The Positron V3 uses the heater cartridge as a dowel pin to prevent the brass insert from falling out. Then the heatbreak screws in and bottoms out on the side face and the nozzle bottoms out on the top face of the insert. 

I'm wondering if something similar could be done for the corners of the trangle material path. I can't think of anything at the moment that's not even larger than the 40mm edition in the previous log, and I fear that PCBWay would charge for every sperate entity.

I still remember asking myself seeing...

...and being like "Ok, you've done that... but y tho?". 

Now I've come to the same conclusion. I'm planning for a 2mm ball though. Maybe even 1.5mm, if I'm safe to assume that E3D is using a 2.5mm bearing to plug a 2mm channel. (I'm using 1mm channels)

The nice thing is that the threads at the bottom of the blind hole no longer need to exist, so I can let 2 holes slightly cut into each other. There's at least 3.5mm of thread space, the outer diamter is now 40mm with 37mm across flats. The mass of the block is now 50g, up from 39g.

For the single M3 grub screw that actually went all the way in, it looked fine.

For all the other ones tho... 

Some screwed out with plastic:

Some screwed out with plastic remaining in the hole:

And some didn't budge at all and it looked like I could risk breaking the M3.

I'm thinking that the small size of M3's are unsuitable for this kind of work. I'd imagine that M4's would be more reliable, and they certainly can handle more torque. The hex M4 grubs that hold the heatsinks seem sturdy enough.

[X] = Experiment / Test

I wake up. 

It's 9am. 

I think "If I test the hotend now, I might be able to make it to the uni if things go wrong." 

It took 6 hours, with the first 3 hours just finding my seldom used soldering gear because the thermistor I have wasn't pre-soldered to the included dupont connector. It's not like Trianglelab's CHC that conveniently had everything ready to go. 

I found the CR10 mounting plate intended for the #Revolving Hotend [gd0012] and used that to mount the Coaxial Hotend on the currently unused DP2 printer, just like the last time.

I got my FysetC AIO out and the firmware is conveniently already configured to the correct thermistor. I started a PID tune, but the temperature climbed really slowly when only a single 40W cartridge was powering the hotend.

Since 80W at 24V is 3.33A and the AIO is good to 5A, I plugged in the second cartridge and now the temperature was increasing at a standard rate.

Usually, the auto PID swings around the target (in this case 196) by a couple of degrees, but this time it only differed by +/- 1 degree. I can only assume it's because there's a lot more thermal mass than for a typical hotend.

The first PID was rather poor so I ran it again, starting from 180C not room temp.

So the first thing was to purge with cleaning filament. It took a while, but I eventually saw something actually poke out the nozzle! (image below)

This was me running the hotend at 200C because 180C was kind of tough. I'm also starting to think I've just bought a lifetime supply of cleaning filament because I've still got loads left. 

I put in some PLA to try extruding it, but then I worried about the fact that PLA clogs easier than other materials for all-metal heatbreaks and switched to PETG instead. Before I did that, I cut of precisely 1 ruler length (305mm) of material to see how much I needed for the colour change.

The first change was rather sharp. That's the dark blue PLA I used before switching to the "Solid Grey" PETG.

This was the first, but since I already started with some PLA beforehand, it's probably not accurate.

I didn't do the length test for the Dark Grey PETG I put in next because it's so similar to the Sold Grey in colour. I'm starting from the lowest filament path and progressively going upwards. Interestingly enough, some material still got to the next cleaning filament input.

The next filament I put in was the extremely wet copper filament, and it didn't take long to see it's identifying bubbliness though the nozzle.

The last input looked like this. I then realised that the pressure inside the hotend is probably pushing up the other inputs since there's nothing actually holding them in place. Thus, when I was extruding one filament, I was periodically extruding a tiny bit of the other 3.

I'm starting to think 24cm is the magic number. It means that 6.5cm of filament needs to be purged for a colour change.

This is white... ish. It seemed off-white when it came out of the nozzle. I also think it's safe to say that the purge length is 6-7cm. 

Lastly, I just wanted to make sure that the grey input was working correctly. Since I've already extruded before, the measurement would be inaccurate.

Conclusions

Perhaps some kind of high-temp Loctite will suffice, but I'm thinking of having a grub screw to prevent the heatbreaks from potentially twisting out of place. I feel like the 2 inputs that weren't mounted have probably loosened, and have precautionally tried to turn the heatsinks in the tightening direction before turning off the heat.

