The DIY YAESU Rotator Controller (G1000DXC version) has withstood the test during the recently concluded 2025 CQWW SSB contest.One aspect is the hardware test.This version uses a 12V-5V DC-DC converter, and the power supply for the Arduino board is also 12V, relying solely on natural cooling without forced cooling. It has been running continuously for at least 60 hours without issues, indicating that this architecture is viable.(There are various opinions on supplying 12V to the Arduino board; some say it’s fine, while others question the reliability of the voltage regulator components on domestic boards, suggesting it’s better to reduce it to 9V. Therefore, I used an external 5V on another G1000DXA version to see what effect it would have.)The other aspect is the software (function) test.The rotator controller works well with n1mm, responding quickly when executing the alt+J command, indicating that the final modification (waiting for external commands) in the code was correct. The controller’s response is significantly more sensitive than before, with delays now within an acceptable range.Now, I am considering two important upgrades for this controller.One is related to the software, mainly addressing the issue of how the rotator turns when the target angle is between 0-90 degrees.This issue may not be clear to those who do not use rotators.Generally, we assume that a full rotation is 360 degrees.Assuming the current angle is 330 degrees and I want to rotate to 30 degrees, normally, after issuing the command, the rotator would turn counterclockwise 300 degrees.This is not a problem, but during contests, this rotation time can feel a bit long.YAESU rotators support a maximum rotation of 450 degrees, which is designed to solve this problem. This means I can rotate to 390 degrees and turn clockwise 60 degrees from 330 degrees, which speeds things up significantly.The hardware supports this, but the software needs to consider a bit more.For my position, the most frequently used angles are around 330 degrees (Europe) and 30 degrees (North America), and during contests, most of the time, I am rotating around these two angles.The angle calculation function built into n1mm is very convenient; the European station angle is over 300 degrees, and the North American station is around 30 degrees. Just press the alt+J shortcut, and the rotator automatically turns to the correct position~~~.However, the program issues the command for 30 degrees, and then the rotator starts turning counterclockwise without hesitation.I encountered this problem during the contest, watching the pointer continuously rotate counterclockwise, lost in thought~~~.Why not… think of a way to add more judgments in the controller program?This is the first significant upgrade~~ software upgrade:To add a judgment in the controller program, when the current angle is around 300 degrees and the target angle is around 30 degrees, it should not turn counterclockwise but directly turn clockwise, reaching the position after rotating 360 degrees.(In fact, this software upgrade has already been completed.On the first night, with no propagation on the high bands and too much noise on the low bands, I was reluctant to go to sleep~~~. While bored, I thought, why not just get started and modify the program right away?— So I did.)After testing during the latter half of the contest, the results were good; the controller automatically rotated along the shortest path.The second upgrade is related to hardware.The angle indication of commercial rotators is based on a precision variable resistor, which identifies the current angle through voltage feedback. The device reads the feedback voltage, drives the rotation mechanism through the circuit, and finally indicates the current angle with a pointer.There are two potential factors that can affect the accuracy of the angle.One is the wear of the precision resistor and the wear (or obstruction) of the rotation mechanism, which can cause inaccurate angle indications. (A similar issue was found with the previous G800SA);The other is the initial positioning deviation during antenna installation, which can lead to subtle angle errors. (Other factors include the instability caused by not securely fixing the mast, which can also lead to significant deviations.)The inaccuracies mentioned here are not just a deviation of 1 or 2 degrees but can be as much as 20 or even 30 degrees.So I was wondering if it would be possible to use an electronic compass module to obtain relatively accurate angle values digitally, thereby completely abandoning the original analog method?I have already purchased the electronic compass module (thanks to the powerful manufacturing industry, this thing is only the size of a fingernail and costs only about 10 yuan each).I also bought a 3.2-inch serial screen as a display, using a leftover Arduino UNO board as the processor, and put it all in a small plastic box to create a portable electronic compass.It has been successfully tested, indicating that the main program is running smoothly.Next, I will install the electronic compass on the main beam of the antenna and add the processing program to the existing controller’s Arduino board (the board’s interfaces are all ready), changing the previous analog angle code to the already functioning digital angle reading code.~~~ This way, the direction of the main beam will be relatively accurate, and the rotator will just be a rotator (of course, I also need to consider the original 0-degree and 450-degree limit issues).The main problem to solve now is the long-distance transmission of the digital angle signal.The electronic compass data uses the I2C protocol, which only supports transmission distances of a few dozen centimeters and does not support long-distance transmission.However, we need to transmit over distances of several dozen meters (generally 30-50m). After checking online, a reliable method is to add an I2C~485 conversion module on both ends and use the 485 protocol for long-distance transmission.This conversion module may be in low demand, and it surprisingly costs several dozen yuan each.I also found an I2C signal amplification module, which is also used in pairs and is a bit cheaper than the 485 conversion module, costing around 20 yuan each. I will buy two pieces to test first.If the test is successful, we can use this signal amplification module within a distance of 50m; beyond 50m, for distances over a hundred meters, we will still need to use the 485 conversion module.Wish me success in testing.———————————————————Striving to make communication more interestingDE BA5AU –… …– – ..-