MK3S Mosquito integration  

RE: MK3S Mosquito integration

Posted by: vintagepc

It is by nature also a very crumbly mineral. I'd have concerns about flecks of it falling into the print/on the printer. Mica dust on your heatsink definitely won't do anything for its cooling ability... 🙂

Only if it isn't a section of a single crystal. I've gotten 3"x5" sheets of mica from McMaster that appear to be a section from a single crystal. It doesn't crumble, it peels into super thin flexible sheets.

The amazon link above is actually a mica composite (the description even says it is made of "Mica, Paper, Silicone Resin"). That stuff might fleck, or the resin binder might fully entrap the mica. McMaster does sell some composites, they may or may not flake. I don't have any experience with them, but the ceramic and glass composites claim to be machinable... I suppose as long as your machining tools are harder than the ceramic or glass (mica is quite soft).

You may be more familiar with natural micaceous schist where the mica crystals have been twisted and mangled by the metamorphic process that created the schist. My guess is the single sheet stuff that I got from McMaster is probably mined from granitic pegmatites, which also explains why it is significantly more expensive than the processed crap in the amazon link. (OK, "crap" might be a little strong. It is probably plenty fine for it's application, namely a microwave transparent cover for the waveguide into a microwave oven's cooking cavity. But probably not the best thing for high temperature applications.)

RE: MK3S Mosquito integration

It survived being cooked at 230 deg C without anything happening and it does not flake that easily else you would have mica in your microwave meals, like sembaruzu stated, the microwave guide cover is a composite and appears to hold the mica together quite well.

I'm going to have some more trials with it because it may work quite well as a cooling deflector for 3d printing.

I'm going to start by drilling a hole so that just the conical part of an E3d nozzle pokes through and suspend the mica sheet below the heat block and see what effect it has on temperature stability of the heater block and whether it could improve part cooling (Try overhangs from various angles).

It's cheap enough to try a few experiments, and you never know, it may just be useful

RE: MK3S Mosquito integration

You may have already thought of this idea, but just in case you haven't:

It might be an interesting experiment to try to make an enclosure box. The pieces of mica composite might be able to be held together using high temperature RTV silicone sealant (often sold as a liquid gasket material at auto supply stores to create oil pan or valve cover gaskets). I guess it depends on how well the silicone sealant adheres to the mica composite.

Not quite sure how to hang the box below the hotend such that it doesn't flop around when the x-axis changes direction. I guess my idea is just half-baked after all. 😉

RE: MK3S Mosquito integration

I imagine Chocki will be making geiger tubes with his left over mica. 

RE: MK3S Mosquito integration

Nah, I thought I might use it as a clean imaging substrate in atomic force microscopy to check the layer adhesion on my prints 😜.

Far more advanced use than a window to a geiger tube...

RE: MK3S Mosquito integration

Yes, but you can show off how well you 3d printed the vacuum tight geiger tube. That would be an ultimate demo of your printer's capabilities.

My 1st silicone sock for the Mosquito came out of the mold in 1 piece. Dawn dishwashing detergent - not only good for cleaning build plates but works as an effective mold release agent. Sock is further curing overnight in my filament dryer. Hopefully will be able to test it tomorrow. BTW Permatex smells a bit evil straight out of the mold.

RE: MK3S Mosquito integration

Installed a Mosquito on my MK3S yesterday (using https://www.thingiverse.com/thing:3515782 printed with Atomic CF-PETG). I did use Slice's 50W heater and 450C thermistor, as well as the vanadium 0.4 nozzle. I used the boron nitride paste.  I also took the opportunity to replace the Noctua fan with the oft-recommended Sunon.

I was a bit puzzled that the heater and thermistor just sit loosely in the block, though I suppose the idea is that the boron nitride does the conduction.  There are screws that keep them from sliding out, however. There are also no installation tips provided by Slice, and an odd-looking zip tie is included that I can't figure out what it's for.

My initial experience is very positive. It heats up VERY quickly and performs well. I'm interested in the silicone sock outcome.

