Remote direct drive extruder question

Hello every apology in advance if this may seem like a stupid question but not long ago I came across remote direct drive extruder and have really like the idea and am considering converting my cr10 to one but I am wondering why are all of them always geared? Can't you have a 1:1 setup?

... you'll need much more force/torque, than a small motor can apply -- so high gearing factors to get more torque ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

The extruder motor on my cr-10 is a nema 17 JK42HS40-1004A-02F would that still be to weak? Another question exactly how much more torque is needed on the typical setup? is it just 2 or 3 times more or are we talking about a 10 time plus factor here

... more into 10x or plus ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Remote direct drive extruders use high gearing and smaller, less powerful motors. The reason for that in my humble understanding is because of the motor speeds involved you need a lower moment of inertia to prevent step loss in high velocity direction changes.

The gearing involved in 'remote direct drive', which is usually a worm gear, is used to mitigate the effects of mechanical variations that come from using a flexible shaft (torsional/rotational flex). Because of the required gearing, you use a smaller motor that can handle the inertial requirements and the loss in torque is reclaimed through the gearing itself. So the gearing kind of serves a dual purpose in this application.

I'm not a mechanical engineer though, so if anyone more formally educated on the subject cares to come in and point out the flaws in my understanding, I would appreciate it.

Yea that is what I seen they often use pan stepper, not full-size steppers. I was hoping to still use the full-size stepper that comes with my cr-10 I think it would have close to double the torque of those pan motor. So the gearing also works as a damper? Hmm, I guess that is why flex3drive and Zesty Nimble use plastic gears and not metal. Now if a smaller mass is required to limit how many steps are lost which requires the gearing to get back the torque then why not start off with a bigger motor that has more torque which would allow you to still use 1/1 gear ratio gear for damping.

When I used remote extruder (long bowden) then I needed torque of about 0.6 - 0.9 Nm. My full size Nema17 stepper had torque of 30 Ncm at the current I used (about 1.5A) so I used gearing with about 2.5 ratio. I was using Ø 0.5 mm nozzle and printed at about 120 mm/s, 0.2 mm layer height. The longer the bowden, the smaller the nozzle, the bigger the printing speed, the higher the layer height ... all that needs more torque on the remote extruder.

Interesting a 2.5 ratio is closer to what I would like. ^^What was the print quiality with your setup?^^
My 42HS40-1004A-02F spec is:
Item Specifications
Step Angle 1.8°
Temperature Rise 80ºCmax
Ambient Temperature -20ºC~+50ºC
Insulation Resistance 100 MΩ Min. ,500VDC
Dielectric Strength 500VAC for 1minute
Shaft Radial Play 0.02Max. (450g-load)
Shaft Axial Play 0.08Max. (450g-load)
Max. radial force 28N (20mm from the flange)
Max. axial force 10N
Holding Torque 4.8kg. Cm
I normally print from 60 mm/s to 100 mm/s with a .4 nozzle (considering going bigger to deal with the cr-10 print bed size). I am looking at cutting the flexible shaft I have from an old Dremel that I never used down to 400mm long.

What I mean by low moment of inertia is a characteristic of the motor, not mass of the gears at the hotend. Low moment of inertia allows the motor to make rapid high velocity direction changes without lost steps.

If you gear the flex shaft at the motor to reduce the motor shaft speeds involved, you could use a full sized motor. You don't really need high torque motors for this because the gear ratio multiplies the motor torque at the 'business end' of that style of setup. I'm not sure how much gearing you'd need at the motor, and that would be dependent on the gearing at the extruder.

I have my own remote direct drive extruder I custom designed that uses 1:24 gear ratio worm drive. I'm still prototyping/testing but testing has been very promising.



Here's a video of the first time I fired it up on my printer.
First run

Edited 2 time(s). Last edit at 08/24/2020 12:39PM by obelisk79.

I am aware that you talking about the motor that is why I said if high gear ratio is mainly used just fix the issue of having to use a lighter motor with less torque that needs to speed faster. Why not just start with a motor with more torque which doesn't need to be run as fast, therefore, the weight won't be much of a problem to cause you to worry about missing steps and you can use less aggressive gearing. My current Bowden setup has no gear at the extruder end the stepper is directly attached to the motor. 
What motor are you using on your remote extruder?

The worm gearing is to compensate for twist in your flex shaft not to allow the use of a smaller motor. The smaller motor is necessary to meet the inertial needs of the high gear ratio at the hotend. I use a cheap pancake nema 1.8deg motor with decent specs for the purpose.

Can you elaborate on how the worm gear helps with the flex shaft twisting?

I'm going to use some hypothetical numbers to help explain. Every flexible driveshaft will have different characteristics.

Let us presume your have a flexible drive shaft which can twist up to 90 degrees at one end compared to the other. This is known as torsional deflection.

90 degrees of 'sloppy' rotation can be cause for a great deal of inaccuracy in a positioning system. But when you divide that up using a reduction gear of 1:30 now you've reduced the effective deflection down to 3 degrees, you've also reduced extruder step distance which provides an even further reduction in impact to extrusion accuracy.

