Will there be support for the Trinamic 2130 drivers in the near future?

Controllers Smoothie
,
1

Will there be support for the Trinamic 2130 drivers in the near future? Working on some designs and figured that they would be a good applicant, however I know that their current SPI setup isn’t part of Smoothieware.

Thanks!
Griffin

2

We have code for several trinamic chips via SPI in smoothieware : http://smoothieware.org/advancedmotordriver

I expect the protocols to be very similar for chips of the same manufacturer, maybe it’s not too hard to add your specific one ?

3

@Douglas_Pearless and I tried porting his 2130 code back into smoothie1. It’s a whole different animal. Never got the SPI setup working properly.

4

I have it working for SmoothieV2 but have not had time to back port to SmoothieV1 and also don’t have suitable v1 hardware to try.

5

@Douglas_Pearless I’m working on getting a few boards together to send to you and @cprezzi

6

Most trinamic drivers are the same so there should not need to be any changes needed to the existing TMC2xx driver already in smoothie.

7

@Wolfmanjm 2660 (26x) and 2130 have different SPI protocols according to @Douglas_Pearless
Also of note is that the enable pin has to be wired as usual and is not handled via SPI for the 2130.

8

^ that’s been the battle so far. I know the 60’s work, but aside from one Marlin post of someone forking the 2130’s into the firmware, there is little to read on the 2130’s being used to date…

9

I don’t think it does have a different SPI protocol, I checked the docs and it looks exactly the same. the entrire range of trinimacs seem to be the same SPI protocol. Having enable as a pin rather than over SPI is trivial and does not warrant a new driver.

10

@Wolfmanjm to be clear: “it’s different registers and stuff in the spi communications” plus it uses a dedicated enable pin.

11

I have added the TMC2130 drivers to my SmoothieV2 repo and it works. The internal registers and SPI communications are different and I needed to make a number of changes to get it working :slight_smile: And it
does work, my FirePick delta is running V2 code. https://github.com/DouglasPearless/Smoothie2

12

Note: When driving this at speed and 256 micro steps on three 0.9 degree stepper motors, the internal timing of the code is critical as we are talking more than 102,400 step and unstep calls per revolution of the motor(s). I had to increase the slow ticker by 10x speed otherwise the mechanism drove into the endstops even at relatively slow speeds. I’ve been considering at either moving the stepping code code to the M0 or using a Spartan 6 to implement NCO’s, endstop and probes, etc.