Notifications
Clear all

MK3S Mosquito integration  

Page 5 / 20
  RSS
Evan
 Evan
(@evan-2)
Eminent Member
RE: MK3S Mosquito integration

Here you go!

Prusament PETG Jet Black printed at 240/250C, 50% fan, 0.4mm E3D nickle-plated copper nozzle with 0.1mm layers. Sliced with Simplify3D. 

I hate hearing about your predicament. You've put in so much hard work for us, I'd hope you of anyone would be getting some great prints! New thermistor and heater sound like a sensible next step if you already have them around, though it still sounds a bit odd.

For what it's worth, my heater was somewhat loose, too. Err...At least until I started filling the holes up with the boron nitride paste. That stuff comes out fast if you're not ready for it. I pretty much filled the holes with paste before I added the heater, then dealt with some cleanup.

I don't have enough time for another print tonight, but tomorrow evening I'll try a treefrog with an E3D Nozzle-X instead of the nickle-plated copper. Its thermal characteristics should be closer to the Slice vanadium nozzle since they're both forms of hardened steel. I'm curious to see what it does to the print.

Ce message a été modifié il y a 5 years par Evan
Publié : 21/06/2019 2:51 am
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

Nice glassy finish. The underbelly overhang roughness is consistent with 0.1 mm layer. So  yes it is nice print result.

E3D nickle-plated copper nozzle <--- might be a big factor.

That's a lot more conductive than the vanadium Slice nozzle. I think my brass nozzle works OK with the sock, but it is just the vanadium nozzle that is a below usable performer. If you get a chance to do the Nozzle-X, that would be a closer match to the problematic setup here.

I'm going to tear down and swap in a different heater and thermistor later tonight. 

 

Publié : 21/06/2019 3:02 am
Evan
 Evan
(@evan-2)
Eminent Member
RE: MK3S Mosquito integration

Thanks! Still some tweaking to do, and I'll probably try swapping back from the Trinamics to the 0.9° steppers, but I think I'm finally starting to get the hang of this.

H'okay, dumb idea:

Reading through the thread, it feels like airflow cooling the heatblock is a pretty solid hypothesis. There are two sources of airflow, though: the part-cooling fan, and the Noctua extruder fan. I briefly considered switching to the Slice-supplied fan in lieu of the Noctua. The Mosquito even has direct mounting holes for it on the outer heatsink. That gives it a tighter fit, significantly limiting the possibility for airflow to spill out the bottom near the heatblock.

Now, that's a bad idea. Prusa-supplied Noctua runs at 5V, Slice fans run at 12V or 24V. (Correct me if I'm wrong.)

Instead, would it make any sense to tweak the BNBSX so the airflow goes directly from the fan through the Mosquito's heatsink, perhaps even having a part mount to those same screw holes in the outer heatsink? I don't have a good sense for how much spill there is in the current design -- it's already fairly tight -- but I could imagine the turbulence at the heatsink sending some not insignificant portion of the air downward. The wind tunnel in there is already going to have more pressure than the outside world anyway, and it'll be looking for somewhere to go.

There are probably some good reasons not to do that with the full extruder casing -- it could get really tricky to preserve V6 compatibility, for instance -- but you might be able to do a PoC with a part that fits between the fan and heatsink and funnels the air directly, possibly removing/replacing the hot fins on the inlet. It doesn't have to be perfect, the Noctua already has about 30% more airflow than the Slice fan anyway.

For that matter: bigger fan + more efficient heat break might itself be a source of trouble...

Like I said: probably a dumb idea, but at least now I can get some sleep without it still rattling around in my brain. 😉

Publié : 21/06/2019 3:32 am
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

Switching out the heater cartridge and thermistor didn't change a thing about how the vanadium nozzle is behaving. We can rule those out as being defective. 

 

Publié : 21/06/2019 8:55 am
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

Guys, could I ask you to print the tree frog using my gcode to rule out my print settings as the factor. It's a long shot, but I'm fairly perplexed at how poorly the vanadium nozzle is behaving here. Brass nozzle I can get working ok with a silicone sock. 

I want to know that my gcode is able to produce a glassy looking PETG tree fog with a Mosquito equipped with a VANADIUM nozzle. Could I be extruding extra fast? I don't think I am, but we'll know if my gcode causes someone elses vanadium nozzle equipped mosquito to misbehave.

Attached is a zip file with gcode and 3mf. Printing with the gcode will yield the most useful data.

Thanks for you help.

