One thing I really like about a threaded connection like this (or like the Makerbot mk7 threaded thermal barrier cooling bar approach) is that a tightened thread is basically a kinematic locating feature, because of the thread flank taper. So you could machine and tap a single bar of aluminum, screw the heatsinks in, and get a precisely-known nozzle offset for dualstrusion. The set screw approach DOES NOT achieve this, because the threads become radially loaded, not axially loaded.
Consider using a slightly thicker than nut thickness threaded plate which can be used for screwing down into the effector or designed into effectors and then two locking nuts on top and bottom for preventing movement. This still requires lengthening the threaded area on the hotend, increasing the adjustability and can also lowers the total mass when the larger adapter is taken into account. See http://www.boltscience.com/pages/twonuts.htm
If the mounting plate is threaded, you only need one lock nut. The usual advice about jam nuts doesn't apply here, because we're not stretching a bolt in a thru-hole. We have extremely mild working loads in 3D printers so simply tightening two threaded parts against each other is perfectly adequate. (Consider that the heat break, hot block, and nozzle have no special thread locking mechanism... why does the heatsink mount need one? It just needs to be held still.
What's the point of being able to unscrew light a lightbulb if the heater wires can't disconnect? Maybe connectored heaters are in the pipeline? I'm not sure if I was clear before, but you could still mount through a plate like on DC's printer. I'm proposing this: Except with 3 screws instead of one. Just getting one to balance was tricky. But if that's no good, my next idea would be like this: Take a normal 1.75mm heatsink, and put a M12 thread on the top portion. Then include both this: as well as a 16mm OD, 3.5mm tall part that has internal M12 threads. Then if someone wants to use the new design they can, and if they want to use it in an old style groovemount they can.
I have been discussing the details with Rory for a little while. I can see two use cases at present: 1. My immediate use case is our new PCB delta printer effector. This needs a mounting arrangement that takes up minimal space, because groove mount takes up too much space on the effector given the other things we want to include i.e. nozzle contact sensor and associated electronics. We also want the hot end to be rotatable to the desired orientation before tightening things up. It would be an advantage (for delta printers at least) if the total distance between the nozzle and the mounting plate can be reduced. A good solution for us is a screw thread at the top of the heatsink, then we can just put it through our PCB and put a washer and a nut on top. We don't need the separate collar, although it could be a substitute for the nut. 2. In dual nozzle machines - including IDEX machines if you want do to mirrored prints - it is important to get the two nozzle heights the same. The arrangement with a collar and grub screw partly achieves that, by letting you screw the heatsink into the collar more or less to adjust the height and then tightening the grub screw. However, it doesn't let you control the angle of the heater block independently, except by a rather messy process of screwing the heater block on more or less. A solution would be to use 2 nuts on the threaded end with the mounting surface clamped between them. The difficulty here is the length of the threaded section needed. One M12 half nut is 6mm thick. So two half nuts, plus one or two washers, plus the plate you are fixing the hot end to (say 2mm thick), plus enough extra for adjustment comes to a total of 16mm or more. This makes the hot end longer that the current design. It's less if you use a threaded plate and just one lock nut as Ryan suggested, but then you use the ability to adjust the heater block angle independently and easily.
Counterpoint to the loading assumption. FSR calibration makes it a dynamic force alternating between light downward load and an intermittent stronger reverse loading. Forgive me if I'm being obtuse, I'm an EE, so mechanics are definitely not my forte and I tend to over engineer on issues I'd like to never return on. My personal experience with the E3D hotends is that the hot block, nozzle, or heat break have become loose on me. I secured the wiring to the effector to prevent the wiring from causing loosening from movement which drastically reduced the frequency of the loosening of those parts, but it still occasionally occurs which I just attribute to using fsr bed leveling or possibly simple vibration. I think the thermal expansion makes the nozzle and hot block the least likely to loosen which is why right now it's at least a month of heavy printing before the heat break grows loose in its interface with the heatsink. While the correction itself is a trivial adjustment to make, its a far bigger annoyance in throwing off the last calibration run and I'd be doing myself a favor to just give it a twist before blaming the belts again. It's probably trivial in all practical senses but my number chaser instincts make me near ocd staying under .03 deviation across the bed just because.
