Clem's Stepper Motor Puzzle!

Your idea is correct but schematic is misleading a little and Clem later specified that EN is pulled permanently low. Note that on the photo are only two buttons while schematics has three.

Interesting puzzle!

By leaving CFG0, CFG1, CFG2 open we configure the driver as "16 microsteps, interpolation to 256 microsteps, stealth chop", with a "chopper off time" of 332 clock periods.
So we need 16 pulses to make a full step, assuming 1.8°per step we need 3200 pulses to complete a revolution.
STEP, DIR and ENABLE pins are connected to tactile switches and pullups.
Default active edge is the rising edge, so the step occours at button release.
ENABLE (CFG6) pin is active low: we need to keep the button pressed to enable the driver.

If we only press the STEP button, the motor won't turn at all.

References:
www.trinamic.com/.../TMC2130_datasheet.pdf
www.trinamic.com/.../TMCSilentStepStick_SPI-TMC2130_v10.pdf
learn.watterott.com/.../

Oopsie, my puzzle solution comment went into moderation 

I tried to tackle this puzzle for several days and came to the conclusion that solving this puzzle requires some assumptions made -there are a few unknown variables that are very important and may affect the outcome-. I will explain in detail.

Looking at the Puzzle’s description

• The Arduino Mega is there to show how Power and GND are connected to the stepper driver and has no other connections to the circuit. This also tells us that there is no software debounce for any input
• There is no mention of a hardware debounce in the Puzzle description, the practical example or in the Puzzle schematic but, there is an internal analog filter in the Step and Dir inputs. The datasheet mentions that this filter is designed to help with normal noise in the circuit but doesn’t seem designed to filter noise generated by push-buttons
  
• The Stepper is enabled
  
• The Motor requires 200 steps per revolution
• From the puzzle description, CFG0 and CFG1 are left open
  
  
• CFG0 left open sets the duration of Slow Decay Phase to 332 clock cycles or around 25µ sec (internal clock is around 13.2MHz under normal temperature operation), this doesn’t affect the outcome of the puzzle IMO
• There is no mention of the state of CFG2 in the Puzzle -or well, it is incorrectly mentioned
  
  In the Datasheet, CFG3 is actually the input that sets mode of the current setting, GND sets it to Internal reference voltage. CFG3 I consider irrelevant for the puzzle solution as well

So what about CFG1 and CFG2 then?. Before continuing we have to make some assumptions which are nowhere clarified in the puzzle or any of the supporting images:

Let's assume that the Stepper Driver is correctly set up in Standalone mode (SPI_MODE = 0 or connected to GND), which I believe is what Clem is attempting considering he mentioned MS1/MS2/MS3 as a way to configure the stepper and part of the description of the puzzle and also mentions configuration pins in different states (low, open)

Based on the above, if CFG1 is left open the number of steps per revolution has 3 possible values, each dependent on the state of CFG2 which again, is not mentioned precisely in the Puzzle description. We can ignore Interpolation for the calculations as this is just how the stepper-driver smooths internally each step

• If CFG2 is connected to GND then 400 steps are required per revolution (2 half steps per step * 200 steps per revolution)
• If CFG2 is connected to VCC_IO then 800 steps are required per revolution (4 micro-steps * 200 steps per revolution)
• If CFG2 is left open then 3200 steps are required per revolution (16 micro-steps per step * 200 steps per revolution)
  
  

From the table above we can see something very important to solve the puzzle later on. Note that no matter the state of CFG2, if CFG1 is left open, the 3 possible values are always in terms of microsteps (2 half-steps, 4 quarter-steps or 16 microsteps) and Interpolation is always Active (Interpolation = Y).

How many OR how often?

Let’s see the Puzzle question in detail because I believe here is where things get a little spicy on purpose

I usually see a “how often” question answered with things like “every 15 minutes”, “daily”, “every other week”, etc., so lets double-check that:

Since we are in an electronics community and this is a puzzle is about electronics, I also verified “frequency vs period” and here with the important part highlighted:

At this point, I’m pretty much convinced that the answer to the “how often” of this puzzle has to be in terms of Frequency and not in terms of a quantity (how many).

