Mount for the Diamond 5 colour

Just a quick post to say that I’ve uploaded the OpenScad file for my Diamond 5 colour mount design to thingiverse. You can find it here https://www.thingiverse.com/thing:2586159.

This is a cut down version of what I use on my machine. That is to say, I’ve taken out the mounting arrangement that is very specific to my machine.

As it stands, it consists of the ducts and shrouds to take the 5 heat sink/brass nozzle assembly plus a standard 50mm fan. Whatever bosses or brackets are needed should be added to the “mount” module. I’m no OpenScad expert so it may not be pretty but it gets the job done. I’ve annotated the file so everything should make sense.

Hope someone finds it useful

Ian

Advertisements

The Diamond 5 Colour – part 2

This is a follow up to this post The Diamond 5 colour (“Fullcolor”) hot end (and other related stuff).

One of the issues I mentioned in the earlier post was that I suspected that the 5 colour version of the Diamond hot end suffered from “heat creep”. Or at least I was experiencing problems with PETG  swelling in the area of the heat break which caused partial jamming of the filament inside PTFE liner. So I decided to investigate.

While I write this, I’m still convalescing so what comes out may seem a bit garbled. I hope not but feel free to point out any errors or typos (in a nice way please).

My theory for why the 5 colour version had problems which the 3 colour did not, was that it may be because the lower heat sink fins of the 5 colour were very much smaller in diameter than the 3 colour version. A possible cure might be to use a higher flow rate fan.  Rather than just fitting a bigger fan and trying to print with it, I decide to try and actually measure the temperature in the area of the heat break. I’m glad I took this approach as will become apparent.

So starting with the 3 colour version, here is the test set up that I used.

testSetup

Basically, I removed the hot end from the X carriage and fitted thermocouples touching the inner part of the heat sinks (the smallest diameter) on the side away from the air flow. One was fitted above the lowermost fin. Any lower and it would have been affected by radiant heat from the hot end. The other was fitted close to the top but also touching the inner part of the heat sink and opposite the air flow. The thermocouples were held in place with cocktail sticks, one wrapped in Kapton tape to make it thicker. Maybe not the best method but needs must and all that……

Here is a close up

3colourCloseup

By the way, in case you are wondering the brass nozzle is fitted with my home made silicone sock. The 5 colour doesn’t have a sock but I doubt if it would affect the temperature of the heat sinks. If anything, I’d guess it might make the readings higher with a sock so the 5 colour might be better off without? Dunno.

The tests were all run without any filament loaded and of course, the temperature inside the tubes will be different to that on the outside, so it’s not a scientific measure of actual temperature under printing conditions, merely a method of comparing one thing with another.

With the setup as shown, I then heated the hot end to 200 deg C and took readings at 1 minute intervals up to 10 minutes, then at 15, 20, 25,30, 45 and 60 minutes. Here are the results. T1 is the lower thermocouple and T2 the upper one. Note that the “Time” axis on the graph isn’t a linear scale.

3 colour data

3 colour graph

 

Next up was the 5 colour with the same spec fan. Here is the setup

5colourSetup

5colourCloseup

And here are the results

5 colour data

5 colour graph

Note that in both cases the hot end gets up to temperature within 3 minutes so comparing T1 should be valid. Also, of note is the fact that I’m using a 40 Watt heater cartridge on both hot ends, rather than the 80Watt heater on the 5 colour that for some unknown reason is thought to be necessary.

T1 for the 3 colour version reaches about 58 deg C after 4 minutes or so and stays at that temperature for the rest of the time. Actually it drops about a degree at the end of the hour. Not sure why that should be – maybe just a drop in ambient air temperature? What is also of note is the T2, the temperature close to the top of the heat sink is only slightly above ambient so the lower fins dissipate most of the heat and the upper fins do very little.