Taking the heatblock off, I havent' seen any signs of leaks yet, though I'd likely need to put in a lot more than the 50cm or so I've done in this test to see that. Just like how SpaceX's main target for their last Starship launch was "it leaves the launchpad", my main target is "Filament actually comes out of the nozzle via all 4 paths, and the hotend is still shiny afterwards".

I'm now taking off what grub screws I can to see the damage report.

So I'm now learning about the wide world of thread taps, such as this article or the video below, and I'm thinking "maybe this is the real reason behind E3D choosing to go with a ball to seal the deal?".

The reason why E3D might have opted for a ball between the hole and the grub screw is to avoid the issue which is that taps aren't that great at creating a thread all the way to the bottom of a "blind hole". 

The heatblock would need to be a 36mm hex to account for this, along with M3x3mm grub screws.

The university technician said that it took 24 hours in total to create this part, and it looks amazing. Thankfully, he sounded like he was enjoying himself. He also milled out both the M6 brass grubs into the inserts, so I've got one spare. If I remember correctly:

• He had to do some operations on the manual tool changing machine because the £80K CNC one didn't have the holders for the tools
  • He said those tool holders are £800 and up.
• He drilled the 1mm paths in 0.2mm increments to prevent the flute clogging up and the drillbit breaking.
  • These two points contributed to the long working time.
• He liked that this was what us in the 3D printing world would call a torture test because of its complexity, and it was found that some GCODE commands weren't working correctly with the machine and someone else set up a post processing script to remedy the issue.

Also, the thermistor didn't fit and he was like "ah, that's why you specified 3.1. I only had 3mm" and another tech said that a 3.1mm drill bit came in yesterday, conviniently enough. Thus the hole was enlarged and it slides in now.

I wasn't expecting him to fill it with grub screws, and all but one were straight, which for a first attempt is rather good. It actually straightened out when I replaced most of them with the brass ones, since, as I feared, these small hex screws are really easy to slip on. I was able to put in more force with the slotted grub screws, and I'd even imagine that if I used stainless or hardened steel grubs, I could use them to self-tap the holes.

Unlike the black grub screws, brass is non-magnetic, so you'd need to use tweezers to hold them in place. 

All the screws went in, but some deeper than others:

2 heatsinks wouldn't slot in as the heatbreak tilted ever so slightly when it bottomed out, and it was fixed by tightly screwing in an M6 bolt which I assummed cleared out any swarf that might be in the threads.

Bamme! Here it is, all together. For some reason, that grub screw on the top left heatsink is really tight compared to the other 3. I might see if an M4 bolt can widen the thread or if I could use an M4 tap at my uni tomorrow. The full weight is 116.5g, excluding the cartridges.

It feels good to get some tangible progress in one of my projects. This feeling of "I'm onto a winner here" is what I needed in my literal dozens of other projects that haven't made it as far. #Tetrinsic [gd0041], I'm looking at you when I say this.

The main ones are the internal grooving tools and the 20mm grub screws that make the insert, but the heatsinks, 4mm grub screws and cleaning filament have all arived. 

I've now gone to my uni's engineering department and the technitian that was looking over my design said that it was going to be a long process and will probably take till next Friday at the earliest (keeping in mind that the UK has a bank holiday next Monday). The long processing time is mainly because there are so many positional and tool changes required. I may be an engineering student, but I don't study mechanical engineering, so the fact that the design is complex but creatable (given the right tools and skill) is a win in my book.

These tools are actually suprisingly small.

Seems the fans were the only thing not in the AliExpress combined delivery and also the first to arrive. Royal Mail was notified of this parcel on the 16th May and the combined delivery 21st May (yesterday), so it could be another week+ before I get the components I need.

I think... that's actually everything I need now for a complete hotend solution. Took just over an hour to plan and model, but wayyyy simpler than what I had to do for the #Revolving Hotend [gd0012].

I've now modelled the cooling ducts. The airways themselves are not optimised at all, but it should serve as a good starting point. 

I also did this to see if there was anything else I might need before I bought the AliExpress BOM, since I've found it unrecommendable to buy components before the assembly file has been completely modelled.

I'm planning to self-tap M5 screws into the 6028 fans. Here's what it looks like with the fans inside:

Hopefully, the plastic being 3mm away from the hotend is good enough for it to not melt.