RE: MK3S Mosquito integration

Is yours printing OK with your usual temperature settings or have you needed to bump up the temps?

The loosely in the block, but packed in with boron nitride solution still seems pretty lame to me. When we do CPU cooling, nothing beats metal to metal contact for thermal conductance. Thermal paste is only to avoid insulative air. In the Mosquito, Slice is relying completely on conduction through boron nitride.

Rather perplexed that I'm having such a bad experience thus far. 50 watt cartridge won't be here until next week. Silicone sock test later today.

Yeah, it is so peculiar that Slice doesn't have detailed assembly instructions anywhere on their web site.

RE: MK3S Mosquito integration

I haven't printed with it enough to notice temperature issues, trying more today. I do note that the hotend and heatbed temps disagree by 5C at room temp (was no more than 1-2 before), making me suspect the table for the Slice thermistor isn't quite right. I may try to do some independent temperature readings to see what I get.

I sent an inquiry to Slice about the mounting and will see what they say.

RE: MK3S Mosquito integration

Thus far the silicone sock seems to be making the Mosquito hotend actually usable. Black PETG tree frog is actually coming out shiny black instead of flat, charcoal gray. Not done with print yet, but overhangs are also looking much better. This is with my usual temperature 240/245C and parts cooling fan settings. Without the sock, I had to bump up at least 5C and drop cooling fan by 50% to get layer fusion. That produced PETG parts that were satin instead of glossy and overhangs were poor at the reduced fan. I'm still on the brass nozzle. Started with the easiest to succeed first. What I'm seeing is super encouraging.

I think I understand what is happening.

1. The Mosquito is exquisitely sensitive to airflow and loses a large proportion of cartridge heat if print cooling flow strikes it. This is probably a combination of highly exposed heater cartridge and relying on boron nitride rather than tight metal to metal contact for heat transfer.

2. 40 watt E3D heater cartridge is barely adequate to keep up with heat losses if any print cooling air is near it.

3. Mosquito works with R4 nozzle because that is a front cooling nozzle which doesn't move much air near left side of Mosquito where heater cartridge lies. This is also consistent with it working well with a right sided cooling nozzle.

4. The BNRHD nozzle is a two sided cooling design derived from the RHD. The RHD does two sided cooling to improved far side overhang performance. Consequently the left side of the nozzle flow goes right past the heater cartridge. The Mosquito's massively loses heat due to the airflow. Although it can keep up the temperature, it simply cannot transfer enough thermal energy from a 40 watt cartridge to work properly with that airflow.

5. A 50 watt cartridge MAY give enough thermal input increase to overcome the airflow, but I have not tested that.

6. Silicone sock on mosquito dramatically improves its thermal efficiency and resistance to parts cooling air.

I would say that a silicone sock is essential if one wants really good parts cooling AND a well functioning Mosquito hot end. 

BTW Slice Engineering has yet to be back in contact with me.

Mosquito sock mold I used is by sadkins1981

https://www.thingiverse.com/thing:3562134/remixes

and

http://www.thingiverse.com/thing:3562134

RE: MK3S Mosquito integration

Interesting. I should watch the temp graph when the cooling fan comes on to see what it does. I could well imagine that more forceful cooling, especially from both sides, would have an effect.

RE: MK3S Mosquito integration

Changed from brass to slice Vanadium nozzle (both without boron nitride). Thermal conductivity worsen enough to make functional print difference. With same gcode, PETG no longer is fully melting before extrusion. It IS better than without the sock. Without the sock, black PETG came out so milky it looked dark gray rather than black.

RE: MK3S Mosquito integration

I understand that your extruder is very different from mine (basically the R4), but I am not seeing the bad results you are. Here's that same (I think) tree frog printed in AmazonBasics PETG at 240/245. No sock. I do use the boron nitride. Furthermore, an examination of the temperature graph shows no budging when the part fan comes on.

(Damn this forum and its aspect ratio bug!)

