By Paul Walker
December 2025

In the early 1990’s, the concept of a B-17 stunt plane took shape. The first model was powered by four OS .15 FPs with four separate tanks. A few years later the second B-17 was built using the same four engines, but with a pressure fuel system. It was complex, and unreliable. It did fly at the 2000 World Championships and finished a reasonable ninth place. It was used a year later at the 2001 US Nationals. It was problematic, and no longer fun to fly. That was the last flight on it for decades!

Then, a few years ago I got the bug to fly a B-17 again. But this time it would be electric powered.

I started by testing concepts in the first B-17. The plan was to verify that the four electric motors selected would be enough power. Motor mounts that would attach to the existing internal combustion engine mounts were constructed. It was “estimated” what size motor was necessary along with ESCs (electronic speed controllers) and timers. Individual batteries were used in each nacelle  and were not large enough for a full pattern. The flight tests from this showed that the motors were sufficient. It flew very much like the IC version.

Based on this success  the B-17 that was flown in Landres, France, in 2000 was converted to electric. This time the battery (singular), ESCs (four) and timer (singular) were mounted in the fuselage. To complicate this installation, it was a take-apart plane where the wing was in two pieces and the fuselage in three pieces. This led to MANY connectors in the power system, which led to added weight. It was flown with one of the "new" Igor timers. It flew well, however it never completed a full pattern.

When the original two B-17s were flown, they required larger muscles to fly them. Lifting weights regularly since I was 25 did help, however I still had to add more new lifts to build the muscles necessary to fly it more than once! The extra muscles helped, but the line tension only allowed three flights in a day, even being in decent shape and younger. Now, 25 years have passed, and my body can no longer take flying those older B-17s. Still wanting to fly a B-17, the only choice was to develop an updated version that was smaller and lighter.

It was decided that the weight should be around 70 ounces. Wing area was set at 700 square inches and the "exact" taper in the wing was decided by just how much the tip chord could be enlarged without looking wrong. The fuselage side view has a scale outline as well as the horizontal tail location. Horizontal tail vertical location was raised a bit so that the elevator horn would not be outside the fuselage. For weight reasons, various versions of the B-17 were considered. The C/D model was enticing as it would provide a lighter empennage, which is important considering how far back it is. In the end, an F model was selected for control system clearances.

At this point, the general layout was drawn up, and Excel was used to calculate weight and balance. It was clear from the start that the center of gravity was going to be the biggest issue to deal with. Motor choice was another critical issue. Any extra weight there is multiplied by four. Again, a few calculations were made and a decision was made on a motor, the Badass 2305-1050. But were they enough? That question had to be answered. Having many old Impacts around, one became a donor. Quickly it was converted to a four-engine test bed, and flight testing began. It was able to fly level laps fast enough, however there was zero added power to maneuver. A call to Badass and four new 2310-1220's arrived. This proved to be the difference. However, they were 0.60 ounces (17 g) heavier. This meant 2.4 ounces (68 g) more than expected. These tests did prove that this setup would function properly.

It was now time to finish the design work. The most difficult part of the design was finding a way to install four ESCs, the wiring, the battery into the fuselage that was accessible for setting the timer and ESCs. A one-piece fuselage with removable wing panels was seriously considered. Eventually that concept was dropped due to electronics installation and adjustment issues. The final solution was a non-removable wing configuration with a removable nose section, which houses the battery. This also provides access to install and adjust the electronics.

All proceeded well. The fuselage was molded with 1/16-inch (1.58 mm) sheet from a male mold. Carbon mat was attached to the inside of the shell. It was then placed back in the mold to have the carbon mat adhered to and conform to the shape exactly. The rest of the construction was straightforward for a balsa model. An assembly jig was built to locate all the fuselage bulkheads before the shell was added. There were no serious issues with the construction. Added jigs were used to locate the wing, the fuselage, and the horizontal tail. It was a joyful day when the jigs were removed and the final process of preparing for painting finally started.

Once construction was finished, a quality finish was required, but again weight was extremely critical. That process went well, and the finish weight came in under the estimate. It was a standard Dope finish, with many hours of sanding between every coat. Then there were the ink lines. Lots of ink lines. Doing ink lines around a curved, in two directions, fuselage was not fun. I obtained a thin flexible straight edge and set up a vertical line on both sides of the fuselage. I then taped the straight edge to one side and applied the ink. Then moving to the other side of the fuselage the same process was done. The horizontal connection on the top or bottom connected those two lines maintaining a straight looking line from the side. This took many weeks of time to complete. The circumferential lines on the engine nacelles were easy as they are simple cylinders. When inking was completed, the oil and exhaust stains were added, then the clear coats went on. Just enough went on to look acceptable, as the weight was already higher than I anticipated. Then the sanding and rubbing to a nice shine. Then the construction was done.