T1 for the 5 colour version reached around 66 deg C after 4 minutes which is 8 degrees higher than the 3 colour but it continued to drift upwards, reaching a peak of 70.8 after 30 minutes which is 12.9 deg C higher than the 3 colour version. Also, the temperature near the top of the heat sinks (T2) was about 11 degrees higher than the 3 colour version. The temperatures at time = 0 (therefore ambient) were around 25 deg C for the first test and 23 degrees for the second so despite a 2 degree drop in ambient, the recorded temperatures for the 5 colour version were significantly higher. The fact that the temperature at the top of the heat sinks for the 5 colour version is higher than both ambient and the 3 colour version, reinforces my belief that the smaller lower fins are not as efficient at dissipating heat.

What I don’t know is what the temperature of the filament would be at the “T1” position other than that it would be higher than was measured on the outside of the tube. The Tg temperature for PETG is around 80 deg C as far as I can ascertain. This is purely speculation but taking an arbitrary value of 10 degrees C as the difference between the outer temperature as measured and the inner (filament) temperature at that point, then the filament could reach 80 deg C using the 5 colour version but only 68 deg C using the 3 colour version. So effectively reaching the Tg temperature with the 5 colour but staying below it with the 3 colour version. Thanks to Rob Miller for pointing that out and I think these tests back that up.

The next logical step was to try a higher air flow fan. Of course, if the limiting factor is the surface area of the heat sink fins, then increasing the air flow may have limited benefits. However, it is worth a try……

The highest flow rate 50mm 24V fan that I can find is 20cfm. Apologies if you see it referred to as 22 cfm – that’s what I thought it was but when I checked, it’s actually 20 cfm. So to get a higher flow rate, I had to go to 60mm which meant making a 50 to 60 mm adaptor. It was only after I had designed it that I discovered there are numerous designs on thingiverse that I could have used – oh well….

By the way, if using a 60mm fan works, this works then I’ll have to redesign this hot end mounting arrangement as it was an absolute pig to assemble.

The other little issue is that the fan blows directly down onto the hot end which is only “protected” from the draught by 3 layers of fire blanket type insulation. Also the heater is fitted in the centre of the hot end so would normally be directly under the fan. It seems that RepRap.me had thought of that and their “official” mount has a tube running down the centre to mask the air flow from the heater. So I used that idea and expanded on it slightly. Here is a picture of my mount, with the 50 to 60 mm adaptor fitted.

5colourDuct

The top of the tube is the same diameter as the fan motor, then it tapers down so that the heater cartridge would just fit inside the bottom of the tube.

These OpenScad screen shots maybe show it better.

mount 2

mount1

On the basis that more is better (when will I ever learn?) I sourced two fans. The first was 50.4 cfm and the second was a whopping 84.4 cfm. In hind site, I should have taken more notice of the noise level, the first being 54dBA and the second 66dBA compare to the one I have been using which is 40.5 dBA.

Anyway, here it is with the first (50 cfm) fan fitted.

5colourBigFan

Then I attempted to run the first test again but didn’t get far before I ran into a heater fault. After 4 minutes, the hot end had only got up to 183 deg C and was struggling to go any higher. I attempted to re-tune the heater PID with this big fan running but to no avail. Just too much air. Actually I’m glad it didn’t work because even if it had, there is no way that I could have lived with that noise level.

So, I’ve ordered two more fans of much more modest air flow (and noise level). One is rated at 27cfm which is about 35% higher flow rate than the 50mm fan I’m currently using and at 39dBA, is actually quieter. The other is 33 cfm so about 65% higher flow rate but is slightly noisier at 44dBA.

Quite what I’ll do with the other fan, I don’t know. Maybe a leaf blower? As for the whopping 84cfm fan that I didn’t even fit I’m open to (polite) suggestions.

So part 3 of this blog will follow. If the increased air flow dosn’t work, then for me at least, I’ll be limiting any printing with the 5 colour Diamond to PLA.

Ian

Recent reader rants

I’ve just a spate of quite nasty, negative comments which I’m a bit upset about. Luckily (or unluckily depending on your point of view) these appeared while I was laying in a hospital bed where I had no Wi-Fi access and only my ‘phone. Otherwise, I’d likely have simply deleted this blog in it’s entirety – I really don’t care and it’s only one mouse click to do so.

