Confirmation of Y stepper wiring please

DanL

Duet WiFi arriving Monday. Converting from Marlin on MKS 1.4 board.
Installing it on Cartesian printer with 1X, 2Y and 1Z stepper at 24VDC.
The two Y steppers are 1.8, 3.75V, 1.5A, 2.5ohm/ph, 5.0mH.
The two steppers are currently wired in series and run off one driver at 1A and 1/32 step setting.
I am planning to plug both Y steppers in series through one connector, into the single Y driver on the Duet.
Is that the best setup or should I consider parallel or another way to connect?
Pretty new to this...wish me luck in the transition...

Veti

if you wire them in series you half the Speed at which torque starts to drop.
see https://reprapfirmware.org EMF Calculator

since you have 2 Y you probably want the most torque you can get, which mean 1 Stepper per driver.
And the Duet Wifi can only do interpolation at 1/16 microstepping

dc42

As @Veti says, the speed may be limited if you wire them in series, but no more so than using Marlin. Also, when you exceed the speed at which torque starts to drop, the motors will become noisier. So if you are not using the E1 motor output for anything else, I suggest you connect the Y motors to separate motor outputs on the Duet.

PuterPro

Welcome to the Duet world!

If you're just starting with RepRap Firmware (RRF) it can be a bit of a ride. For example, if you've adjusted VRef on your board before, now it's done in a line on the config.g file. That's also where you would define the motors and what outputs do what. Everything is software controlled and configured by G-Code commands.

It's a very flexible system since, as @dc42 points out you can use an unused extruder output to drive a motor, or a "Fan output" for a sensor!

Your main guide will be here: https://duet3d.dozuki.com/ which has all the info you need to setup your system. That said, it can be a bit confusing at first. If you've done any programming or are familiar with command line structure like Linux or DOS you'll be fine.

The main G-Code reference page is here: https://duet3d.dozuki.com/Wiki/Gcode and the wiring & differences from Marlin is here: https://duet3d.dozuki.com/#Section_Getting_started_with_the_Duet_2_WiFi_or_Ethernet

You might bookmark those, you'll live there a lot at first. Read the getting started while you're awaiting your board.

If you get stuck there are many experienced hands here to help, it's a very friendly forum! Have fun!

DanL

Thank you for the quick responses...much appreciated. And thanks too, PuterPro for all the beginner tips. I'm nearly 70, and early in my career I used Fortran and Basic in research - had to write all your own programs then. Today, it's all forgotten and as I start into 3D printing, G-code and Arduino present a whole new learning curve. Ha....great hobby.
I did use the EMF calculator spreadsheet...that's why I switched to 24 vs 12V. Great bits of info in that stepper article.
So, Y's then will go on separate drivers. I'll look into how to do that ....synchronization issues?
~~On microstepping, why is DuetWF only 1/16 when "The TMC2660 drivers are crazy quiet and capable of up to 256th microstepping for smooth and precise machine motion."?~~ ...found the answer....
Thanks again, esp. DC42, for all the tremendous support on this forum.

Veti

@danl said in Confirmation of Y stepper wiring please:

On microstepping, why is DuetWF only 1/16 when

on 1/16 and only on that mode does the tmc2660 provide interpolation to 1/256 microstepping.
You can run directly on 1/256 microstepping but it puts a high strain on the cpu. normally 1/16 with interpolation is fine

see
https://duet3d.dozuki.com/Wiki/Gcode#Section_M350_Set_microstepping_mode

PuterPro

@danl - You're most welcome.

Yeah, I worked for Wang Labs back in their heyday, they were big in the Engineering world because they had a firmware based on modified Fortran with a bunch of baked in engineering routines.

Then there was CPM ... LOL.

So, Y's then will go on separate drivers. I'll look into how to do that ....synchronization issues?

Yeah, I run dual Z that are setup to sync. One way is to have endstops (switches usually) that define home for each. You have to tell the Duet the same info for each axis. There's some others on here much more heavy than I, I'm still a noob at this. Perhaps one of the heavies will help you out configuring it.

BTW - if you refer to someone here type an @ and a list will pop up of those who have posted, just click them. It also lets them know they were mentioned up in notifications (the bell symbol, top right).

Later!

DanL

This is not a good idea, is it?

Run Ymotor1 on Y driver, run Ymotor2 on E1, use only one endstop but add parallel output to both Ystop and E1stop.
The unstopped side can then get out of sync?

As a noob, should I maybe just start by assigning Y's to the series 2xZ driver and assigning Z to the Y driver?

Searched, but can't find best approach to get exact match on two stops - one on each Y stepper. Is it just a matter of tweaking the microswitch positions to match activation points?

Veti

the duet firmware can keep them in sync while running no problem.

i use 2 z motors in parallel without a problem.

dc42

@puterpro said in Confirmation of Y stepper wiring please:

Then there was CPM ... LOL.