 

Attachment removed
 
PS, I should add that my startup gcode does a PINDA preheat sequence before it does bed leveling. It will move print head into center and appear to stall. It's simply heating the PINDA to temp.
Ce message a été modifié il y a 5 years 2 fois par Bunny Science
Publié : 21/06/2019 2:26 pm
james.hess
(@james-hess)
Trusted Member
Topic starter answered:
RE: MK3S Mosquito integration

Hi guy-k2,

Chasing mythical and opportuned rabbits can be very interesting.  Though I'm getting good to excellent prints (received the .8 Moons while on travel) with the Mosquito, I also don't fully understand the thermal properties and required extruder configuration parameters.  I need to print out several extruder assembly configs and test them in a static xyz state during min max filament flow.  I'm in full agreement with the threads focus on, "where is the fan (heat break / layer) air flow and how does it effect the heat block and nozzle temps at a known ambient room temperature.  Think I'm going to take a Mosquito sabbatical and install the x/y Moons first to see where it takes the machine's performance.  Hate to take the firmware out of stock configuration but you guys know the rest of the story.  I'll post some before and after Moons photos.  Any requests for the baseline gcode?  Intent is to use a Prusa sliced model.

Publié : 21/06/2019 2:53 pm
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

I just thought of one more thing to check. Maybe my mosquito heat break's copper base is loose from its heat block. That would dramatically shorten its melt zone. Then again, it is conducting heat well enough to grow the melt zone up into the heat break and cause a jam if I interrupt airflow to the heat break.

PSA don't do the combination of

1 Letting Mosquito grow its melt zone up higher by reducing airflow past the heat break. I did that while testing Evan's idea of airflow past the top of the mosquito)

2 Subsequently restarting with normal airflow (and consequently a smaller melt zone). Filament will be firmly fused to inside of heat break tubing. Normal airflow maintains reduced size melt zone and mosquito won't melt filament free. Bondtechs will grind away. Filament will not budge.

Solution was to completely block airflow to grow melt zone until I could advance filament with my fingers. Was a bit scary to intentionally cause melt zone to grow like that.

So, maybe my copper bit being loose theory is not true. Can't check now because in midst of long print (with a brass nozzle)

 

Publié : 21/06/2019 2:54 pm
james.hess
(@james-hess)
Trusted Member
Topic starter answered:
RE: MK3S Mosquito integration

Hi guy-k2,

Yes, print the tree frog is on..  Will tune the hotend to a +5C bias to compensate for the thermistor offset.

Publié : 21/06/2019 3:00 pm
holmes4
(@holmes4)
Estimable Member
RE: MK3S Mosquito integration

And I am doing the tree frog gcode exactly as-is. I'll be back in an hour or so with the results.

Publié : 21/06/2019 3:08 pm
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

Thanks! That will be very helpful. I don't think my print settings are at weird, but at this point I'm quite perplexed.

If your prints come out looking good, my next step is switching the part cooling fan to an R4 style.

BNBSX extruder design above the melt zone just extrude filament and can't be causing my vanadium nozzle issues. 

I got the Mosquito planning to set up one machine with it and a vanadium nozzle for higher temp, abrasive filaments. Gotta get it fully melting PETG before there is any hope of doing high temp materials.

Ce message a été modifié il y a 5 years par Bunny Science
Publié : 21/06/2019 3:13 pm
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

Looking again at the Mosquito's melt zone. It actually starts up at the copper segment. Slice's little fan hits just the cooling fins. The BNBSX is designed to channel Noctua airflow past the entire Mosquito above the heat block. When I tried to implement Evan's idea last night, I dropped airflow to both the copper segment and the cooling fins above the heat break. (Which gave bunnies an adventure in unjamming a Mosquito)

Bunnies must repeat the experiment with just the copper segment protected from Noctua airflow. That copper segment is probably a big portion of a Mosquito's melt zone, particularly if the nozzle is a low conductance material and doesn't effectively participate as part of melt zone. If the BNBSX is cooling that copper segment too much, it could also explain why PETG isn't melting despite the heat block being at full temperature. 

 

Publié : 21/06/2019 3:29 pm
holmes4
(@holmes4)
Estimable Member
RE: MK3S Mosquito integration

FWIW, I am using the recommended Sunon fan, not Noctua. Slice's fan doesn't come in a 5V version.

Publié : 21/06/2019 3:40 pm
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

That's the same size as the Noctua, not a tiny one like that from Slice, right?

One more thought. Those that seem to having the easiest time with Mosquito are on a R4 style body. That is consistent with needing to reduce airflow over the top of the Mosquito heat block. The R4 suffers an issue of reduced airflow in the lower portion of the plenum. That's why it tends to melt its hot fins and PINDA mount. Ironically, that is probably helping to keep the Mosquito hot enough to melt filament.