Now that I have access to my computer again and have had a few days to calm down, I’ve simply deleted those comments. As it’s my blog, I have full moderating rights. If people really want to take what I write out of context and rant about it, then I’ll simply disallow all comments.

So I’d just like to re-iterate the purpose of this blog by quoting my opening post:

The intended purpose of this blog/web site is that it will be a place where I can share my thoughts, ideas, experiments and general 3D printing related stuff. I have no affiliation to any 3D printer brands or manufacturers. What I write here will just be my own personal views and experiences. Hopefully some of the stuff I put on here will be useful or inspirational to the wider RepRap community. If not, well nobody is forcing to read any of it……..

Now it takes fair bit of my time and effort to put together some of these posts (which nobody is forcing anyone to read), for which I get no financial gain. I simply want to log my findings, as much for my own benefit than anything else. By putting them on the web, others can read them if they choose. I also give people the opportunity to comment if they wish to do so. I really don’t care if you agree/disagree but there is no need to get nasty. I’m simply logging my own experiences and of course many other users will have different experiences. Lets try and be a bit grown up about it though.

For example, my method for assembling the Diamond hot end is just that – it’s my method. Many people have problems with leaks around the heat sinks which I have never had. That is what prompted me to write my post. Merely to share the way that I do it, hoping that it may help others who have difficulties. Of course, other methods may work equally well or even better – I’m just saying “this works for me”

Despite making this clear in the opening paragraphs, I’ve been flamed for “telling people how to to assemble their hot end” and then blamed when it didn’t work for them. All I can say is that if you buy a cheap clone, then remember the old saying “you can’t make a silk purse out of a sows’ ear”.

Another example is my experiences with certain filaments. I started that particular post with this:

As ever, these are just my personal experiences and I have no links with either eSun or E3D other than being a paying customer. Neither do I have any “axe to grind” or grudge against any company. What follows are just my own findings, on my machine and should not be taken as any recommendation or otherwise.

Yet I’ve just been blasted for “flaming eSun”. For sure there are bound to be hundreds maybe even thousands of happy customers but my own personal experience (which is entirely what this blog is all about) was not good.  It doesn’t take much intelligence to work out that what I received was a bad batch but that doesn’t detract from the fact that I had a less than satisfactory experience.

So what should I do? Tell lies to keep everyone happy? That isn’t going to happen so if you don’t like it, stop reading what I write – I just log my own personal findings.

Then to the person who summed up my printer as a pile of crap which would be the laughing stock of the TCT show, well sorry to disappoint you but take a look at this from about 7:50   https://www.youtube.com/watch?v=v3Xnb0XzlAs

It takes me time and effort to write these posts for which I get no reward. Nobody says you have to read these posts. If you have an alternative point of view, then start your own blog but don’t start a rant on mine.

I apologise that you can’t get back at me as I’ll simple delete any rants and if they get worse, I’ll disallow all comments. There is also the option of going to “settings – general” and clicking the link “delete your site permanently” which I came very close to doing.

Finally, it’ll likely be while before I post anything else as I just got discharged from hospital, with what seem the entire contents of a pharmacy, and a longish period of convalescence and numerous out patient tests yet to be conducted.

Ian

 

The Diamond 5 colour (“Fullcolor”) hot end (and other related stuff).

Apologies to anyone who follows my blog for not posting much for a while but life has been very hectic lately. Anyway,  I’ve finally got a bit of time to do this post. To make up for the lack of frequency of posts, this one will be quite long as I have a lot to talk about.

What follows are my experiences with the 5 colour version of the Diamond hot end. There are also some other observations to do with mixing hot ends in general. I hope the information will be of use to others.

Before I go any further, I want to make it clear that the Diamond 5 colour hot end was generously supplied at no cost to me by RepRap.me for evaluation purposes. However, my acceptance was on the clear understanding that anything I write about it, good or bad, would be completely unbiased by this fact. I think readers will see that I have honoured this and that everything that follows is a true and unbiased account of my personal experiences.

