Scale Wendell Hostetler 30% Piper Cherokee Glass Fuselage Build.

Over on the radio forum on another site came across a discussion using OpenTx with an electronic carburetor used on Stihl saws. Followed this back to originating forum and one thing lead to another while I was on vacation so I have the carb in hand now. It is a Zama carb with 17 mm bore. One fellow has it on a 26 cc engine, an another has it on a 32 cc engine, both two stroke. It is used on 50 cc saws



Fuel flow is controlled by a Seeed Studio XIAO board using Arduino code. The code was written by a fellow that goes by the name of raleighcopter on several forums and GitHub. Others on the site are folks named 1967brutus, Cat1, and Glowgeek, there may others an I've only read though 400 of the 2500 page thread. The project is mainly focused at using the solenoid from the Zama carb in a custom housing to control the fuel flow on gas convert glow engines. It so happens to work with the Zama carb as well. There is also temperature and barometric pressure in the code. A BMP 280 is used to measure the temperature and air pressure near the carburator. The devices are rather small as seen below.



I have limited experience with Arduino so I need to relearn everytime I use it. I have the code downloaded and written on three XIAO boards, that took the better part of half to set the sketch up and .27 seconds to write it. Need to some soldering and get this wired up. As far as I know this is the first conversion done on a four stroke engine, the Zama doesn't come with a spring on the fuel pump membrane so I need to work out a solution for this. The Zama has a rather thin plastic like membrane while the original carburator on the Saito engine looks to be reinforced elastomer.

So cool! I keep seeing EFI with DA and GP starting up at commercial drone industry prices. There have been a few others trying to produce a product and some making kits like Ecotrons, INF Inject, Loweheiser. Pretty neat that this will one day be in our model airplanes. I think the midrange is what's been lacking on EFI for aerobatic planes on what I've heard.

There are no needles, the adjustment is done using a curve, in OpenTx this can be up to 17 points. I need to work out the pump however. The original carb had a spring in the pump. The Zama does not and the pump membrane is thin plastic sheet while the original looks like reinforced elastomer. I found a similar kit and have it on the way hoping it is different material.

Oh man, this technology sounds great.

I have the controller built for the electronic carb. It seems to be working, need to get the carb mounted and do some test runs. There are four telemetry points, haven't figured out how to decipher the numbers yet. Can hear the carb pulsing proportional to the throttle stick position.

You fancy.

A little update here. The code for the controller was written for using the BMP280 baro/temp sensor that communicates with I2C protocol . I have no idea what I2C is but when the sensor would fail it would lock the controller up. I reached out to the author of the code and his comment was it was the library causing the problem. I purchased several different sensors and with their respective libraries failed in the same manner. I decided to try analog sensors much like found in our current automobiles. I've been studying and rewriting code for the past few weeks, it takes a lot longer when you have no idea what you are doing! With the original code, Google AI, YouTube and the like I have the code rewritten with the analog sensors. Also included a moving average to eliminate the noise spikes in the data stream. I have failed the sensors in each possible broken wire senario and it works as expected. Today I decided to add a rolling average to the final solenoid output to further smooth the output although I'm not sure if it is necessary but why not? Today when I went to upload the new code, I guessing my cord failed and lost communication with any device plugged into the cord, I tried three different devices and they all failed to communicate with my PC. I have new cord coming tomorrow Hopefully that is the problem.
Photos of the analog sensors are shown below, the barometric sensor is the rectangular device and the temperature sensor looks like a transistor. Both are mounted on opposites sides of the same board.
Started sanding the flaps and ailerons today, would like to get the wings painted before winter, with the scale contests and practice flying along with working on the code I little interest in sanding! Plus I sent about a month doing rework on the Champ for nought.
Temperature is dropping here and building season is on the horizon. I have the fun flyes to attend this month and that's it.

This coding thing catches on, thinking about making RPM/Voltage/Temperature hub for the plane as well. Thinking 1 RPM, 1 Voltage and 4+ temp sensors on the same board used for the controller. With enough temp sensors one can better evaluate the effectiveness of the baffling.

The summer heat here is waning, flying season about to restart………. Interesting work going on here.