To answer the “how often” of the puzzle, we need to look at a few more places in the Datasheet. First, let's take a look at the timing section, which tells us that the Step input can operate at a maximum of ½ the frequency of the TMC2130 clock. See the image below

We can safely assume that dedge=0 is the only possible scenario; remember that all over the puzzle Clem is trying to tell us that he is trying to operate the stepper driver in Standalone mode (the MS# pin mentions, no IO connection to the Arduino, CFG# pins left open) and dedge=1 can only be possible by setting the 0X6C CHOPCONF - Chopper Configuration register which can’t be done in Standalone mode, it can only be done in SPI Interface mode.

So, how often then?

From the timing characteristics in the datasheet, the internal clock is the option that makes more sense to me -also there is no mention of an external clock in the puzzle. I also think that a 50°C TJ (Operating Junction temperature) is a safe bet.

With all of the information presented so far, this leaves us with the answer to the puzzle:

What happens if you just leave MS1/CFG0 and MS2/CFG1 floating (open) -- how often would you have to push the STEP button for a full rotation?

When operating at maximum micro-step resolution, the STEP button can be operated at a maximum of ½ the frequency of the TMC2130’s internal clock (13.2 MHz/ 2) or as fast as every ~152ns (0.15µs) in the best case scenario (no bounce or missed actuations in the switch and when operating with the internal clock at around ambient temperature).

What if how many?

Answering how many times the Step button has to be pushed introduces more unknown variables and several possible answers to the puzzle, here are a few factors to consider:

• Are we working “theoretically” with a perfect switch?
• If we have to consider switch bounce, the amount of noise varies by switch
• Should we assume the motor is under no load?
• Should we assume the motor won’t miss any steps?
• When working with stepper motors, the higher the VMOT, the better the motor will respond -this is one of the main objects of a chopper-type driver (more torque, potentially faster response time and potentially less missed steps)
• We know VMOT is 12V (from the puzzle’s description) but we don’t know how well the stepper will respond to this voltage

Just take into consideration the following example:

In a best case scenario (no switch bounce and no missed steps), if CFG0, CFG1 and CFG2 are left open, the Step switch has to be pushed 3200 times for a full revolution (16 micro-steps per step * 200 steps per revolution):

• If we introduce missed steps from the motor, the value goes up
• If we introduce switch bounce, the value in most cases goes down
• Combine the two above and we can only provide an exact answer (3200 times) for the best case scenario

Other possible solutions to the puzzle

Since I mentioned other possible solutions when looking more closely at every detail of the puzzle, I will cover them too.

The TMC2130 operation mode

The TMC2130 has a few operation modes based mostly on how the SPI_MODE is configured. Some youtubers point out that the TMC2130 Stepper Driver’s default operation-mode varies by brand. Here is another youtuber highlighting the same.

There is no mention about the SPI jumper-pad in the puzzle. I can only assume that Clem is actually using the watterott TMC2130 pictured in the practical example (the breadboard prototype picture), on which the jumper-pad is not visible in the Breadboard picture but Stepper Driver datasheet shows it at the bottom of the PCB,

This manufacturer states that the factory’s default state of the SPI jumper is left Open -see the image below-.

Standalone operation

To Solve the puzzle for Standalone operation, we have to assume that the Stepper Driver is correctly set for Standalone operation by bridging the SPI solder-jumper, in this case SPI_MODE=0 (connected to GND) -again, the puzzle doesn't mention this.

The Standalone operation mode is seen in good detail in the TMC2130's Datasheet

In standalone operation, the interface pins of the stepper driver act in a TRISTATE detection (they can be connected to GND, VIO or left OPEN).

The Watterott TMC2130 Datasheet shows that the CFG4 and CFG5 are connected to GND but I consider them irrelevant for the puzzle solution

• CFG4 is connected to GND (sets chopper hysteresis to 5)
• CFG5 is connected to GND (sets the Chopper Blank Time to 16 clock cycles)

From the puzzle’s description and the practical example we also know that CFG6 is connected to GND which is an Active LOW input, indicating that the stepper-driver is enabled but here we have to ignore the fact that the ENABLE pin is Pulled-Up in the Puzzle’s schematic.

The rest of the solution to the puzzle in Standalone operation is already covered.