The drawback, is that your motor now has to step/pulse/rotate 30 times as fast for the same amount of extruded plastic, these higher speeds also result in high speed direction changes for functions such as retraction. The step motors ability to deal with these forces (from my humble understanding) is called its moment of inertia. Smaller steppers have lower mass in their rotors (thereby moment of intertia) allowing for those high velocity direction changes. I thought the same as you, higher torque = better, right? When it comes to these types of remotely driven extruder setups that just isn't the case. I run a smaller weaker stepper at lower current and it far outperforms the much larger 84oz-in motors I attempted to use at first.

Edited 1 time(s). Last edit at 08/24/2020 02:24PM by obelisk79.

If you wanted to do something like a 1:1 gearing on a remote direct drive type setup, you'd need a gimballed motor with a telescoping spline shaft to drive the extruder hob, that is likely possible, but would come along with its own uniquedesign/engineering challenges.

Edited 1 time(s). Last edit at 08/24/2020 02:32PM by obelisk79.

Quote
obelisk79
I'm going to use some hypothetical numbers to help explain. Every flexible driveshaft will have different characteristics.

Let us presume your have a flexible drive shaft which can twist up to 90 degrees at one end compared to the other. This is known as torsional deflection.

90 degrees of 'sloppy' rotation can be cause for a great deal of inaccuracy in a positioning system. But when you divide that up using a reduction gear of 1:30 now you've reduced the effective deflection down to 3 degrees, you've also reduced extruder step distance which provides an even further reduction in impact to extrusion accuracy. 

The drawback, is that your motor now has to step/pulse/rotate 30 times as fast for the same amount of extruded plastic, these higher speeds also result in high speed direction changes for functions such as retraction. The step motors ability to deal with these forces (from my humble understanding) is called its moment of inertia. Smaller steppers have lower mass in their rotors (thereby moment of intertia) allowing for those high velocity direction changes. I thought the same as you, higher torque = better, right? When it comes to these types of remotely driven extruder setups that just isn't the case. I run a smaller weaker stepper at lower current and it far outperforms the much larger 84oz-in motors I attempted to use at first.

Oh ok, you accounting for flex shaft error thru the gear ratio. The way you described it previously I thought you were using gear to dampen the movement of the shaft itself or something. The motor ability to deal with rapid change is related to the motor spec it can be heavy but if it is powerful it can still change speed rapidly but the same goes the other way less power but light = rapid change. 

Quote
obelisk79
If you wanted to do something like a 1:1 gearing on a remote direct drive type setup, you'd need a gimballed motor with a telescoping spline shaft to drive the extruder hob, that is likely possible, but would come along with its own uniquedesign/engineering challenges.

I think the only setup I have seen like this was in an Ultimaker some time ago.

Here are some really old parts which were printed with the remote extruder and long bowden (about 80 cm). The yellow part was painted with acetone and printed without support. The white part does not have any post-processing. The parts are small, so the maximum printing speed may have been lower to keep the minimum layer time ... probably at 20 seconds. If you want so see some later parts then search this forum for my posts, I posted some detailed pictures when we talked about about patterns on the vertical walls and ringing. But those were probably printed with flying extruder (bowden length about 7 cm).


Edited 1 time(s). Last edit at 08/25/2020 07:27AM by hercek.

Well then today I learned about flying extruder didn't know that was a thing. I quite like the idea. While googling remote extruder I also came across this video another idea that I find interesting.

[www.youtube.com]

You want to use a flying extruder when you intend to print parts which have a lot of isolated isles in their horizontal cuts (i.e. in one layer). That leads to lot of retractions. Each retraction adds some error to the amount of extruded filament especially at high printing speeds. The errors are bigger as bowden gets longer. My first flying extruder had about 7 cm long bowden, I shortened it to about 4 cm later.
Also if you are designing your extruder then take care so that it is speedy. It needs to retract and un-retracf filament in at most 0.1 s. For that you will need e axis acceleration around 7000 mm/s^2 or more, and extrusion speed of around 20 cm/s (is is not possible to extrude filament that quickly, but you want to be retracting/un-retracting that quickly). You will still need retraction length of around 5-6 mm to get good results (even with a short bowden and flying extruder). IIRC, I used about 9 mm with the 80 cm long bowden.

I'm quite pleased with mutley3d's Flex3drive

I would like to try a flexible shaft approach. I just do not have time now.

The problem I see is how it will behave under high speeds. The shaft torsion results to significant errors. One needs a lot of gearing to mitigate them. The gearing slows down the final retraction speed. Based on the quick on the corner of a napkin calculation it did not look like a clear winner.

What E-axis accelerations, and speeds do you run? What is your retraction length? What printing speed and accelerations do you typically use?

Sorry, I missed your reply. I don't check in very often.

The flex3drive has a 1:40 gearing.

I have to double check the acceleration I don't remember what I usenow, speed 40-50 mm/s retraction length 0.5-2 mm depending on filament but has not done any real tuning. Printing PETG without stringing, with nylon I get some stringing, but I think that is mostly due to not being dry enough.

Printing speed usually 40-60 mm/s, since I usually print relatively small things I have to dial back because of my (lack of) print cooling.

I have not done any extensive tuning of retracts, hardly any at all, But before going to 24V supply for the stepper drivers(used 12V before) I was limited to perhaps 20mm/s. I think using a small stepper with low inertia would be better, with the gearing not much force is needed, and would allow very fast retractions.

Thanks, if you find the retraction acceleration then please let me know.

Acceleration 250 mm/s2