.  To top it all off, his wife made a home-made supper that couldn’t be beat, and I got to visit with one of his fellow TPS friends and hear about their Oshkosh trip.  That area is crawling with a bunch of really smart airplane folks, and I could certainly imagine an alternate life where I could live around them and have lots of airplane fun.  Overall the trip was absolutely worth the effort, and I’m looking forward to applying the experience and some measurements from Russ to make a much more numerical comparison between the capabilities of the 540 and 360 Bearhawks.

Flying Update

I have found that the engine is running a little less smooth than I think it should. This is a challenging problem, in part because it may be running just fine. I have a fair amount of experience in single-engine piston airplanes, but most of it is not in recent times. At idle, I hear an occasional miss, even with the mixture leaned. The other day, I took off with the intent to fly for a few hours, but at 2500 RPM and full throttle, it was running rough enough to make me turn back towards the airport. To be fair, I’m not sure if a non-pilot passenger would have noticed it. I leaned the mixture a little bit and the roughness went away, but this didn’t make sense, because it was exceptionally cold that day, and the density altitude was probably well below sea level. On another day I found 150 RPM drop during a ground mag check, when I turned off the left mag. I tried running the engine up a little bit and leaning aggressively, but had no change. I taxied back on the rough mag and hopped out to check the exhaust pipe temperatures. All except number 1 were around 350 degrees, while number 1 was more like 300. I swapped the bottom plug from that cylinder to the top, and tried another mag check. This time the drop was on the other mag, confirming that the plug was to blame. I replaced it with a spare that I had on hand and took off to conduct an inflight mag check, as described by Mike Busch.

His process involves using the EMS data logging, so I figured getting that started would be a good first step. I had noticed earlier that when I went to the data logging menu in the D180, instead of the usual choices of “on” and “off”, I had a message that said “Error 4.” With me not having the tools to address that problem right then, I took off anyway and conducted the check as Mike describes. I wrote down the EGT changes, but as he points out in his articles on the subject, that’s not really enough information. The good news is that during those checks, I didn’t get any significant roughness. The bad news is that I had some wacky EGT numbers. The cylinders were all fairly even when running ROP, but one was about 300 degrees lower when running LOP. I resolved to get the data logging working, then fly on another day to get more valuable numbers and see what’s going on. I never have installed the little blocking plates in front of the front two cylinders, since the Vans baffle kit instructions say to not install them unless you find that you need them. I have been reading reports from Husky owners that they get much better LOP operations with similar plates in place, because the cylinders run more similarly to each other. So that’s another avenue to explore.

To address the “Error 4” in the data logging menu, I did some Googling to find that it is a bug with the Dynon. Speculation is that the data log file becomes corrupted. The fix is for Dynon tech support to send a file, which I then send to the D180 as a “restore firmware backup.” The repair file is firmware version specific, and Mike at Tech support sent the version that corresponds with the latest firmware. I was due for an update anyway (though I didn’t realize it just then) so I downloaded the latest support program and went back out to the airport. I went through the usual ritual of powering off both the D180 and D100, then updating the HS34. Then I updated the D180, then the D100. After all of the firmware updates, the Error 4 message was still there, so I sent the repair file. The good news is that the new file fixed the problem, but the bad news is that there wasn’t any log file there. The D100 did have a good file though, and it has been extremely fascinating to study. Now I’m looking forward to getting to fly again and get some good data from the D180, which will be much more valuable, since it should have EMS data too.

In fact, with having so much information available in the logs, I think I’ll plan to redo the bootastrap glide tests, to try to provide better numbers for the data plate.

The Bootstrap Method of Performance Analysis

The first part of flight testing was relatively easy to plan.  For the first few hours, I was just interested in getting the airplane safely into the air and back on the ground.  I began to find the edges of the center part of the envelope in as incremental of a way as possible.  After those few hours, what was to come next?

I’ve read several articles, books, and Advisory Circulars that all have good guidance, but what I really wanted was a good way to start fleshing out the performance data that would come in the POH of a store-bought airplane.  I have read about sawtooth climbs and other “fly and record” approaches, but I didn’t see how that was going to give me as much detail as I knew I would want to have available.