SPI Mode

According to the manufacturer, this mode of operation is the default operation mode of the watterott TMC2130, seen in good detail in the STEP/DIR application diagram -pictured below 

From now on, I will assume Clem left this solder-pad in unbridged (Open). In that order, the TMC2130 will operate in Step/Direction Driver Mode and SPI Interface (considering that in the TMC2130's Datasheet SPI_MODE has a pull-up Resistor so SPI_MODE=1)

Clem’s practical example

Taking a closer look at Clem’s practical example only (the picture of the Breadboard prototype which I have added again with labels) and assuming power is actually connected to the DC Power Input of the Arduino Mega then:

• The Stepper Driver is enabled for operation (ENABLE is connected to ground and is also stated in the Puzzle’s description)
• SPI_MODE=1 tells me that the Interface pins will operate in SPI Mode (not in the 3-state configuration mode) -again, I’m assuming this solder-jumper was left open
  
  
• From the Breadboard picture then:
  • SDI and SCK are connected to GND
  • CSN and SDO are left open, but these have pull-up resistors when not working in tristate-detection -see the “(tpu)” in the datasheet picture above, so CSN and SDO are Pulled-up to 5V.
  • CSN is the Slave Select input on a Serial communication, which is Active LOW, so unless this connected to GND, this driver will ignore any incoming serial data

According to the getting started section of the TMC2130's Datasheet, the TMC2130 needs initialization when operating with SPI Interface

And here is another place from the datasheet that confirms that the registers are reset on power-up

Based on the information presented, if Clem is using the watterott TMC2130 stepper driver with the SPI solder-jumper left Open, the answer to the puzzle in this case is “the motor won’t turn at all”. 

This makes total sense since to operate any stepper-driver, at minimum the current-limit needs to be set, which in this case needs to be configured via SPI (0x10 IHOLD_IRUN - Driver current control register).

And here is a small demo, where I try the Stepper Driver as is from the factory (SPI Interface). There is no movement from the motor and also no voltage at one of the coil's output because it hasn't been initialized

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

Comment on Clem's Stepper Motor Puzzle!

Poster

Upload Preview

Ok

Cancel

The confusing things of the puzzle

I found a few misleading things in the puzzle that maybe Clem left there for a reason, either way I’ll cover some of them

Taking a look at the Schematic only

• All the switches are NO (Normally Open) and Pulled-Up
• The ENABLE input is Active LOW (see the stepper driver symbol) and the Enable Switch (SW3) is Pulled-Up… just looking at this part, unless you push and hold the Enable  switch (SW3) before pressing the Step switch (SW2), the motor won’t move
• In the puzzle’s description nowhere is stated what is the state of the Enable Switch (SW3) but, there is mention of the Enable signal connected to GND for practical purposes, this discards the previous bullet
• We can also see in the schematic that all the interface pins have their SPI Interface names -one can think ok we are in SPI mode-, but the puzzle description references the Interface pins by their Standalone configuration names -see image below for comparison-, so we don’t know exactly in which Interface mode the stepper is operating when putting all the puzzle's information together
  
  
• We can also see that the pin numbers and the question marks referencing MS1/SDI, MS2/SCK and MS3/CS match precisely the Pololu A4988 pinout. The puzzle precisely mentions MS1/CFG0 and MS2/CFG1 left floating (open) but, on a careful comparison these don’t precisely match -see the image below where I have added pin-numbers matching the Puzzle’s schematic
  
  

If we put all of the above together, perhaps the real answer to the puzzle is that the motor won't turn at all because Mr. Clem failed spectacularly at translating the schematic into the Breadboard 

Jokes aside, thank you so much for this puzzle!. I really enjoyed trying to solve the puzzle and learning about the TMC2130 which I will certainly consider in future projects with stepper motors. 

Luis

In theory considering how stepper motors work, yes!. In practice, stepper drivers can drive at different current values in certain events; the most common are Run Current and Hold Current. The TMC2130's Datasheet states that driver will switch to IHOLD (Hold Current) after 2^20 clock cycles after the last STEP pulse.

So according to the Datasheet, the minimum frequency required to operate the motor at Run Current (IRUN) without it switching to Hold Current (IHOLD) is roughly the Frequency of the Clock / 1million (~2^20 clock cycles) -see below