My 6 Input Hot End – Part 11

1. Introduction

This is going to be a very quick post (by my standards) detailing my tests using PET-G instead of PLA and what I plan to do next.

2. Experimental design changes

From the last tests in part 10, it is clear that heat is being conducted through the filament itself, and those filaments which are loaded but not moving forward during a print were softening and swelling inside the heat break tubes and eventually causing blockages. To be clear, this only happened with filaments which were loaded and heated but not moving forward. There was no jamming inside the heat break of those filaments which were being used. Although some of the filaments may not be moving forward, they are being retracted and then un-retracted because that is how retraction has to work with a mixing hot end. I have previously explained the reasons why so I won’t go over it again here.

So although the thin wall tubes of the Mosquito heat breaks are very efficient, they cannot prevent heat being conducted up through the filament itself. The copper heat sinks on the Mosquito heat breaks work well when the filament is moving forward. I repeat that I have never had a blockage with any of the filaments that were in use – only those which were loaded and heated but otherwise not in use.

What I believe is needed, is to increase the cooling capacity above the heat break zone in order to dissipate the heat that is being conducted up through these “static” filaments. As an experiment, I decided to wrap some thin copper wire around the heat breaks, between the fins and run this wire from heat to heat break, much like threading a pair of shoes. Here is a picture which might explain it better.

In theory, this has increased the surface area of the heat sinking capacity. In practice, it might not be by much because the wire itself takes up some of the surface area between the heat sink fins. And I’ve probably caused havoc with the air flow blowing over those parts. Anyway, I thought it was worth a shot because it was easy enough to do.

3. Testing

Having first replaced the blocked heat break, then wrapped the wire around the heat breaks, I re-assembled the hot end and loaded the printer with PET-G, instead of PLA.

I needed to make an adaptor to connect a 50mm diameter pipe to a 14mm tube as part of the irrigation system on my wife’s “living wall”. That is largely irrelevant but suffice to say that I wanted to make a part using just a single filament.

The print went well although the first layer was too low and I had to use baby stepping to correct. I’ve never had to do that before and I need to investigate. I suspect that it is just due to thermal expansion of the hot end. As readers will know, I use the nozzle itself to probe the bed when homing Z, and so I heat the nozzle to around 140 deg C just to soften and plastic that may have “oozed” from a previous print. But the print temperature is much higher and the hot end is big, so maybe I need to home at print temperature, rather than 140deg C to allow for extra thermal expansion.

As the print progressed, I tried to take the temperature of the copper wires and heat sinks by poking it with a thermocouple. It wasn’t easy because the head was moving but as near as I can tell, it was around 59 deg C. That is about 17 degrees hotter than the temperature tests that I did back in part 3 where I had around 40 deg C inside the empty heat break tubes.

At the end of the print, I moved the head away and was able to extrude filament from the other five inputs which had been static (for about 40 minutes). So there were no blockages.

I think the main reason for this is simply that I used PET-G rather than PLA. Depending on what source one believes, the glass transition temperature for PLA is around 60 deg G, whereas for PET-G, it is around 80. If the heat break temperatures are close to what I was able to measure, then that would explain why I had blockages with PLA but not PET-G.

Another good thing to come out of all this is that there were still no signs of any leaks, so at least that is a step in the right direction.

4. Next Steps

So it seems that I might be able to print with PET-G and not get any blockages. But that single print lasted only about 40 minutes so it could be that I might have had problems if the print time had been longer. Therefore my level of confidence is low.

My current thinking is that, with a mixing hot end, where some filaments are “static” in that they don’t move forward as the print progresses, an efficient heat break zone on it’s own is not enough. I believe that some mechanism is needed to actively dissipate the heat which gets conducted through the filament itself, up past the heat break zone. Either that, or the heat break itself must be lined with PTFE or similar (which of course will limit the temperature that can be used).

Possibly a larger copper heat sink on a Mosquito type heat break tube might work, if enough air can be directed to flow over it. But with six inputs, that would mean spacing the heat breaks further apart, which makes the entire hot end assembly much bigger than it already is.

A more compact solution would be to use liquid cooling. I could probably make a liquid cooled block that the Mosquito heat break tubes would pass through, but I would need to remove the copper heat sinks from the (very thin) tube, and I have no idea how I could do that without destroying the tubes.

I have spent a lot of time over the last two days looking at all the various heat breaks that I could buy, but almost invariably they are designed to be threaded into a hot block by hand and then the nozzle is tightened against the heat break with a tool. What I need is a heat break that has flats or a hexagonal section that I can use to tighten it in the hot block. Preferably one that has a smaller, non-threaded section at the end, then I wouldn’t need to use a separate plate but could screw the heat break directly into the hot block, but that is a secondary consideration.

The reason that I approached Slice Engineering is that I didn’t think that I had the skills or tools to make an efficient heat break. So I wanted to use “off the shelf” parts. The Slice Engineering heat breaks are undoubtedly the most compact, which is a major consideration when one is trying to fit six of them into a single hot block. Unfortunately, that hasn’t worked in this particular application where filament can be heated but mostly static for many hours. I always knew that might be the case and I am grateful to Slice Engineering for supplying the initial batch of heat breaks at no cost (it’s just a shame that I foolishly managed to break them and had to purchase replacements).

So in conclusion, it’s starting to look like I will have to make my own heat breaks (and a liquid cooling block). But I don’t know if I have the skills or tools to do that. The cooling block shouldn’t be too difficult but low conducting metals such as stainless steel are not the easiest of materials to work with. I certainly couldn’t make a bi-metallic type heat break like a Mosquito (I have no idea how Slice Engineering manage to attach such a thin walled stainless steel tube to a copper nut in a manner which is leak proof at 200 Plus Deg C).

I’ll continue to look around and maybe I’ll find an “off the shelf” heat break that isn’t too expensive (I need six of them) and that I can modify. If not, I guess I’ll have to try and make some from scratch. Then I’ll need to design and make a liquid cooling block. So the next update (if there is one) is likely to be some weeks or months in the future.

Ian