RE: MK3S Mosquito integration

Hi

Certainly hear you on Slice Engineering's lack of technical documentation.  It is a bit limited. I was able to find the picture showing the application of the supplied tie wrap. Looks like it's intended use is for a Bondtech Mosquito configuration.

As you already know the paste becomes a solid with an electrically insulative property when it dries.  If the cartridges fitment were to tight it would not be electrically isolated. The paste also provides the heater and thermistor with a strong thermally conductive bond while securing the cartridges mechanically.  This is a standard during the manufacturing of high temperature components. During installation the nozzle threads should include a thin application of the paste. If not applied it can become a huge thermal transfer bottleneck from the block to the nozzle.  Having problems with high speed prints? It should also be noted when properly applied the paste will reduce heater-sensor loop latency extending the life of the heater cartridge.

I'm hoping that if properly installed, tuned, and cooled my Mosquito installation will provide long term service for most print materials.  I'm having excellent results. For myself I'll never go back to the threaded heat break architectures. They simply lack a desired continuous improvement engineering path.  I'm waiting for the 30cm superduper volcano that can melt a single spool of filament in a single bound! Buy one now!

When returning home from travel I'm looking forward to additional testing.  The excellent and knowledgeable contributors to this thread have peaked my curiosity with several of their comments.  Everyone has been super helpful.

RE: MK3S Mosquito integration

Now I'm really confused. I used a Thermapen to measure the temperature right between the heater and the thermistor. I set the temp at various values and then read the Thermapen's result. I got temps considerably lower than the setting:

50 48.1
75 63.3
100 82
125 107
150 121.7
175 143.3
200 168
225 199.7
250 204.9
275 229.0

But this can't be right based on the way it prints - no way I'd get good PETG results at under 200C.

Here's my thermistor table - did I do something wrong?

const short temptable_67[][2] PROGMEM = {
{ 22*OVERSAMPLENR , 500 },
{ 23*OVERSAMPLENR , 490 },
{ 25*OVERSAMPLENR , 480 },
{ 27*OVERSAMPLENR , 470 },
{ 29*OVERSAMPLENR , 460 },
{ 32*OVERSAMPLENR , 450 },
{ 35*OVERSAMPLENR , 440 },
{ 38*OVERSAMPLENR , 430 },
{ 41*OVERSAMPLENR , 420 },
{ 45*OVERSAMPLENR , 410 },
{ 50*OVERSAMPLENR , 400 },
{ 55*OVERSAMPLENR , 390 },
{ 60*OVERSAMPLENR , 380 },
{ 67*OVERSAMPLENR , 370 },
{ 74*OVERSAMPLENR , 360 },
{ 82*OVERSAMPLENR , 350 },
{ 91*OVERSAMPLENR , 340 },
{ 102*OVERSAMPLENR , 330 },
{ 114*OVERSAMPLENR , 320 },
{ 127*OVERSAMPLENR , 310 },
{ 143*OVERSAMPLENR , 300 },
{ 161*OVERSAMPLENR , 290 },
{ 181*OVERSAMPLENR , 280 },
{ 204*OVERSAMPLENR , 270 },
{ 229*OVERSAMPLENR , 260 },
{ 259*OVERSAMPLENR , 250 },
{ 290*OVERSAMPLENR , 240 },
{ 325*OVERSAMPLENR , 230 },
{ 364*OVERSAMPLENR , 220 },
{ 407*OVERSAMPLENR , 210 },
{ 453*OVERSAMPLENR , 200 },
{ 501*OVERSAMPLENR , 190 },
{ 551*OVERSAMPLENR , 180 },
{ 603*OVERSAMPLENR , 170 },
{ 655*OVERSAMPLENR , 160 },
{ 706*OVERSAMPLENR , 150 },
{ 755*OVERSAMPLENR , 140 },
{ 801*OVERSAMPLENR , 130 },
{ 842*OVERSAMPLENR , 120 },
{ 879*OVERSAMPLENR , 110 },
{ 910*OVERSAMPLENR , 100 },
{ 936*OVERSAMPLENR , 90 },
{ 948*OVERSAMPLENR , 85 },
{ 958*OVERSAMPLENR , 80 },
{ 975*OVERSAMPLENR , 70 },
{ 988*OVERSAMPLENR , 60 },
{ 998*OVERSAMPLENR , 50 },
{ 1006*OVERSAMPLENR , 40 },
{ 1011*OVERSAMPLENR , 30 },
{ 1013*OVERSAMPLENR , 25 },
{ 1015*OVERSAMPLENR , 20 },
{ 1018*OVERSAMPLENR , 10 },
{ 1020*OVERSAMPLENR , 0 } 
};