And there was MP/M and concurrent CP/M - way way beyond anything that Microsoft had for very many years. I made good money out of improving and customising Concurrent CP/M for OEM customers. We had a version that would support 80 terminals connected to a single PC. Then a customer paid us to extend it to 100 terminals.

Software was written to be much more efficient in those days.

DanL

@dc42, found this from previously addressing this question:
*dc42 ADMINISTRATORS 16 Jun 2017, 22:48
In particular, when you power the printer on and first energise the motors, the motor currents will flow in a particular ratio between the phases and the motors will jump to the nearest multiple of 4 full steps that matches those currents. If the motors were mid way between two such positions when you powered down, then it is possible that the two motors will jump in opposite directions, thereby losing sync by 4 full steps.

I have it on my list to add an option in the firmware to move the motors by up to 4 full steps when you power down so that they are in the right position for power up. Ths may help to keep them in sync.*

What is the update on this please? Is there a synchronization solution now available?

Thank you.

PuterPro

What is the update on this please? Is there a synchronization solution now available?

@dc42 +1 on that. I knew you were an old school guy, LOL. Assembly language was the real way to go ... super compact but not so portable ...

Assembly language? Luxury! (spoken with a Yorkshire accent). We had to enter our programs in binary...

[explanation for the benefit of non-English speakers especially: https://www.youtube.com/watch?v=_nL6isGPhzk]

DanL

@puterpro I kept an old punch card deck wrapped in its doc sheet just for the memory.
It was a Fortran program for mass and heat transfer modelling in a bed dryer for grain.
But all these interacting 3d printer variables seem actually more complex....
@dc42 @Veti
Back to the issue....if up to a 4 step misalignment occurs at powerup, and at 1/16 axis movement is 40steps/mm, that's 10% error, or 0.1mm. Maybe still no issue at 0.1, but eventually would add up to significant misalignment.
Even if some sort of digital e-synch were used, mechanical slop in these cheap steppers would probably result in rotational offset.
So using two endstops - not just as endstops but for homing calibration to prevent drift over time - seems necessary. But how does one eyeball alignment of microswitch endstops to maybe 0.1mm and also get repeatability from the switch actuation?
A reality check, though....the current setup has a two stepper Y setup, steppers in series on one driver and only an endstop on one side. It's run many prints over many hours and the X-frame still seems in alignment. Maybe over-thinking this?
Are permanent mechanical hard-stops for alignment a workable idea?

Veti

how did you do the alignment on the old board? you have the same problem if you put your motors in parallel.

it should work like this
https://duet3d.dozuki.com/Guide/Independent+Z+motors+and+endstop+switches/18?lang=en

DanL

@veti my printer came assembled with firmware loaded. The two Y steppers were wired in parallel. I changed that to series, went to 24volts and readjusted Y current to 1A (1.5A motors). I didn't do any marlin adjustments. Calibration and test cube geometry were excellent.
I did check mechanical alignment when I first got the printer and squared the Y manually to be perpendicular to the X rail. That's it. Never thought it through til this thread gave me a lot more info to think about.
Looked at the link - thank you.. Need to digest it....the inclusion of Z leveling function and related G-coding still confuse me. I'm early on a learning curve here...
Excuse my dullness, but still don't get how to position and adjust two separate endstops to get geometrically (squared to X) matched activation with accuracy, precision and repeatability.
Getting the Duet installed and playing with it will help me learn more.

Here's an interesting link discussing this issue:
https://www.zaber.com/technical-articles/driving-parallel-axes-with-lockstep-movement

Maybe I'll just add a crossover belt and drive both sides from a more powerful stepper.

PuterPro

Maybe I'll just add a crossover belt and drive both sides from a more powerful stepper.

Yep, that's one way out of the mess. Lots of Creality CR-10 users go that route to get a dual Z, mine is a CR-10S which came with dual Z setup with the motors in parallel. Fortunately, all I needed to do was add another switch (endstop) for the second Z then separate them in the config file.
Got the switches aligned pretty close and the software aligns the rest of the way. (I run a BLTouch).

I suppose over time there will be some slight wear due to it being not perfectly aligned (because the X carriage is ever so slightly tilted) but sometimes you pick your battles!

PuterPro

I kept an old punch card deck wrapped in its doc sheet just for the memory ...

LOL. I remember watching my older brother who worked as the lead tech for a Honeywell data center in NYC hand a data entry girl a 300 card stack of a diagnostic he had written and punched by hand. She dropped it and they went everywhere. I thought he was going to jail for murder... We don't have words for that angry.

DanL

Crossover belt option seems good, but difficult to implement.
So, to start out, since both Y steppers are already wired in a series harness, they will simply be plugged into the Y driver.