As the first flight neared, the concern for the electronics increased. Igor sold Jeti ESCs that I used in the weight estimates. They were to be Spin 22s. The problem came when he informed me that Jeti no longer produced them. Ouch. The lightest Spin ESC available was a Spin 44. Problem is that they were bigger (space issue) and each was about one ounce (28.3 g) heavier, times four. So now, it is 6.4 ounces (181 g) heavier than planned. My temporary solution was to use two Spin 22s and two Spin 44s. Sounded simple: two motors would stay constant and two would speed up and slow down.

The first flight went well. It was to set up the level laps speed. The settings were "guessed" at, well maybe better than a guess after 19 years of running electric power, which produced the wanted 5.3 seconds a lap. The battery capacity was also a "guess” and the first flights were all two-minute runs. After several two-minute runs, it appeared the battery was large enough. It was a 4S4200 made from Molicel 21700 4200 cells.

Maneuver time came and immediately there was an issue. Two motors held their rpm, and two didn't. Several flights were tried, and nothing solved the problem. Back to Igor to get two more Spin 44s. With four Spin 44s running, it was pure harmony. It was hard to tell that there were four motors running. It sounded like just one!

Now, it was finally time for a full pattern. My friend Tim Just happened to be at my field on that day. He was totally surprised at how it flew. After all, how well should a large four-engine plane really fly? It does fly quite well. I look forward to the 2026 US Nationals in July as it is scheduled to fly there. When asked, which happens often, how does it compare to the IC, the response is that it simply doesn't compare. By any measure it is clearly superior. It will take a little time to adjust to it, but it will be fun! There have been a few more flights on it, and with each additional trim adjustment, it keeps getting better. It WILL be fun. A plane at that weight simply should not fly as well as it does.

Was it worth all the time and effort? It absolutely was. It is fun to fly and looks great in the air. After all, how many four-engined F2B planes are out there competing?
 
 Here are the “final” specifications:

• Wingspan is 67 inches (1701mm)
• Wing area is 700  in^2 (45 dm)
• Moment Arms: The HT hinge line to flap hinge line is 20 inches (508 mm). However, the flap hinge line sweeps forward so it is effectively longer.
• The finish is Dope from Randolph.
• Motors BadAss 2310-1220
• Battery: first flight was on a 4S4200 but now have multiple 4S5000. These are made from 21700 Molicel 21700 5000 batteries.
• ESC Jeti Spin 44s (Igor configuration)
• Timer New Igor timer, version 1.0
• Propellers APC 8-6 electric in black
• Line Length. 70 feet center to center (21.3 m) .018 inch (.457 mm) 7-strand braided wire.
• Launch RPM, currently 10,010 rpm.
• Lap Times 5.3 sec/lap
• Weight (including battery) is approximately 75 ounces (2,126 g)
• Handle is a Walker 18-gram carbon hard point.

Photo gallery

This is the first part made of the plane. It was one of the critical weight items and I wanted to be sure it would be within the weight estimate. And it was!

The VT leading edge being added to the remainder of the Vertical Tail. The assembly came in on the estimate.

The wing ribs and spars in the jig. The inboard leading edge sheeting is being fit-tested at this point. The wing jig are four steel rods.

This is the basic wing construction finished. The electronics bay is obvious. Inside the wing are four aluminum tubes routing the power leads to each motor.

The BadAss 2310-1220 mounted to the motor bulkhead. This is one of four. Note that the motor mounts from the rear of the bulkhead.

These are the fuselage bulkheads ready to load in the fuselage assembly jig.

The completed wing is now in the jig inverted. The fuselage jig is the white section below the wing. It is the length of the fuselage, and has clamps to hold the bulkheads in place. A laser level is used to verify the vertical and horizontal positioning of each bulkhead.

The finished wood work.

This view shows the two tube routes on the right side for the wire runs out to the two motors on the inboard side. There are two more that are not visible on the left side of the outboard wire runs. There are added balsa pads on the end of each tube so there will not be an electrical wire touching the edge of an aluminum tube.

Well, how will all of this go in here? Note that there are two 44s and two 22s at this point. Each ESC is mounted to a tray that slides into slots in the opening. Then all the additional wiring has to go in, like the connections from the timer to the ESCs,the main power from the battery to four ESCs, the arming plug and the on-off switch.

Every now and then when the sun was out, I couldn't resist taking a few pictures. At this point I was trying to decide if it needed any more weathering.

This is another shot after the buffing was complete. Lots of hours spent getting around all four nacelles, etc.

This is just before the first flight. A bit tense, but ready for whatever comes.

---

Back to Aerobatics section

Back to Walker on Stunt index page