Nimble + 1.2mm Nozzle - extruder stalls unless Amps increased 4x

Kodachrome

Ive just (as in today) swapped over my E3D Revo Obxidian 0.6mm nozzle to a Revo High Flow 1.2mm nozzle and tried some prints. All ended up looking way under extruded, like the filament snagged then released over and over.

I was watching it print on the 3rd test and noticed the Extruder Stepper stops moving and makes a buzzing sound. Sometimes, though not always, this was accompanied by a console error:
"Warning: Driver 3 warning: phase B may be disconnected"

I thought the Stepper itself maybe was a bit undersized for the new nozzle or just defective so I fished out a larger unit from my bag of parts (an old Wantai 42byghw811p2, 2.5amp rating, 4200 g-cm holding torque). Came with a new cable, no change to issue.

Exactly the same issue. I then tried slowing down the extruder limits set in Duet, no change.
I then swapped out the Zesty Nimble for a second identical Nimble I used to use. No Change.

Finally I tried increasing the amps from the tiny 500ma that Zesty insist you should limit the Nimble (v1.2) to, first to 650ma, then 750ma then finally at 2000ma. At 2amps... everything just works fine.

But this is 4 times the guideline Zesty advised (and what was working beforehand)??
https://zestytechnology.readthedocs.io/en/latest/nimble/nimblev2/tuning.html

Should I care about this, is there some problem somewhere or is this what is required to drive a larger nozzle (hard to understand, its so big there is little resistance)?

~~EDIT: Ive tested down to 1200ma now and appears to be working, so the min value is somewhere between 750ma and 1200ma. Still way outside of spec though.~~ Further testing shows the required figure is indeed around 2000ma, maybe 1900ma or something.

T3P3Tony

@Kodachrome are you asking for 4x the flow rate given the bigger the nozzle? I.e. if you swap out the nozzle but keep the flow rate the same as a test does it work as before (I expect it will). for the same movement speed and layer height going from a 0.6nozzle to a 1.2mm nozzle is an increase of almost exactly 4x the surface area of the nozzle so if you have not dropped the flow rate you are asking the extruder to push 4x the filament through for the same movement speed.

(0.30.3 pi = 0.28), 0.60.6pi = 1.13) , 1.13/0.25 ~=4

oliof

might also need to increase temp to help the filament flow

Kodachrome

Well yeah, but while the flow rate is higher, the effort to push it is lower. Like it's really easy to push it by hand on the 1.2mm vs the 0.6.

I thought these would kinda negate each other. I did try slowing the print significantly though too, to no avail. Also the Zesty instructions make no reference to flow rate or nozzle size at all, only that the torque it generates so huge you need to start at 500ma and drop it from there till you get clicking sounds.

I had also tried ramping the temp way up, which is what I thought would solve it but made no difference.

T3P3Tony

@Kodachrome AFAIK the key limitation is the melting of the filament, hence why high flow systems have longer melt zones, or metal elements in the melt zone (like CHT) or combinations of those. It is easier to push by hand, because you are comparing the filament input - i.e. the same flow rate.

The nimble (from memory) has a really high gear ratio compared to other extruders so that could be part of it. Is it potentially the acceleration to higher speeds that is being restricted, rather than the top speed itself? worth doing some static extrusion tests at varying accelerations and top speeds to get a better handle on the system Alternatively i there a downside to running at a higher current?

Kodachrome

@T3P3Tony Sensible stuff, though your last paragraph has me thinking, while its not actually acceleration, it actually only craps out on sustained straight lines (ie top speed, sustained for a few seconds). But its a crazy looking thing when going (with the amps high), the filament reel is just spinning constantly, it is actually quite impressive. There is also a crunchy-creaking sound coming from the Zesty, but I think thats the filament which is a quite hard PLA (branded as High Speed PLA+).

As to the drawbacks, heat in the Stepper is one, which may eventually cause its to also shutdown when over heating and also this poorly explained logic (IMO) from Zesty themselves:

Because the Nimble has so much torque available, you can/must run a much lower vref for your stepper than normal. This also helps getting the pulses across properly as you are not fighting decay in the pulses caused by too much current.