The problem with having a plate mounted directly on top is that it makes the push-fit lock nearly inaccessible if you want to fiddle with the ptfe tube. Also if the top stays at 16mm then there is not really enough room for M3 bolts. https://cad.onshape.com/documents/2...72c5fe239ead3c8b78/e/c60cf3a719c44a3d763fa926
Is fiddling with the PTFE tube done that often? Its a once every few months thing for me, if that. And 3 more screws doesn't seem like too much more time. The M3s would be tight, I agree - but I think it is possible. If not then option 2 would be my preference, as it at least allows for one hotend to work for groovemount and through hole mounting with the addition of a small internally threaded ring for groovemount. Telling tens (hundreds) of thousands of existing users "too bad you all have to design new mounts" is a tough ask. I guess E3D could continue producing groovemount heatsinks for folks that want them, but then that becomes a mess from an inventory and ordering perspective. How about something like this? Clamping style shaft collar, put the triangular hole pattern in it, you get 360 free rotation independent of hotend height.
Being able to easily fiddle with the PTFE tube is not a problem up until the moment you have a problem and you need to fiddle with it! I like the clamping style. Check out the crazy way I clamp the hotend on my machines.. This style ensures a tight fit as well as keeps the push-fit above the effector so it's accessible. This style is likely way too expensive to do in metal though.
Why not replace the press fit collet with a standard 1/8 threaded one. No additional part manufacture needed, the matching push fit connectors are inexpensive and ubiquitous.
Having the plain section should also address both issues, there should be enough thread to go through a 3mm plate and use a nut on the other side (can increase the length of thread if needs be). The reason the plain section is there is so you can use the collar (optional) and not clamp on the threads as this is very bad practice and will wear out quickly.
True, this is something I've been thinking about. My thoughts were that the nut is surprisingly big, 19mm AF and 6mm thick (even for a half nut). Also a grub screw would likely be easier to access for adjustment? For through mounting there isn't anything preventing you using a nut, the collar is just a nice add on for levelling dual head machines.
Yes, quick connect heaters are in the works, we already have these on the fans and sensors, makes sense to do if for the heaters as well. Can't put a timeframe on this as of yet. Back compatibility is great and hopefully something we can do, but I am worried about it becoming too much of a compromise. Option 1 is likely to increase the part cost a fair bit which is something I want to avoid for people who are happy trucking on as is. Option 2 is promising, although I have some concerns about mating against the threaded section and how this will effect wear. e.g the groovemount socket on Titan. I think I'll put a heatsink in the lathe and give it a go, see what comes out!
I don't know what the paradigm we'd go down yet, suffice to say we would NOT ask everyone to redesign everything. Making it back compatible or stocking both variants are both 100% doable, although its good to discuss this kinda thing early on. For now this project is just about working out what good alternatives may be. The shaft clamping idea is another interesting one, again i can make this myself so should be quite easy to try out.
An alternative way to make the height adjustable is by using a stiff o-ring between the heat sink and the effector and tighten the nut on top. By squeezing the o-ring you make it stiff enough to hold it properly, the Nut on the other side makes it stay vertical and height adjustment is done by simply tightening the nut. this gives you a few mm of height adjustment. It also allows you to position the heat blocks any way you want, before tightening. As to diameter, by all means, keep it M12. Then I do not have to redo all the effectors I have. Even so, re-doing the effectors to get rid of the groove mount is worth it.
I have 3 sets of M14 heatsinks and collars left to give out to testers, who would like one? No strings attached, all that I ask is that whoever receives one gives some feedback from testing.
Hi Rory, if you are still looking for testers I'd be happy to help. Currently building a corexy with dual nozzle lifting setup and would be happy to install it on that.
Hi, I'd also like to test it. Right now I am in the process of converting my custom 3D printer to IDEX. I am trying to make the X carriages as small as possible and this kind of mount could be useful.
