[BUG] TMC 2130 does not work in SPI mode

[dronecz]

Did you test the latest bugfix-2.1.x code?

Yes, and the problem still exists.

Bug Description

I designed custom board for FluidNC, but find out that there is not support (and it is not planned) for features which I need for my pick and place machine so I decide to switch to Marlin.

My board is based on ESP32 andn it is using 74HC595 shift registers to control stepper motors, but issue which I´m facing now is that I can not establish connection via SPI. This is defintely SW issue as if I upload FluidNC firmware, all axis are working fine.

Here is my log from Printerface:

Connecting...
Attempted to write invalid text to console, which could be due to an invalid baudrate
 External Reset
 Brown out Reset
 Watchdog Reset
Marlin bugfix-2.1.x
echo: Last Updated: 2022-07-17 | Author: (none, default config)
echo: Compiled: Jul 18 2022
echo: Free Memory: 313444  PlannerBufferBytes: 1408
echo:Hardcoded Default Settings Loaded
echo:; Linear Units:
echo:  G21 ; (mm)
echo:; Temperature Units:
echo:  M149 C ; Units in Celsius
echo:; Steps per unit:
echo:  M92 X40.00 Y40.00 Z40.00 A4.44 B4.44 C40.00
echo:; Max feedrates (units/s):
echo:  M203 X500.00 Y500.00 Z250.00 A50000.00 B50000.00 C5000.00
echo:; Max Acceleration (units/s2):
echo:  M201 X1000.00 Y1000.00 Z2000.00 A3000 B3000 C3000
echo:; Acceleration (units/s2) (P<print-accel> R<retract-accel> T<travel-accel>):
echo:  M204 P1000.00 R3000.00 T3000.00
echo:; Advanced (B<min_segment_time_us> S<min_feedrate> T<min_travel_feedrate> J<junc_dev>):
echo:  M205 B20000.00 S0.00 T0.00 J0.01
echo:; Home offset:
echo:  M206 X0.00 Y0.00 Z0.00 A0.00 B0.00 C0.00
echo:; Stepper driver current:
echo:  M906 X800 Y800 Z800 A700 B700 C800
echo:; Driver stepping mode:
echo:  M569 S1 X Y Z A B C
echo:  M569 S1  A
echo:  M569 S1  B
echo:  M569 S1  C
Testing X connection... Error: All HIGH
Testing Y connection... Error: All HIGH
Testing Z connection... Error: All HIGH
Testing I connection... Error: All HIGH
Testing J connection... Error: All HIGH
Testing K connection... Error: All HIGH
Printer is now online.

I searched issues here, searched Discord channels (looked for "I2S", "All LOW", "All HIGH"), but did not find solution.

I tried use TMC_USE_SW_SPI, tried put X CS pin on real HW pin, but without change. I´m using BIGTREETECH TMC2130 V3.0 .

As I´m using custom board definition, I´m including files so you can compile same code as me.
marlin_configs.zip

I ordered ESP PROG so I will be hopefully able to use debugger and see what FW is doing.

Thanks for help or any direction.

Bug Timeline

No response

Expected behavior

No response

Actual behavior

No response

Steps to Reproduce

No response

Version of Marlin Firmware

Marlin bugfix-2.1.x (Jul 18 2022 02:26:39)

Printer model

none

Electronics

custom board

Add-ons

No response

Bed Leveling

No response

Your Slicer

No response

Host Software

Pronterface

Don't forget to include

• A ZIP file containing your Configuration.h and Configuration_adv.h.

Additional information & file uploads

No response

[Bob-the-Kuhn]

You may end up using the TMC2130s in standalone mode. Is there a SPI only feature you need from these chips?

TMC_USE_SW_SPI assumes all SPI interfaces have a common TMC_SW_MOSI, TMC_SW_MISO & TMC_SW_SCK with dedicated chip selects. I see the dedicated chip selects but not the others.

I'm not aware that Marlin has the ability to control/use the 74HC595 shift registers.

I'm thinking this request is outside the scope of this forum. This is dedicated to Marlin bug fixes.

There are a couple of Facebook Marlin support goups (3Dtechtalk is one) plus a RepRap one that are better suited to this type of question.

[EvilGremlin]

Which SPI actually used for TMC? If you try using one shared with flash i wouldn't expect it to work at all.
@Bob-the-Kuhn it's very old code for ESP, step stream output through I2S, sorted out by shift redisters

[ellensp]

@Bob-the-Kuhn esp32 has built into the HAL shit register support, to extend the number of IO pins.
I dont know if you could software spi over it though.,,,

[Bob-the-Kuhn]

I'm amazed at what they've been able to do with the ESP32 platform. Definitely a world I haven't explored.

[EvilGremlin]

CS on extended pins should be possibe anyway, with hw or sw SPI (though no idea if current code/framework can do arbitrary CS pins for HW SPI)
Yep, random 4 year old cite from reddit: With the current esp-idf, you could simulate it by adding your devices as one device (without a CS pin) to the spi driver, then defining a pre- and post-transaction callback that pulls down the GPIO connected to the CS of the device you actually want to control. And esp-idf reference don't say anything about extra CS pins (above 3 native) possible.
So it's not a bug, but platform limitation that you must work around somehow.

[dronecz]

Which SPI actually used for TMC? If you try using one shared with flash i wouldn't expect it to work at all.
You mean internal SPI flash which is in ESP32 module? That one is on SPI1 (AFAIK) and I´m using VSPI which should be SPI3 (pin 18,19,23).

I understand that there is probably limitation in using shift registers in Marlin with current code, but if you support SPI drivers and settings current via SPI, I should be (at least) able to communicate with driver.
As I meantioned, I tried initialize driver on X axis with GPIO CS pin (33), but result is the same as with shift registers. If I´m not mistaken, I supplied all necessary pins to communicate over SPI (SDI, SDO, CLK, CS).

With the current esp-idf, you could simulate it by adding your devices as one device (without a CS pin) to the spi driver, then defining a pre- and post-transaction callback that pulls down the GPIO connected to the CS of the device you actually want to control.

ESP-IDF change in last 4 years so it is possible to control CS pin via I2S and shift registers. FluidNC has code for that.

I would like to help to resolve this, but I would need some point where to start as I do not know how Marlin internally works.

[stef-ladefense]

• To use TMC2130, TMC2160, TMC2660, TMC5130, TMC5160 stepper drivers in SPI mode
• connect your SPI pins to the hardware SPI interface on your board and define
• the required CS pins in your pins_MYBOARD.h file. (e.g., RAMPS 1.4 uses AUX3
• pins X_CS_PIN 53, Y_CS_PIN 49, etc.).
• You may also use software SPI if you wish to use general purpose IO pins.

[EvilGremlin]

LOL, this much is obvious. Please read about esp-idf SPI peculiarities.

[dronecz]

I finally figured out what was my problem. After a lot of testing I found out that I can not toggle CS pin at TMC driver if this pin is driven with shift register and I2S. When I control just STEP and DIR pin via I2S and use some GPIO for CS pin, it works as expected. Does anyone know code for I2S?

I will try to make PR for documentation with my findings as it will definetely help other people.

[EvilGremlin]

esp-idf framework is not aware about I2S pins, you'll have to redefine/overload any SPI CS related functions in HAL to use them

[github-actions]

This issue has had no activity in the last 60 days. Please add a reply if you want to keep this issue active, otherwise it will be automatically closed within 10 days.

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.