Kodachrome

To reiteration, the key issue in my case isnt melting of filament, its an electrical issue with the Stepper, the Zesty isnt skipping or hitting resistance with the filament. The Extruder stepper itself visually stalls to a standstill and "screams" on fast moves with extrusion. I dont know if this is a Duet Driver or Stepper motor issue. Further testing is showing that even 2200ma isn't enough.

I will try a 3rd stepper motor soon, ordered a SteppersOnline model, which has about 25% more holding torque.

engikeneer

@Kodachrome you are facing a melt rate problem. Just it's manifesting itself in the fact that the back pressure caused by your high volumetric flow rate is too much for your extruder to overcome, hence it skips steps.
If you increase the extruder torque, you probably will manage to get to higher flow rates. However, you will probably soon start running into problems with poor layer adhesion etc because you the filament simply isn't being properly melted through to the core because you're pushing it through so fast.

Have you tried calculating the volumetric flow rate you're actually demanding and compared to the figures e3d advertise?

Have you also tried increasing the nozzle temp? When you try to do long sustained extrudes, the nozzle temperature will drop because so much heat is going into the filament. Lower temp then means poorer filament melting and increasing back pressure for your extruder.

Equally, higher nozzle temps mean the melted filament is less viscous (more runny) so will flow better at high extrusion rates.

Finally, at high extrusion rates you are needing both high torque and high speed. Generally stepper motor torque drops off at higher speeds, especially if you are only running 12V. I can imagine that the mechanical losses in the nimble cable and gearing also increase with both speed and torque, hence you may be needing the motor to do even more than you think...

Kodachrome

@engikeneer
I'm really confused on why both of you are mentioning melt rates and temps. The issue isnt an Extruder skipping problem, I said the Stepper motor driving the extruder stalls. The stepper stops spinning (well, it vibrates slightly) and the stepper motor (not extruder) makes high pitched noises. If I just had the basic and known clickly-skippy extruder noise then yeah, whack up the temp (which I tried but as I said, its irrelevant due to the fact the stepper motor isnt feeding the extruder anything when its stalling out!). But I suspect this has nothing to do with the extruder actually. I am running at 24v. And as I said, the problem is "circumvented" with a now confirmed 2150ma (of the 2500ma the motor is rated at). But this makes be uncomfortable due to high heat of stepper motor and the fact its so way over the top of Zesty doc's (which dont mention any nozzle or flow specs).

Did you misread me or did I misread (misunderstand) you both?

EDIT: Just found out that Zesty, who I assume are now out of business based on having no stock for a year now, once made a High Volume version of the Nimble, the Zesty Giddy. Never heard of it before right now!?
https://zesty.tech/products/the-nimble-giddy

engikeneer

@Kodachrome doesn't matter whether it is the motor skipping steps because it doesn't have enough torque or the extruder gear slipping against the filament. Fundamentally the extruder system cannot push the filament through the nozzle at the rate you are asking. Apologies I have kind of used the terms interchangeably

Kodachrome

@engikeneer said in Nimble + 1.2mm Nozzle - extruder stalls unless Amps increased 4x:

@Kodachrome doesn't matter whether it is the motor skipping steps because it doesn't have enough torque or the extruder gear slipping against the filament. Fundamentally the extruder system cannot push the filament through the nozzle at the rate you are asking. Apologies I have kind of used the terms interchangeably

Ok, cool, so it will be interesting the change when swapping out the 4200g-cm (at 2.5amp) torque Wantai 42byghw811p2 for the StepperOnline Nema17 rated 5900g-cm (at 2.0amp) stepper. On paper, thats maybe closing on 50% more torque for a given Amp rating (since Im not running the Wantai at the full 2.5amp its not actually pushing 4200g-am torque).

droftarts

@Kodachrome to me it sounds like the stepper motor is being maxed out by the number of steps sent, which is indicated by the screeching noise. This is common in high inductance motors, which are high torque and high current, but often can’t keep up with the step rate demanded. You said the nimble has a high gear ratio, so the motor is spinning fast. Can you post your config, and the speed you’re trying to run the extruder at? If you can post the nimble motor specification, that might be useful too. You can check motor performance using the back-EMF calculator: https://www.reprapfirmware.org/emf.html

You can reduce the step count by reducing the microstepping, generally.

