The current MK3S extruder body is 0.4 mm loose(r)  

Re: The current MK3S extruder body is 0.4 mm loose(r)

Has anyone done a measure of the clamp dimensions, because it looks like 12.2mm diameter on the R4 .step files, and 12.1 on the older R3?

Re: The current MK3S extruder body is 0.4 mm loose(r)

There are some interesting things in how the extruder motor and the hottend have been moved between the R3 and R4 parts.

R3 E-Axis

R4 E-Axis

The motor has been moved 2.5mm towards the Z Frame and the hotend has been moved upwards by 5.5mm. As the X-Ends and the Z-Axis tops have not changed between R3 and R4 this effectively gives an increase in the Z-Axis of 5.5mm. However as can be seen below from 1_75mm_MK3S-EINSy10a-E3Dv6full.h the Z-Axis is still 210mm rather than 215mm as physicaly posible with the R4 parts.

// Home position
#define MANUAL_X_HOME_POS 0
#define MANUAL_Y_HOME_POS -2.2
#define MANUAL_Z_HOME_POS 0.2

// Travel limits after homing
#define X_MAX_POS 255
#define X_MIN_POS 0
#define Y_MAX_POS 212.5
#define Y_MIN_POS -4 //orig -4
#define Z_MAX_POS 210
#define Z_MIN_POS 0.15


Regards,
Martin

Re: The current MK3S extruder body is 0.4 mm loose(r)

I think that’s just an oversight. What about diameters at the e3d throat?

Re: The current MK3S extruder body is 0.4 mm loose(r)

I think that’s just an oversight. What about diameters at the e3d throat?

I've checked, the throat diameters are fine. From bottom to top: 16.2 mm, 12.2 mm, 16.2 mm and 10.0 mm.
There is a slight difference between the geometry of the two, because R3 was made for sliding a heatbreak in, while R4 grips it with a lid that is screwed on top. R4 is even a bit tighter due to that difference. But the issue here is the bottom opening that does not hold the bottom part properly, so after a few collisions with a grid infill while travelling over a big print, the neck grip will weaken.

Currently there is still absolutely no feedback from Prusa regarding this issue, but I am pretty confident in my estimate, so no matter what they eventually decide to do with it, I'll be using my own tighter mod. 🙂

Re: The current MK3S extruder body is 0.4 mm loose(r)

Isn’t r3 12.1mm diameter?

Re: The current MK3S extruder body is 0.4 mm loose(r)

Isn’t r3 12.1mm diameter?

I've measured 12.1 - 12.2, depending whether you measure in X or in Y axis. It might be an error due to approximating a circle with a polygon. Still, it is in the range of printing tolerance, and when you tighten it with a screw, it will be leveled. Either way, that is nothing to be worried about.

Re: The current MK3S extruder body is 0.4 mm loose(r)

Depends. I usually run 0.15 for a tight, but free, fit, and the Prusa hits that without an issues with default profiles.

12.1 is .05mm tolerance (per side), which is ok for a clamp fit.

12.2 is 0.1. That's starting to get pretty loose for a clamp fit.

and at 12.3, or 0.15mm per side, it's not likely to be much of a clamp at all.

On my modded extruder I'm running 12.1 and it's only a solid clamp under pretty high screw tension.

Re: The current MK3S extruder body is 0.4 mm loose(r)

Depends. I usually run 0.15 for a tight, but free, fit, and the Prusa hits that without an issues with default profiles.

12.1 is .05mm tolerance (per side), which is ok for a clamp fit.

12.2 is 0.1. That's starting to get pretty loose for a clamp fit.

and at 12.3, or 0.15mm per side, it's not likely to be much of a clamp at all.

On my modded extruder I'm running 12.1 and it's only a solid clamp under pretty high screw tension.

You convinced me, so I've updated the model https://www.thingiverse.com/thing:3428292 with the tighter +0.1mm diameter tolerances on the neck.

Re: The current MK3S extruder body is 0.4 mm loose(r)

My upgrade set finally has arrived. I have compared how the heatbreak fits the original R4 body, versus my modified R4x1, with tighter neck and nozzle-side opening. It made me even more sure that the tighter version is the way to go:

1. The heatbreak easily fits the original R4. There is no noticeable slack.
2. The heatbreak is also easily fitted into the tighter R4x1, with no special force required to insert it, but you can notice the firmer grip.

So if the same part easily fits the tighter version, without forcing it into the place or bending the plastic parts, I am much happier knowing that I've chosen the tighter grip, since after some use (especially after hitting the grid infill while traveling across the print) it is very likely to loosen up.

Re: The current MK3S extruder body is 0.4 mm loose(r)

I'd really like to hear the answer of the prusa developer who designed that model. Didn't someone want to report back here with his answer?

Re: The current MK3S extruder body is 0.4 mm loose(r)

I'd really like to hear the answer of the prusa developer who designed that model. Didn't someone want to report back here with his answer?

While I'd be curious to hear their response, I wouldn't blame them for making a decision not to answer every question about the design. If there are, in fact, actual reported problems with the design as-is, by all means they should make revisions. If it's working without issue -- and I've yet to read up on an issue or even seen a comparison print -- then it's working as designed and what was delivered is what was promised. I'm willing to bet they have more pressing priorities.

