TAZ 4 Extruder Weight Reduction?

So I am trying to decide between a Taz 4 and an Ultimaker 2, and one of the (few) advantages of the UM2 is that it’s bowden head is much lighter, so it is able to retain accuracy at higher speeds, resulting in overall higher quality prints.

I am aware the Taz4 is a direct drive extruder and thus will never be as lightweight as the Ultimaker 2’s… but I am curious if:

  1. what (if any) modifications are possible that could reduce the overall weight of the extruder, and

  2. would there be an advantage to such weight reduction (in terms of a significant difference in print quality at higher speeds?)

You could use a bowden on the TAZ4. SInce we use common aluminum extrusions you should be able to mount the bowden extruder mount on the side. You’ll need to print a different mounting plate, Bowden extruder body/mount and potentially replace the filament guide tube with a more appropriately sized Bowden tube.

Do you there could be significant advantage to such a lighter print head? Would it be difficult to calibrate since I assume the firmware and various profiles would be “thinking” that the extruder is heavier than it now is?

We think the disadvantages to a bowden (where the motor is off the print head) over what we have. It doesn’t do retraction as well as it has to retract over a much longer distance, and not as many materials are going to be able to feed through the machine. If we thought bowden was better, we’d just be using those (we have built bowden TAZ in the long past).

I wouldn’t get too obsessed over the weight of the print head, unless you are trying to set a world speed record or something.

-Jeff

Your last post surprises me Jeff, I thought the whole point of open source was to improve 3D printing, using the community to source innovation at an accelerated pace :wink:

For me one of the biggest problems with 3D printing is the long wait times… any improvement in this regard is welcome.

I’m also looking to reduce the weight of the head, especially in with the development of a dual head. I find with the current weight I get speed wobbles unless I print quite slowly.
I’m looking at a smaller stepper with a different gearing, or a very small stepper with a bowden attached to another stepper, in an attempt to gain the advantages of both setups.

I’m also looking at reducing bed weight and increasing stepper size there.

Anyway, not getting on your case, I just thought I’d add my thoughts :slight_smile:

I thought the whole point of open source was to improve 3D printing, using the community to source innovation at an accelerated pace



I find with the current weight I get speed wobbles unless I print quite slowly.

The UM-TAZ difference for the most part here is not speed but acceleration. UM2 is set at 3000, I believe, and TAZ4 is at 500. I think this thread can tackle both of your comments. :smiley:

https://forum.lulzbot.com/t/linear-bearing-backlash-big-x-axis-issues/966/1

I’m also thinking about a dually bowden rig. But given that Lulzbot has a fleet of production printers with direct extrusion the tradeoffs to go to a bowden must not be that great.

I don’t think making the X axis more zippy (higher acceleration) will give much benefit without doing the same for the Y axis. … And, since the Y axis acceleration is limited by the mass of the bed, I think lightening up the extruder carriage could only help if its mass already exceeds that of the bed… Maybe in cases like the (flexi)dually… maybe.

You could make your own flexidually that uses a bowden for one head (for stiff filaments) and a low-mass stepper direct for the other (stiff and TPE). If it could be made light enough, it could go on a CoreXY design and match the UM2 quality while offering a broader range of filaments (hint hint Lulzbot :slight_smile: ).

I don’t think making the X axis more zippy (higher acceleration) will give much benefit without doing the same for the Y axis. … And, since the Y axis acceleration is limited by the mass of the bed, I think lightening up the extruder carriage could only help if its mass already exceeds that of the bed

A very true statement–with one or two caveats.

1)Technically there is no acceleration limit on a TAZ or any open source printer. I could set firmware parameters to run at 500mm/s with an acceleration factor of 9,000–hummingbird like movement. The problem with that is that components will obviously start to break due to the forces being so great that both axes become slide hammers. Looking at Configuration.h the X&Y acceleration limits and given that the entire X carriage weighs 0.86kg the max force generated is roughly 770N. For reference I have a physics book that says a baseball bat hitting a baseball generates ~3400N–probably not Hank Aaron’s bat.

426#define DEFAULT_MAX_FEEDRATE          {800, 800, 8, 50}      // (mm/sec)
427#define DEFAULT_MAX_ACCELERATION      {9000,9000,100,10000}  // X, Y, Z, E maximum start speed for accelerated moves. E default values are good for skeinforge 40+, for older versions raise them a lot.
428
429#define DEFAULT_ACCELERATION          500    // X, Y, Z and E max acceleration in mm/s^2 for printing moves
430#define DEFAULT_RETRACT_ACCELERATION  3000   // X, Y, Z and E max acceleration in mm/s^2 for retracts

2)Acceleration is an all encompassing parameter for the X&Y axes. So yes the heavier axis would be the limiting factor except when the heavier axis remains in the most stable area of the machine and the other axis goes up and down the frame. The print head can be way up in the air where the machine will experience lots of frame flexing. Plus the X-axis is connected to the Z-axis which will flex quite a bit given that some people say that the weight of the print head will make the stock X-axis rod flex a little. So making the printer head lighter should help with higher accelerations on taller prints.

Now that I’ve thought about all this I currently run my TAZ at 150mm/s with the acceleration cofactor at 1500. I have yet to print anything over 50mm tall with this setup but maybe I should see if this theory is correct.

Mark, I didn’t mean to discourage hacking away at it, go for it for sure. :slight_smile:

JJ,

Hmmm…the thing is you have one of the lead engineers for Lulzbot telling you that they tried the bowden setup and found it not worth it for listed reasons… hacking is great, but working for nothing isn’t… atleast consider his advice. I don’t have that much time playing with extruder setups per say but I am a designer and common sense tells me there’s downsides with bowden setup. To me, your best bet would be to stick with the standard xcarriage but maybe try to reduce some of the material used for the mount.