Printer does not exist yet, but is based on the CR10S PRO V2.
mainboard octopus V1.1
SBC Rpi 4
Klippy.log not applicable yet as it is not up and running
I read several times that daisy chaining motors is to be avoided. I want to understand this better.
Does this refer to 2 motors on one driver or is this relating to something else? The original CR10S PRO V2 has both z-axis motors linked to one driver.
If I were to increase the motor size and amps, I would need to assign 2 individual drivers and configure them separately. Drivers I have are TMC 2209.
Should I expect trouble due to the driver adjustments being slightly different? I would expect it will be near impossible to configure them exactly identical, given the crudeness of the onboard potentiometer?
Would love to hear from those of you who have experience with this question. I am still in the design stage, currently soldering my control box cables. So any considerations can still be made at this point.
Kind regards,
Kees
Ok, I am in total agreement with your reply. Just because the original CR10 did it, it does not mean I have to also. The reasoning behind choosing separate drivers is too solid.
But, what about the difference in stepper driver adjustments? Will this potentially cause synchronous errors or a mismatch in resonance? If two motors are doing the z-axis there is no risk unless you have a non planar machine. But if you choose to run two Y-axis motors to sling the bed, they are very much part of the printing process. So my question is will this potentially cause trouble?
By the way, this “daisy chaining” of motors, is this the same matter or is it related to something entirely different?
If this was an issue then all printers that rely on multiple synced steppers would be in deep s$!%, like Delta, Core-Something or printer with 3+ Z-axes.
If you have enough pins, you want to address your drivers with either a dedicated UART pin for each or in case of SPI with a dedicated Chip Select pin.
If you are short of free pins, you can also attach multiple drivers to one UART pin or in case of SPI, only use one Chip Select and pass on the driver selection via the MOSI - MISO pins from driver to driver. This is typically referred to as “Daisy Chaining”
Sorry I am not trying to sound pedantic but mechanically there are a few differences here between steppers working along side each other not locked, or in synced with each other and mechanically locked. I feel I have not stated my intended use of linked motors correctly, so excuse me for this, probably over simplified, explainer.
To elaborate first on the delta, the three z-axis are independent from each other and the resulting motion of three independent motors correlate to a motion in one plane, which is used to print the layer. If one motor is out of sync with the others in a small way, it does not immediately cause a problem, it simply results in a minute discrepancy in the print. Probably this will be so minute you will never even notice.
In the core whatever it is basically the same situation, because the 2 motors driving the x and y plane are free to do whatever they want, which results in a position in the x-y plane. The 3 motors working the z-axis are never even important, since it isn’t part of the printing process in a direct way. They simply need to end on the same position in order to have a new layer that is perfectly planar with all the other ones.
Now imagine driving a bed with 2 motors that are fixed on the same belt or ball screw!
This is a fixed mechanically closed system that possibly (here I am not sure) requires very well tuned settings of the drivers.
This is what is my question relating to. In servo systems this is used very often. 2 motors driving a mechanically closed system from either one or two drivers. But settings in servo drivers are easier in this respect.
But in stepper drivers I do not know of this application and I am wondering how much of a problem it would cause.
So please accept my apologies here for not being clear in the first place, and for trying something that is not according to the accepted norm.
There are printer designs that have 4 Motors for X and Y, mainly to increase torque and to shorten the belt paths. So one belt is driven by a stepper on each end of the loop.
And still it is not an issue.
What likely makes sense is to couple motors with the same spec and not try to couple a Nema14 with a Nema23
Wouldn’t you be able to do this if they were wired in series? In parallel, you would have a problem due to the mismatched back EMFs from the two different sized motors.
Actually, thinking about it, I think you’d always want to wire them in series because any difference in loads across the two steppers will result in different BEMFs.
Regardless, I’d prefer having a separate drive for each stepper.
@mykepredko: The discussion is about two steppers on two different drivers but coupled in a closed loop system, e.g. driving the same belt.
The concern is that minor variations in the timing of the two stepper motors could disrupt the balance of the system.
Edit:
Not if you want to have halfway dynamic axes. Both steppers will only get 12V which severely reduces their dynamic abilities. If you’d supply them with 48V then it might be an option
Correct me if I am wrong, but in my opinion, this is more about the use case. I would never use two motors daisy chained together on, for example, the x or y axis, but on the z axis, sure. It doesn’t do as much moving, and so the compromises would be less than negligible. But on the x/y axis, I would totally use two separate drivers. As @Sineos said, the amps required to drive each motor would have to be doubled, and you would have to risk frying your driver. I also don’t really know why you would even attempt to double two motors together on one driver. The cost to buy a few extra slots on your control board is so cheep I would rather buy them than risk frying drivers.
Just a side note, I do chain together my z motors, but doing so doesn’t allow me to automatically level my x gantry on my CR-10 V2. If I wanted to do that I would have to assign separate drivers. But to be honest I don’t care about leveling my gantry, and because the stress for those motors is negligible, I leave them the way they are.
No problems at all.
Basically from my machine engineering background I know that in moving cnc setups the stiffness of the machine is of utmost importance. Since our 3d printers are often built with extruded aluminium profiles, sometimes of untraceable origins, those machines suffer greatly from issues related to stiffness, perpendicularity and stability.
All these “shortcomings” can be overcome with fancy trickery such as accelerometers and input shaping. That is all very good, but I am trying to see if I could follow the principles of for instance a milling machine build.
Heavy, lots of weight in the bottom. Least weight in the top. And the bed slinging system to be beefed up so the motors will have no trouble with the weight and be stable even at higher speeds.
the 2 bed rails to be hi-win 15 with 4 carriages on an aluminium milled base.
The x rail, single, is hi-win 25 with one carriage. The z axis will be trapezoidal screws, one on each side. But I might change them later to one ball screw. The z rail is also a hi-win 15 with one carriage. The z rail is held by a solid block of milled aluminium billet and is mounted on the 10 mm thick base plate. 4 steel legs under that and the total weight is already almost 30 kg with most of it under the bed. As I do not have a frame that surrounds the machine, I do not have any stability issues related to those extruded profiles. It simply cannot move with the little weight that is on the x axis carriage. So as the print goes higher, nothing changes in the dynamics of the machine except the weight of the printed item itself on the bed. In most other machines the weight distribution during printing changes, and is highly dependent on the stiffness of the frame to keep it reacting the same as it was at the start.
I am keeping in mind that I might have to change the belts for ball screws, since at higher speeds with more weight on it those belts will likely reach their limits. We will see. As I design and build all of it myself I can change things later. I am not dependent on off the shelf items except for electronics, motors and linear bearings and rails.
This machine will be running on BTT octopus V1.1 with 8 drivers, so I have a few spare. It will have a BTT touchscreen, raspberry pi 4 sbc and the bed heater is a silicon mains powered 750 watt mat.
Still a lot to learn before it can fire up. All this klipper stuff with configs and such and electronics have never been my main interest. I am forcing myself to jump into the pool and learn to swim while trying not to drown. All the rest I am comfortable with. If it works the next step will be a multi colour head on it, one direct drive extruder and the others via one hotend and divider from bowden tube extruders.
But that is even further in the future.