Ian

soare0

@Kodachrome said in Nimble + 1.2mm Nozzle - extruder stalls unless Amps increased 4x:

I'm really confused on why both of you are mentioning melt rates and temps.

Because increasing the nozzle diameter means increasing the cilinder surface, wich means higher friction with the walls. As the fluid movement is kind of turbulent, this is even worse, when printing at high speeds. So, even if the pressure needed is lower because of larger nozzle, you have enough other sources of pressure, wich may cause troubles, when printing at larger flowrates. So I assume that this is why it needs higher current.
This is even more so, on sustained extrudes, as engikeneer pointed out.
If the motor is properly built, in specs regarding current and external temperature, this is it, it will be working hot. Even very hot. Stepper life, hehe...
Eventually, if you really needs top speeds of your gear, wich is almost always a bad idea, you may need to cool down your stepper a little, by a method available/possible in your setup.

Stalling problem is another issue, and it may be that droftarts may be right on your needed answer.

Kodachrome

@droftarts

Well that certainly does sound like a promising lead!

The Zesty is a remote drive extruder, my V1 is has a 30:1 multiplier on the torque input.
I found the screeching stall happens when exceeding 32mm/s on a sustained (ie more than second) extrusion on the 1.2mm extruder and 0.7mm layer height. I have a 1000w 24v PSU.

Zesty recommend the following ideal stepper specs:

--
Low rotor inertia
This allows it to start and stop rotating fast.
A value lower than 15 g-cm2 is desirable.
This often isn’t shown on the specification sheet, In this case an overall weight of the stepper of 120 to 150 grams is way to get something close to the target.

Low inductance
This means that the currents set up by the pulses are not too strong.
This allows the stepper to spin fast without the pulses interfering with each other.
Aim for a value of 3.5mH to 4.5mH

--

Now, where this interesting is the spec's of the two steppers I tried.

My original (supplied as part of the Anycublic tower motor I think) Wantai 42byghw811p2 has a Rotor inertia: 68 g⋅c㎡ and an Inductance of 1.8 mH
The replacement StepperOnline motor, which in practice was no better at all, has a Rotor Inertia of 105gcm2 and an Inductance of 3.0 mH.

So basically the StepperOnline motor is much worse for this application, lol. Im pretty sure in my junk box I have a half size Nema17 that was part of a geared setup (it had a little gearbox on the end of it) that most likely better than both of these Steppers.

Kodachrome

@droftarts

Alright, an update and progress. I tried two smaller half height Nema17 steppers from my junk box. I also isolated a piece of movement from the actual print Gcode and made it into a macro. This was the most "offending" section, the last two fast extrusions back to back.

Macro:
G0 F2500
G1 X39.245 Y28.111 E12.53808
G1 X38.845 Y30.238 E.65933
G1 X11.598 Y57.486 E11.7389
G1 X39.245 Y28.111 E12.53808

This piece of GCode causes stalls on both the full size, high torque Stepper motors (the Wantai and the StepperOnline), Ie both fail, no matter the current (amps).

Conversely both the smaller Steppers could complete it with the following settings:

Working with 17HS3001-C5X (without gearbox, so its now a low torque direct motor)
M906 X1250 Y1250 Z1250 E1615 I25

Working with unnamed (it literally only has "XY" written on it) half height stepper
M906 X1250 Y1250 Z1250 E1545 I25

So basically the cheap ass, no name small Stepper works the best in this scenario. Im going to look for a better option online, annoyingly not mentioned in the Zesty Manuals but is on their old product page is a specific recommendation: https://www.omc-stepperonline.com/nema-17-bipolar-1-8deg-16ncm-22-6oz-in-1a-3-7v-42x42x20mm-4-wires-17hs08-1004s

Kodachrome

Im now testing an LDO NEMA 17 motor OverTheMoon LDO-42STH25-1004AC which is a High Temp High Speed pancake motor.

And I think I found the real root cause, aside from not-ideal stepper choices, I was using a Jerk speed of 200 mm/min, which is a value I found online as a recommended "extreme" setting. The default value is 40mm/min. When I dropped this down to 100mm/min and combined with the new light and fast stepper I can now achieve a pretty impressive 400ma current on the Stepper driving the Nimble (and still at high speeds).