Now I'd really like to talk to the Nissan engineer that decided that having a dashboard assembly that protrudes into my knee is a good thing, but they won't answer my emails either, and I paid a hell of a lot more for that damned car!

Re: The current MK3S extruder body is 0.4 mm loose(r)

While I'd be curious to hear their response, I wouldn't blame them for making a decision not to answer every question about the design.

I totally agree with everything you say. But a simple: "We did this intentionally, this did not happen by mistake" would help to clear up things and not cost them too much time.

Re: The current MK3S extruder body is 0.4 mm loose(r)

...
But a simple: "We did this intentionally, this did not happen by mistake" would help to clear up things and not cost them too much time.

This is just my crystal-ball guess about what happened, and how the issue will continue developing in the near future:

Since the part was completely refactored, it is very likely that the engineer who modeled it just took wider tolerances, without thinking about the previous values. After a certain period of testing the part on their print farm, they noticed no issues, so they must have decided to release it as is. There is no guarantee that after some time the part won't start wobbling (nor that it necessarily would), even though that is much less likely to happen with tighter tolerances. But alas, they already printed and shipped a sh*tload of these parts, so making a correction at this early point in time would not make a good picture. So they would probably be waiting for a couple of months for the people to start reporting issues with the wobbling head (if that happens), and only then would they make a tighter release, solving the problem and making everyone happy. In fact, much happier than if someone just said: "Ok, you've invested 5 hours the last weekend to make this upgrade, but now you will have to do it all over again, since these tolerances might start causing some issues 2-3 months from now".

Justifying that decision to go with wider tolerances would just be a very obvious attempt of damage control, so they have probably decided to sit on it.

At least that would be my call if I was a responsible insider.

Re: The current MK3S extruder body is 0.4 mm loose(r)

My upgrade set finally has arrived. I have compared how the heatbreak fits the original R4 body, versus my modified R4x1, with tighter neck and nozzle-side opening. It made me even more sure that the tighter version is the way to go:

1. The heatbreak easily fits the original R4. There is no noticeable slack.
2. The heatbreak is also easily fitted into the tighter R4x1, with no special force required to insert it, but you can notice the firmer grip.

So if the same part easily fits the tighter version, without forcing it into the place or bending the plastic parts, I am much happier knowing that I've chosen the tighter grip, since after some use (especially after hitting the grid infill while traveling across the print) it is very likely to loosen up.

Just wanted to thank you for designing the R4x1, I also printed both R4 and R4x1 and decide do go with the tighter ones. So far, no problems whatsoever.

Re: The current MK3S extruder body is 0.4 mm loose(r)

It's worth noting that there doesn't appear to be any widespread print quality issue or anything that would seem to be caused by this, so it's probably a non-issue. You are, of course, perfectly entitled to fiddle with your printer all you like, but barring objecting benefit it would seem we're being a bit dramatic, no?

Re: The current MK3S extruder body is 0.4 mm loose(r)

I've been fighting with this issue for a while now - since before the R4 parts were ever released even. This is part of why I've been doing my own extruder designs, which go back to pre-R3 as well.

Comparing the R3 and R4 groove-mounts - they are the same (loose-fit) size, but if you go back farther and compare to say an original MK2 extruder body, the fit used to be tighter. I believe the change was made in the R3 release to make it easier to get the E3D in/out at an angle with the PTFE having to bend a bit to fit into its hole. In the older MK2(s) extruders it wasn't easy getting the V6 back out, and probably caused a fair share of bent/broken heat-breaks. It got a lot easier with the R3, but unfortunately it also doesn't grip tightly enough resulting in wiggle down at the nozzle. Then they carried the new dimensions forward to R4, even though there's no reason for it anymore with the new design not needing the V6 to go in/out at an angle.

Having also put together a handful of these using both prusa-printed parts as well as my own - the problems seems to only really appear on user-printed parts. Both the R3 and R4 extruders I've assembled from Prusa shipped parts fit tight enough to not have an issue, while both R3 and R4 ones I've printed myself were so wobbly I didn't even try to mount them on a printer. I've gone back to the SCAD or STEP source files and verified they are loose by design, so the only explanation is that Prusa is over-extruding their parts. Possibly on purpose for strength, and possible that the loose fit in the groove-mount is also by design to compensate for that over-extrusion. I haven't bothered to compare the Prusa g-code files to my own to see if there's any indication of higher extrusion than I normally use. But I do print a LOT of functional parts for customers, and am confident that my slicer/printers produce parts that are dimensionally accurate.

RE: The current MK3S extruder body is 0.4 mm loose(r)

@martin-w15

Beautiful drawings! I don't suppose they are available as PDFs somewhere, they are wonderful references for the dimensions if printed at the correct size.

Thanks,

-Lars

RE: The current MK3S extruder body is 0.4 mm loose(r)

@martin-w15

Hi

might I ask, what application did you use for the drawings?

Solved in R6

This issue is fixed in R6, as all tolerances are set to the values proposed here. R6 introduces also some other improvements, so we can close this issue as solved.