I should probably also mention that the Duet3D team helped me out by generously supplying two E3D Titan extruders to supplement the three that I already had. The reason being that they wanted me to take my printer (with 5 colour capability) to the TCT show to help demonstrate the capabilities of the Duet electronics and firmware. It’s fair to say that when it comes to 3d printing electronics and firmware, I am biased towards the Duet. That’s not because of any freebies but because I personally think it’s the best.

I should also make it clear that I was born in England and so I use English spelling which is why I write “colour” rather than “color”.

Anyway, I’ve already covered the modifications that I made to the printer here 5 Colour upgrade so I won’t go over that again. Assembly of the 5 colour is much the same as the 3 colour except that the heat sinks have two flats on the top which makes it much easier to tighten them by using a spanner instead of messing around with pliers or pipe grips.

The first thing of note is that the 5 colour Diamond comes with a massively powerful 80 Watt 24V heater or 60 Watt 12V heater, rather than the more conventional 40 watt heater that the 3 colour version comes with. As my system is 24V, I therefore had the 80 Watt version. Here is a link to the relevant part of RepRap.me’s web site https://www.reprap.me/diamond-fullcolor-hotend.html.

There is no mention of why it is necessary to use such a powerful heater except for this statement “Special heater cartridge is necessary for the Diamond Fullcolor Hotend”. One thing that immediately struck me was the fact that, presumably if it needs an 80 Watt heater (the 24 volt version), then the 60 Watt 12 volt version might be marginal. Conversely, if 60 Watt is adequate, then why does the 24V version have to be 80 Watt? Anyway, after fitting the hot end assembly, the first thing I did was to tune the PID parameters using the Duet auto tune facility. The result was fairly alarming in that after the tuning sequence, I received a warning to the effect that the heater could potentially reach 647 deg C.

This was all done with the heat sink cooling fan running by the way. Also, I use a fan which shifts about 22 cfm which is higher than the one that RepRap.me suggest at 18cfm, (at least for the 3 colour version). There is no specification that I could find for the fan on RepRap.me’s web site for the 5 colour version, except that it’s 50x50x15 mm and as near as I can tell, it’s the same fan for both the 3 colour and the 5 colour. The price is identical so I’d guess that it’s the same fan. Anyway, the one I use is 50x50x20 (5mm taller) and I haven’t been able to source a 50mm fan that has a higher flow rate – more on this later.

Having tuned the heater, I tried heating the hot end and the first thing I noticed was that it heated up very quickly (nice) but overshot the set temperature and then oscillated quite a lot about the set point (not nice). It seemed that although the PID algorithm was trying its’ best, the heater was simply too powerful. I loaded filament into all 5 inputs and tried again but with the same results. At this point I sent an email to RepRap.me asking them why they thought that such a powerful heater was necessary but didn’t get a reply.

So I re-tuned the heater but this time I used 60% PWM, effectively reducing the power to 60%. This time I got more sensible results with a predicted maximum temperature of 350 deg C. I ran the printer like this for quite a while with no adverse effects. Some time later, I had to strip down the hot end to make a change to the mount, so I fitted a “standard” 40 Watt heater cartridge at the same time. I have done a lot of testing with this, even feeding 5 filaments using a mixing ratio of 0.20:0.20:0.20:0.20:0.20 (equal amounts of each) of PET-G at around 230 degrees C and at quite high speed and haven’t witnessed any significant drop in hot end temperature while doing so. My conclusion therefore is that a 40 Watt heater is perfectly adequate. I have a theory as to why RepRap.me think a more powerful heater but I’ll come to that later.

