TMC2208 on Cetus3d mk3

Here is some info on how to get TMC2208 stepper drivers working with Cetus3d mk3:

https://github.com/Sergey-77/Cetus3d_TMC2208

It's not a perfect solution but it works.

 

 

2 Likes

Great work ! Was thinking about doing something very similar.

Thanks, Brantwinter!
Just uploaded 3 steppers version of the board with individual serial connections to TMCs. I thought separate channels would give better tuning possibilities from micro-controller. But I think setting current with potentiometer still producing better results. Probably because of sensing resistance is too low for 30Ω steppers.

1 Like

Nice work. I’ll be building a PCB soon.

But why? What improvement will this bring? Noise reduction? Better quality?

Huge difference in noise; also better surface quality (less salmon skin effect).

Interesting. How noisy in dB was your Cetus before, and how noisy it is now? Any before and after photos of a sample model?

Virtually silent after, don’t have calibrated dB measurement, nor do I have Cetus, but Up Mini. However drivers and motors are the same, but I think Cetus make a little more noise with its rails.

This post has link to two videos showing homing with original and trinamic https://www.tiertime.com/forum/viewtopic.php?f=28&t=56613&sid=24f7b1b4cb47cda285f6c3d380143405#p51641
Can see sample model with trinamic 2224 driver here:
https://www.tiertime.com/forum/viewtopic.php?f=28&t=56613&sid=24f7b1b4cb47cda285f6c3d380143405&start=20#p51687

Just to be clear ,different printer, but change should be pretty much the same

@bjorn, thanks for the video link, now that is the best comparison I’ve seen thus far. I’ve seen many other videos from other 3D printing channels (not Cetus mods though) and I have to say I am disappointed with the outcome as the noise reduction wasn’t significant (and they use a calibrated SPL meter).

Interestingly, I found some information from the link you shared. The DOORCHECK port… So, I can use pins 1 and 2 only and pin 2 is the signal from a NC switch (door)?

Trinamic drivers when driven in stealthchop at low speeds, its virtually without motor noise. Your fans will be the loudest component by far (and linear rails for the cetus)

Not quite sure about the door thing, i thought it was a hall effect sensor, but it could be a reed switch. I was primarity after mapping the LEDs and the switch one the side of the Up Mini2, but I never looked at the actual door switch, just noted it had three wires and asumed in the back of my head it would be a hall effect sensor.

@sergey77, what are the Vref for the drivers in your design? Will it make the stepper motors operate at lower temperature?

Vref is set to x 0.15v, y 0.1v, z 0.2v. Steppers runs cooler because current is set so low, just high enough not to skip steps when printing.
In Attiny85 code IHOLD_IRUN can be set for different current sensing formula. Second is more precise for low vref settings.
IHOLD_IRUN(0x1F10); // Irms = Vref * 0.707
IHOLD_IRUN(0x1008); // Irms = Vref * 0.37565

@sergey77, thank you for that information. I’m sorry I didn’t read the notes in Github.

My 2208 (3 of them) arrived last week and I just need to brush up on my PCB making skills and I’ll give this a go.

@sergey77, I printed the PDF file for the PCB but it’s small. What’s the right scale?

Scale should be 100%.

I believe a dxf is unit-less so the software that imports the file needs to know what units the original was created in. mm, microns, inches etc. There are a lot of dxf in the world that are for example 25.4 times bigger or smaller than they should be. I had some engine mounds out for laser cutting once and the sub-contractor phoned to ask if they should really be the size of the hood. (I’d modeled in mm, he’d imported as inches, so he was looking to cut my 100mm long mount as 100 inches…)

If on;y there was a standard of measurement the whole world used…

Owen S.

@sergey77, I printed it at 100%, sigh. Anyway, i’ll try to figure it out.

Did you print three or four stepper version? I will check scaling. You can also print PCB from KiCad project.

Printed files with 100% scaling on US Later(8.5x11in.), everything seems to be right. Not sure what would cause it to print in different size.

Printing with default settings on A4 would be fit to page and cause it to be the wrong size, but not by much. Printing with 100% scale on A4 should either be correct size or cropped.


(that in f.ex. chrome should work for you metric types)