RE: MK3S Mosquito integration

Hi holmes4,

After my own temperature observations and the wide variations in this forum thread I've decided to take all readings by passing a probe through the filament path to the tip of the nozzle.  My plan is to use a throw away nozzle (tool box testing nozzle) filled with paste.  Insert the probe and warm the heater to a low temp for a drying period.  Then step through a range of temperatures with fan changes at each step  Can't think of any other way to remove the variables.  With my machine I'm not using the high temperature thermistor as of yet....

RE: MK3S Mosquito integration

I like your idea, James, though it's not something I'm likely to try. Right now I'm finding that if I keep using the temps I have been, it works fine.

I would be very interested if anyone using the high-temp thermistor could check my table. There are some variations in how the tables are set up in the .h file and no explanation of them.

RE: MK3S Mosquito integration

Vanadium nozzle still vexes me.

Went up to 255C --> About 40% of frog surface is glassy looking. Rest still satin finish or flat.

Went up to 260C --> Most glassy surface, but overhangs are dropping under legs and belly.

Trying again with 245 but with boron nitride paste on Vanadium nozzle.

This may also need testing with an R4 fan shroud, but there is another BNBSX user who is getting good results with the same BNBSX extruder + BNRHD nozzle I am using. He is even on regular old 40 watt E3D cartridge and thermistor. Really weird.

I guess I could put in another new 40 w E3D cartridge and thermistor I have here. 

RE: MK3S Mosquito integration

I have a similar setup to @guy-k2, using a Mosquito with a BNBSX extruder. Apologies for being late to the conversation -- I broke off my print fan's wires this weekend, and so my printer has been idle. Up to that point, I'd gotten in a fairly modest amount of printing, and it had mostly been PLA. Initial impressions are pretty neutral.

I just got a new fan installed and running tonight and broke out the black Prusament PETG. I started with an XYZ calibration cube. The top cube is with the stock E3D V6, while the bottom is the Mosquito. Posted without commentary -- feedback and criticism welcomed!

Printed with the same GCode from Slic3rPE at 240/250C with the Prusament PETG profile defaults, 0.4mm E3D nozzle and 0.1mm layer heights.

Note that there have been other changes between the two, including a swap in extruder stepper, belts, pulleys, and toothed idlers. (and it looks like I still haven't dialed in the belt tension either.)

Obviously, this isn't a terribly strenuous test. I chose it because it's a part I have a copy of from before the hotend switch for comparison, and because it prints fast.  For more detail, I have a treefrog printing right now, and I'll be back with the results in about...let's see...an hour and 26 minutes or so. 😉

RE: MK3S Mosquito integration

Thanks for coming aboard, Evan. I'm VERY interested in your tree frog.

I just finished a tree frog using 240/245C, vanadium nozzle WITH boron nitride on threads. Ugh! It is a dark gray, flat finish frog.

If your frog comes out glassy looking, maybe I should swap in another new heater and thermistor.

The ones in my Mosquito are the same ones since I started my trial. Both are fresh, new ones from Prusa, but maybe I lucked into a bad one? Maybe my heater is extra loose. It was wobbly loose going into the Mosquito. Fortunately I had three of each in the shipment. Could possibly also jam it tight with some brass shim.