The next thing to note is that the 5 colour Diamond suffers from the same issue of not actively mixing the filaments, as the 3 colour version. That is to say, what comes out of the nozzle is akin to stripey toothpaste. This was no real surprise to me because as far as I can tell, the only difference between the 5 colour and the 3 colour is that there are 5 inputs going into one, instead of 3 inputs. However, it should be noted that RepRap.me refer to it as a “Diamond Fullcolor” hot end rather than “5 colour”. To me, this implies that it will actively mix filaments to create other colours but in practice, the actual colour varies around the printed object because what comes out of the nozzle is biased towards each input.  Here is an example

ariaLeft

ariaRight

This is the same “Aria” by Loubie (https://www.thingiverse.com/thing:600550) that I printed in FilaPrint Premium solid PLA viewed from different sides. I did it by running a little script on the gcode file to vary the mixing ratio at each layer change. It started with White which it then mixed with Blue, then Blue with Yellow, Yellow with Red and finally Red with Black.

Just looking at the Green on the lower part of the body in the lower picture. This was produced by mixing Blue with Yellow. Yet in the upper picture, there is very little Green but Yellow is much more prominent because that was the side where the Yellow filament entered the hot end.  Arias’ face shows an even starker contrast with one side being Black and other Red.

This effect can be negated to a large extent by using transparent filament. So here is a scaled up version of Aria that I printed in transparent FilaPrint PET-G.

AriaPETG1small

AriaPETG3small

Here you can see that it appears the colours are much better mixed and look very similar from both sides. In fact, the filament still comes out of the nozzle as “stripey toothpaste” but because the bead is transparent, it doesn’t matter if one side of the bead is Blue and the other side is Yellow – it will still appear Green to the eye.

As an aside, there are other issues with the 5 colour Diamond which mean that this particular print was very difficult to produce and needed a somewhat “hands on” approach. This is covered at the end of this post.

So, it seems to me that there is very little point in scouring filament suppliers to try and obtain Cyan, Magenta, Yellow, Black and White because they will not be actively mixed to produce other colours. Instead, I substituted Cyan with Blue and Magenta with Red because these colour are much more readily available.

Moving on, we come to the next little issue with mixing hot ends which is to do with the characteristics of PLA. As far as I know, this is undocumented elsewhere – at least I haven’t come across anything in my own research. It is common to the 3 colour Diamond as well as the 5 colour Diamond and will also affect any hot end that has filament loaded at print temperature but which may not be moving forward, including for example E3D’s Cyclops.

What happens is this. Once the filament reaches about 170 degrees C, it starts to become less and less viscous (more and more runny). I met a very nice and knowledgeable lady called Susan Woods (susi) from Rigid Ink at the TCT show and here is her explanation.

…………….”As you have noticed, PLA starts to degrade (hydrolyse) when kept at printing temperatures for more than a few minutes. Just to really cheer you up, this degradation starts at 175°C and accelerates the higher that the ‘holding temperature’ is set at.”…………

So what does this mean in practice? Well basically, if you mix PLA in anything other than 50:50 ratio, then one filament is likely to be more runny than the other unless you print at highish speed. The greater the difference in mixing ratio, the more prominent this will be. Another thing is that it is essential to use some sort of purge mechanism if you have a filament that has sat in the hot end for more than a few minutes. This is a blow to my attempts to try and print without purge by moving the tool change point forward in the gcode file. Although the script has now had all the bugs ironed out if it, what happens is that when switching between filaments, what is in the melt chamber will come out all at once as a single runny blob. Of course a purge mechanism doesn’t have to be a sacrificial tower. It can still be the infill but in that case it does mean that the infill itself my be weakened.

As an experiment I tried to print an object that was essentially a series of twisted cylinders set in a circle. There were 10 cylinders in total, 2 each of 5 colours with each pair of colours opposite each other. I didn’t use a purge tower. Instead I sliced the object with 3 perimeters, thinking that when changing from one colour to another, the purging would happen during the inner two perimeters so by the time the outer perimeter was laid down, the new “colour” would be flowing. My thinking was that each filament would be in the melt chamber for the same amount of time, so should have “hydrolysed” by the same amount. In which case, maybe varying the extrusion amount would compensate.  Here is the result:

