I am happy to hear that you got everything working. Once it is working, it runs great. I have two copies of the clock running side by side. The second hands are still tracking within a second after running for more than a month. There will be another release of the "silent desk clock" series coming soon with bigger bolder gears. A second back layer of gears was added to give continual motion. The first release will be quite large at 15" by 15", but an 80% scale version at 12" by 12" might be possible. Experiments on the smaller scale clock have just started. I even decided to build a simple circuit board to use in place of the CNC Shield V4. It won't be cheaper, but it will use simple soldering instead of the back side wiring. It will probably be offered as a kit or a fully assembled and tested module. These updates should be available in a few weeks.

It sounds like there may be two issues. The motor current needed to be adjusted slightly higher and you took care of it. The clock needs very little power. Some motors can run with Vref at the lowest possible setting. Other motors need a tiny bit more current. I doubt that any motors would ever need more than 5-10% of the maximum available from the TMC2208. The second issue is the speed. The serial debug monitor will provide a ton of information. It will show the program revision, the target speed including the number of steps sent to the motor, and a display of how well it is tracking against the RTC. How much too fast is the clock? I would only suspect the TMC2208 as the cause of any speed issues if it wasn't set up for the proper number of microsteps. This would result in the clock running exactly 2X, 4X, 8X, or 16X too fast. The root cause might be from a bad solder joint on edit #2 to connect the speed jumpers to Vcc.

The speed selection jumpers were added when the larger size clock was designed, and each clock had slightly different gear ratios. The description of the jumpers is in the latest version of the addendum that can be downloaded from https://www.stevesclocks.com/sp6

SP6 was originally designed using the resistor based controller. There is an addendum describing the CNC Shield V4 that describes the jumper wires. There are 4 jumpers used by the sketch. The CNC Shield V4 has them labeled as CoolEN, Resume, Hold, and Abort. CoolEN sets the motor direction. The other 3 jumpers set the speed. SP6 should have a speed of 011. The Resume jumper should be out. Hold and Abort jumpers should be in.

Different size clocks have different gear ratios. The assembly manual should list the settings for your model.

You set the time manually, just like any of the other clocks. If Vref is set low enough, you can hold the second hand gear until it matches the current time. It should stay accurate once it is set.

That is great news. You should like how quietly it runs in the clock. Please report back if you realize what the final fix was. It might help others.

Seems like a good idea to take a short break and come back later. I just built another controller and the first trial had a bad RTC. The RTC appeared to be running, but at least 10X too fast. Wierd. Swapping to a new RTC and it works great. It's a good thing the components usually come in sets of 3. I may update the algorithm to add a 5th jumper position with extra debug information. And I may build a small circuit board to use instead of the CNC Shield V4. It would have simple soldering instead of jumper wires on the back of the board. It would be a few weeks away before it will be ready.

It is hard to debug remotely. The code is giving step commands to the TMC2208. You can tell by the "+" characters on the serial debug monitor. There could be a number of possibilities preventing the motor from moving. Here are a few guesses with my estimates as to the most likely. Defective TMC2208 No power to the TMC2208. This is the diagonal wire at the lower left corner on the back of the board. Double check the soldering connection. Defective stepper motor cable Defective stepper motor Defective CNC Shield V4 I have heard feedback from one builder that his RTC was defective. He swapped it with another and his clock started working. I have built 4 or 5 boards and all of them work. Defective parts are always a possibility, especially with low cost suppliers.

Maybe you could try to debug one component at a time. I say this because it is possible that one of the components is defective. Testing them one by one could help determine where to focus the debug. The Arduino Nano is programmed with the new algorithm that doesn't hang waiting for the RTC. The "+" sign appears every second after a certain number of step commands have been sent to the TMC2208 stepper driver. It has already been tested outside the CNC Shield V4. Remove the TMC2208 and RTC from the CNC Shield V4 and add them back in one by one. Unplug power between each step. Start by plugging the Arduino Nano into the CNC Shield V4. Does it still show the same message? Add or remove some of the configuration jumpers to see if the algorithm detects the change. Add the TMC2208 into the CNC Shield V4 without the motor attached. It should still work. Unplug power, connect the stepper motor, then add power back. If the algorithm is still working like before, then the TMC2208 is receiving step commands and the motor should be rotating. You can adjust the Vref potentiometer to change the motor current. The last step is to add the RTC onto its header. The serial monitor should change to a mixture of "+" and "-" as it starts tracking the RTC.

They hide the .ino inside a zip file. Try this file.

It is a simple text file. Ctrl-A should select everything. Ctrl-C to copy everything selected and Ctrl-V to paste.

Open a new sketch and delete the few lines of code that appear. Cut and paste from the txt file into the new sketch.

The wiring looks correct, but it still does not seem to recognize the RTC. Try using this code that will timeout after about 5 seconds if the RTC is not present. It should continue, but the time will be about 5% off. I had to rename *.ino as *.txt.

Look at the wiring for the real time clock. There are 4 wires on the back side and one wire on the RTC itself. And it needs to be inserted in the proper position on the 4 pin header.

Does the serial monitor show just the one line and then stop? Or does it show multiple lines? If it is only a single line, then it is stalled waiting for the real time clock. I should add code to the algorithm to print "Waiting for clock..." and "done" when it successfully reads the RTC. The power adjustment screw is just like a faucet. Turn it anti-clockwise for less power.

It is not a monitor, but rather a window that pops up when programming the Arduino Nano.

The serial debug monitor is built into the Arduino IDE. It uses the same USB cable that programs the Arduino Nano. It pops up when you click on the icon that looks like a magnifying glass. Or click "Tools ==> Serial Monitor".

More information is needed to try to figure out what is happening. What does the debug serial monitor show? If it shows a single header line before halting, then it is likely waiting for the real time clock. The I2C driver will stall if the RTC is missing or not working. I would like to figure out a workaround to display an error message instead of just stopping. The serial debug monitor will provide some clues. Does the stepper motor make one tiny motion at powerup and then go weak? I believe this to be the TMC2208 calibrating itself at powerup. This tells you that it is receiving power but not receiving any step commands. The serial debug monitor should give more clues. What position is the Vref potentiometer? Start at about the 10% to 20% position. It may be able to be adjusted for lower motor current after everything is working.

Nice. It looks like a chocolate mint. It is OK to stretch the pen spring to provide extra tension. My next clock has a smaller gap to naturally provide more tension. It's good to know about different motor variations that can be made to work. There seems to be hundreds of gear reduction options. I should modify the flipper arms to add a pocket on the back side to move the magnet a tiny bit further away from the reed switch. I considered using a mechanical micro switch instead of the reed switch. This eliminates the magnetic attraction, but every switch I tested was considerably louder than the reed switches. Steve