Fortunately, some searching led me to a series of three articles by John T. Lowry in the Avweb archives.  Those articles led me to other articles of his, and to his book, Performance of Light Aircraft.  I’ve read enough of his writing to appreciate his approach and his style.  He is a physicist, but he is also a pilot.  He finds an excellent balance between the theoretical world and the actual world, and takes an approach to performance calculations that knows its limits so to speak.

He describes a method that he calls The Bootstrap Method of Performance Analysis.  The short version of the strategy is that we use the measured performance from one set of conditions, combined with his math-rich spreadsheets, to create predicted performance for many different conditions.  I’ll let you read the articles to learn how the strategy works.  If you’d like a little bit more detail, I’m planning to put an updated article in the first quarter 2014 Beartracks issue.  If you’d like to have much more detail, see if you can find a copy of his book, in which he goes step-by-step through his derivation of the various formulas.  Good luck finding a copy, since it has aparently been out of print for a while unfortunately.

The good news about his method is that it doesn’t take very much flight testing to get started.  The even better news about his method is that airplanes with constant-speed props have even fewer measurements to make.  The airborne flight tests amount to just a series of glides.  These glides do require a calm day with no convection, which means either an early morning, or an overcast, stable day.  

The first round of glide tests have shown that the minimum sink speed is somewhere between 50 and 55 knots indicated.  Lowry points out that the curves in most GA airplanes are very smoothly shaped, which is another way of saying that the glide performance difference between 50 and 55 is likely to vary little.  I’m planning to do more tests to see if I can pin down the specific speed, but for now my sample size is too small to say definitively whether it’s 52, 53 knots, etc.  

I’ll have much more to write about the method, especially if it works well.  The great thing about it gives me some insight about how various factors impact performance.  I’m less interested in knowing exactly what my climb rate will be at a particular circumstance.  I’m much more interested in knowing if, for example, the temperature were 20 degrees warmer, how much I should expect the rate to change.  Will it be 10% lower, or perhaps 80% lower?  Short of intuition or actual flight testing, I don’t have any other way to know.  

The Bootstrap Method also allows for comparing input factors like engine horsepower.  What if my engine produced 15 more horsepower at the same weight?  How would that impact my maximum level speed?  Similarly, what if I had installed a Hartzell 80″ prop instead of a 76″?  You can see how this method, even though it does make a few underlying assuptions, will be very useful in helping elaborate on how the Bearhawk can perform under certain circumstances.  Look for much more to follow, if not here, then perhaps elsewhere.

Transponder Check

Yesterday’s flights went very well, and after a second thorough inspection, the airplane still looked like it was ready to fly. I took off and climbed up above the airport again, and all was still well, so I flew over to the avionics shop to get the transponder checked. This will allow me to turn the transponder on, which will greatly enhance safety with ATC and other airplanes. The folks at the avionics shop fixed it right up, and I flew back home. Instead of landing back at Hickory, I landed at our home airport just south of town. I’m glad to have the airplane back in our own hangar, where I can work in more relative comfort and convenience.

First Flight

Last week I drove down to Charlotte to pick up the airworthiness certificate. I got back into town with it just as a cold front did, and it took these 9 days to get a day with the right combination of me being in town and the weather being good. The weather was perfect all day today, and the flight went well. I flew for about half an hour, then came back to do a thorough inspection. I found a few more minor items to correct, such as the direction of the elevator trim. It worked backwards to how I thought it should work, so I just flipped the wheel around to reverse it. In that process, I noticed that the nut in that bolt had no cotter pin, which goes to show that even with several sets of eyes looking over an airplane, things can get overlooked! Everything looked good with the inspection, so I put it all back together and flew again. On the second flight I repeated much of the first, though I expanded the initial stall testing by adding flaps. At the end of the flight, instead of ending after one landing, I flew around the pattern for 5 more. The challenge with test flying a new airplane is that there are several goals to achieve. One is to validate the airframe and systems, but another is to train the pilot. The landings were more for the latter goal, though of course they also helped with the first. I feel very confident with the directional control on the ground, which was one of my main concerns about the flight testing. I find that I have a little bit of a heavy wing, and a few other minor things to fix, but that’s what the next 40 hours are for.