Publié : 21/06/2019 4:15 pm
james.hess
(@james-hess)
Trusted Member
Topic starter answered:
RE: MK3S Mosquito integration

Hi guy-k2,

First frog failed to print (not enough layer cooling)...  Started second print with R4 body, hotend 245, Bed 90, Fan 180 and is coming along nicely.  On the first print the front legs were curling upward on the leading edges.  Will post pictures when complete.

Publié : 21/06/2019 4:58 pm
james.hess
(@james-hess)
Trusted Member
Topic starter answered:
RE: MK3S Mosquito integration

Hi guy-k2,

Second print failed...  Again not enough layer cooling...  Will retry with R4 body, hotend 250, bed 80, Fan 255..

Publié : 21/06/2019 5:04 pm
holmes4
(@holmes4)
Estimable Member
RE: MK3S Mosquito integration

Frog attached using your gcode, and I think it looks great. Yes, the fan is 40mm. I would dispute that I have insufficient cooling with the Sunon, and the plenum is far larger than the Mosquito's tiny heat fins.

Attachment removed
Publié : 21/06/2019 5:04 pm
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

Thanks. This is helping tremendously. I think I am finally grasping what is happening and have a solution for the BNBSX - a modification of the hot fins to better isolate the mosquito heat block and copper portion of heat break. It will be nearly as tight as we do around the bottom fin of the E3D.

James is talking about the print parts cooling, not the heatsink cooling fan. When parts cooling is insufficient, geometry fall apart.

The Sunon is the other cooling fan (for the heatsink). That we think is pretty easy to be sufficient enough to keep the extruder from melting.

Problem on the BNBSX is really looking like the noctua (40 mm) is cooling the top of the mosquito heat block. BNBSX is a Bear derived design and retains the Bear airflow pattern. That keeps it from melting its hot fins and PINDA mount. Bottom of air plenum is well ventilated in the Bears. Current BNBSX mosquito hot fins don't close as tightly around mosquito radiator fins due to the geometry being blocked by support tubes and bolts. I'm going to add two mosquito helper pieces to close the gap. There will be very little air from the Noctua reaching the heat block once the helper pieces are added.

Going to be like this for the hot fins. It will be a couple days before I can test it, but based on all our combined tests, this is the most likely solution to get BNBSX and vanadium nozzle to play nice.

 

Publié : 21/06/2019 5:23 pm
holmes4
(@holmes4)
Estimable Member
RE: MK3S Mosquito integration

Sounds good. For print cooling I am using a Fystec fan, as the wires to my stock fan broke off the fan's circuit board. I don't think it runs any faster, though.

When I was using the E3DV6, my fan shroud was melting (printed in AmazonBasics PETG). The one I printed for the Mosquito mount is out of Atomic CF PETG, which has a bit higher temperature rating, and so far it shows no signs of softening. I don't recall ever having an issue with the PINDA mount but I did get some melting of the PINDA cable jacket when it was too close to the E3D heatsink. With the Mosquito both the PINDA cable and the part fan cable are far away from the hotend (and I have more air moving through with the Sunon.)

Publié : 21/06/2019 5:43 pm
Bunny Science
(@bunny-science)
Noble Member
RE: MK3S Mosquito integration

I went ahead and added the experimental hot fins and helper piece STL's to the BNBSX Mosquito Kit Thingiverse https://www.thingiverse.com/thing:3651160

The new fins have a tighter slot for the Mosquito and the helper pieces tightly encircle the critical area currently exposed.

The new files are not plated because I have not had a chance to print and check how they fit together. They look good in Fusion, and I have made my manual adjustments for fitment. Sometimes real life vs CAD clearances differ. Files are up in case BNBSX users need them sooner than I can test.

To use, insert Mosquito into extruder, add the helper pieces, and then slide on hot end cover. The helper pieces get held in place with tabs like the E3D plenum planes. 

Files to grab are....

experimental BNBSXM extruder hot fin Mosquito use with helper.stl

experimental BNBSXM hotend cover hot fin Mosquito use with helper.stl

experimental BNBSXM mosquito helper 2.stl

experimental BNBSXM mosquito helper.stl

 

Publié : 21/06/2019 5:48 pm
Evan et Dave Avery ont aimé
Evan
 Evan
(@evan-2)
Eminent Member
RE: MK3S Mosquito integration

Hey @guy-k2, here's a treefrog printed with your gcode and an E3D Nozzle-X, started just as soon as I got home from the office this evening. It's definitely got a matte charcoal finish compared to the shiny black. Otherwise, the same roll of Prusament PETG:

I just started a print with the same gcode and plated-copper nozzle just in case, though I doubt we'll be overly surprised with the results.

As for your design changes: dang, you work fast! Looks exciting!

Publié : 21/06/2019 11:45 pm
Page 5 / 20
Partager :