TMC2209 support

[Dilbert0815]

I had problems getting TMC2209 step sticks working with UART configuratio in Marlin 2.x..
Some Bigtree TMC2208 before were working correctly, so I was debugging the case.

Problem is on these TMC2209 sticks the port RST1 is wired directly to PDN_UART from TMC2209 while M3 port is having an addtional 1k ohm resistor to PDN_UART of TMC chip..
So using the normal UART jumper on SKR1.3 we get a total of 2k (1k + 1k) in the TMC TX line, causing it to fail.

Hardware Fix:
Fix is possible in hardware by using a adtional jumper in SPI jumper block between pins 1/3.
This shortens RSTI and M3 ports of step stick, effectively disabling the stick internal 1k ohm resistor. UART jumper is of cause also set.
Communication is now working and TMC_DEBUG with M122 returns correct register values and reports TMC2209 communication as online.

May be you can add this to board FAQ for troubleshooting TCM2209 stick problems of other users.

[Dilbert0815]

Jumper illustration on X-stepper port:

[NaterNore]

Hey, I was wondering if I could get these to work on the skr 1.3 non pro version. That mod you made is that little jumper I'm assuming? And this would work on any standard tcm2209? Thanks I'm a bit new to this.

[ArchSnatcher]

Thanks for this advice Dilbert. I've been getting an "all Low" for Y on my 2208s and a "TMC connection error" on the LCD for the last week. I've been through countless pages of issues that are similar to mine, but we're not able to fix the problem.
I'm going to try this after work to see if it is an extra resistor. I'm using a total of 4, Eryone 2208 v1.2 drivers.

[Dilbert0815]

Hey, I was wondering if I could get these to work on the skr 1.3 non pro version. That mod you made is that little jumper I'm assuming? And this would work on any standard tcm2209? Thanks I'm a bit new to this.

Actually this IS a SKR1.3 non-pro. The Pro is actually revision v1.1.

But there ist no standard for TMC220x. Basically any vendor can route secondary signals to any unused port. This is exactly what happens here with Rx/Tx.
Bigtreetech has routed their board for THEIR TMC steppsticks, and thats obviously not the idea that FySET had in mind when designing their 2209 sticks.
The TMC2209 is a chip standard, but the step stick is a vendor desgin and there are simply a few versions out there how to rout non-essetial ports to stick pins. This are ussually the stick pins that carry the setting pins on older chip designs.

You should have an ohmmeter ready and check how your sticks are routed to the TMC chip pins.

[ArchSnatcher]

And it wasn't a resistor. The XYZ and E all work perfectly well when I turn on 'movement' mode on the LCD. I'm still getting the tmc connection error, but it is interesting. I have 4 total 2208s on xyze. When I run a check through terminal xy and z will be fine, but e will be 'All Low'. I'll move x to y (for example) and xze will be fine with a 'All Low' on y.
I'm starting to think this is a tx/rx issue.

[NaterNore]

So you're telling me that if I get the stepstick drivers I'd have to do this but if I get the bigtree or some other brand I might not?

[jimmyeao]

Curious, do the 2209s still need the pads bridging like the 2208s did?

[thisiskeithb]

To add to this, if you're using genuine Watterott 2209 V2s un UART mode and would like to use sensorless homing without "flying wires", some hardware modification is required:

The Watterott 2209 V2's DIAG pin is in a different spot than what the SKR 1.3 is setup for on other/earlier drivers.

No extra wires required, but you have to be careful if swapping drivers in the future to make sure only a single pin is used for diag/sensorless homing so you don't short anything out.

Disclaimer: This is just for general information and I'm not responsible for any damage to your driver/board/printer.

[MarkV3082]

Dilbert0815 Thank you very much. This jumper config has solved my problem. Endless variations with these chips as you say.

[asdmin]

@Dilbert0815: you left the MS1 and MS2 pins floating. In UART more they are to set the slave address. Isn't it a problem?

[asdmin]

@Dilbert0815: also, it seems that this mod with the cross-jumper drives the SPREAD at times when UART communication takes place. As I see, during communication on the UART interface the driver is 'required' to rapidly switch between spreadCycle and stealthChop many times. Isn't it a problem?

And this is not the only problem: based on 'TMC2209 DATASHEET (Rev. 1.03 / 2019-JUN-26)', chapter '5.1 General Registers', a high level on SPREAD (see our case) inverts whatever value was configured for GCONF.en_SpreadCycle. As I see, the printer will always work in the opposite mode as it was configured by GCONF.en_SpreadCycle

Did you cut the pin on the silentstepstick?

[Dilbert0815]

@asdmin : I don't have the board mounted acutally.
But there may be a misunderstanding. SPREAD ist not connected with this pin, at least not on the sticks I have here. If yous sticks have SPREAD connected to any of these pins do NOT use the mod.
Here on the FTYSET sticks I have no other connection is these pins, they are both just UART (RX/TX).
They just put the RX/TX seperating resistor onto the stick and have the TX port on the wrong pin for the SKR board.
As said: endless variants of these stickts, just make shure what pinout your vendor chose for these 'optional' pins on these sticks.

MS1/2 should have internal pull-down resistors according data sheet. It is normal to leave them open to my knowledge.

[The-Monkey-King]

@Dilbert0815 : Please update this as a hardware fix for the FTYSET sticks. I am receiving the same error messages for the BTT SKR V1.3 and the BTT TMC2209s. I tried your pin fix and the TMC2209s from Big tree Tech still issue null responses from the X, Y, Z, and E0 registers. (0x00:00:00:00)