The main piece of hardware, along with its testing devices

So here’s the problem: I like redundancy and I like control, fine grained, precise control of my devices. These are obviously not the aspects that drove me to choose the EQ3 mount when purchasing it, but since it is the mount I have at hand, and no physical room nor real opportunity for something better, I decided to hack this mount.

Independently of Hernán De Angelis who looks like he did a great job, from the same starting point — so kudos –, I got my hands dirty and my screen full of code.

So, here’s the interface as we in software like to put it: a replacement for the original black box, the dual axis drive, sidereal rhythm for North/South, no goto, no ST4, and buttons and a switch to move at 2x, 4x, 8x sidereal speed. Here’s an image. I hacked  the original controller to expose the traces of the pcb leading to the mechanical buttons, which are mere pullups, to be exposed as an ST4 port, and given the circumstances, it works acceptably — it helped me a lot tracking Venus in the great Venus season of mine in 2020, without a polar alignment, and it keeps helping me out when filming the Sun, also without proper polar alignment. Further more, I hacked it by extending it with a buck-boost circuit to accept a wide range of input voltages, from about 5V to 30V, providing the original PCB with the 6V it specifies. I keep the original motors and gears.

The original EQ3 motorization kit

Given this interface specified above, the job was to build a new box, with physically nice buttons and a screen of some sorts, that implements the ST4 standard, and drives the original motors on the mount. I keep the original steppers for now, for interchangeability’s sake. Any extra functionality is to be on top of the original’s, so that an instant  downgrade by swapping the controllers remains as a possibility.

So let’s go diagram:

A bit of reverse engineering: other than hearing the tick-tack of the RA motor while doing sidereal tracking — deducing that it’s probably doing full steps, I wanted to see how stuff looks like in real life. I mean it is a stepper motor, nothing special about it, no surpises to be expected, it is just a stepper with a huge reduction gear in front, to convert the speed of rotation into torque. But anyway, out the oscilloscope cometh.

Oscilloscope image of what the RA motor is doing while tracking at sidereal speed, with an ST4 correction at 8x

I have a soft heart for ESP32 micros. I particularly like the development board that comes with an OLED display glued onto it. It is unfathomably handy when compared to what we had like a decade or so ago. The downside: it has nowhere near enough GPIOs to do all the job. So: the ST4 and motors stayed on the ESP32 for near real time handling. And the front panel got outsourced to an arduino nano — didn’t have i2c extenders at hand, and didn’t want to wait.

Imaging in Calcium K. The first light of the drive. It doesn’t have a front panel yet, at all.

About the calibration: I did a lazy job. I let the mount rotate for about 147 degrees, then counted the steps: somewhere around 303k. Then assigned hours to the degrees and deduced the frequency of the pulses needed to achieve a rough sidereal approximation. It kinda does it’s job. FOr the declination, the overall gearwork meant that I had to do microstepping to get the fine grained control, but it is done from hardware switches, the ESP32 has no idea it’s microstepping.

So the next iteration will be running on something like an Arduino Mega with plenty of pins, possibly more buttons, with the LCD as a slave device, or even with no front-panel / realtime division.

Some day I’ll include all the boring technical details, source code and circuit board. For now, let this post be a reflection of the adventure I was in.

family portrait

with labels

Dual Axis Theatrics

family portrait in November 2022

A video demo, narrated in Hungarian