ridges 1

ridges 2

Unfortunately this hasn’t photographed well but if you look closely at the Red cylinder in the upper picture, you can see raised ridges with a slight trace of Black. This had me wondering for a while as it had occurred on all of the cylinders. Eventually, I realised that the pattern of ridges for each pair of cylinders was diametrically opposite, which is shown in the lower picture. That is to say, where there are raised ridges on one cylinder, the other cylinder on the opposite side of the circle has “valleys” and vice versa. You can see the effect better on the Yellow cylinders which have some bleeding from another colour although in this case, the ridges themselves aren’t as prominent. I finally figured out what was causing it. At each layer change, one colour is printed before moving on to the next colour so the cylinders were printed in pairs. But the order changes. First it prints (say) Black, then Red then Yellow then Blue then White. After which there is a layer change and it starts again only this time it continues with the White, then Black, then Red, then Yellow then Blue and as it has already done the White, there is a layer change. The next layer starts with the previous colour, in this case Blue. To further “scramble” things the next cylinder of the pair will be the one that is nearest the print head.

It transpires that the raised “ridges” coincide with which tower was printed first. So, for the first tower, the filament has sat in the hot end for a while and has hydrolysed so is more “runny”, leading to over extrusion and the raised ridges. By the time the second tower is printed, new filament has been pushed into the hot end so is more viscous and extrudes “normally”.

The cure for this is would be to always purge the filament before printing with a new colour. It is more about purging the new but hydrolysed PLA than it is about purging the old colour out. That is to say, the purge has to be enough to clear out the remains of the colour that has just been used plus whatever of the new colour has been sat in the melt chamber and hydrolysed.

The good news is that it still isn’t a huge amount of filament that has to be purged compared to other ways of printing multiple colours. Mr Prusas’ smart towers come to mind which use a massive (in my opinion) amount of filament. I did manage to produce this vase for the TCT show which prints with no separate purge tower. Basically I wrote another little script that goes through the gcode file and randomly assigns a tool (colour) to each layer. It was sliced with 3 perimeters and because it is larger than the cylinders I printed earlier, the inner perimeters lay down enough filament to purge out the old as well as the new but “hydrolysed” PLA.

BandedVase1

CloseUp1

At this point I just want to digress slightly. At the TCT show, the Duet stand was very close to E3D’s stand. One of the people from E3D saw this vase being printed and asked if he could have it or buy it for his wife and I was happy to donate it. Then another lady from E3D said that she wanted it, so I printed another for her. I was very flattered by that (printing objects to give to people who work for E3D). To top it all, a Swiss couple who were just looking around the show, also took a shine to the original that I had taken along. To cut a long story short, I gave them that one at the end of the show and received this response a day or two later by email.

 Dear Mr. Pegg

My wife and I would like to thank you again for your generosity. We are really happy that we can exhibit your beautiful vase next year at the jubilee exhibition of Stamparia Strada, the historic printing museum I work for (www.stamparia.ch<http://www.stamparia.ch/>). In case you shall be in the Engadine valley next summer: Do not miss to have a look at your piece of art!

Kind regards

Ted Gut

So fame at last! Please excuse the digression but I’m really flattered and just had to mention it here.

Anyway, back on topic………..

After abandoning my tool advancing script I wrote another quick script to look for a tool change, then retract, move the head to the rear of the bed, un-retract, “draw a line” about 200mm long, retract, move back and resume. My longer term plan is to physically move the print bed back which will allow the print head to be in a position where it isn’t above the glass build platform. Then I can simple “squirt” some filament into a “bucket” to purge the hot end. My slicer of choice, for various reasons is Slic3r but it doesn’t do purge towers well.

With that purge mechanism, I was able to produce this:

flagsWithPurge

Obviously, with something like this where there are 5 colour changes on every layer and it’s viewed from above, there is nowhere to hide any purging so it has to be done outside of the print. You can see the purge “lines” at the back of the bed so it illustrates how little filament is actually needed, even when the new filament is PLA which has hydrolysed.

Finally, there is one more thing I need to mention which is specific to the 5 colour Diamond hot end and it’s not good. When using PET-G I had big problems with what seemed partial blockages. To cut a long story short, the filament was swelling somewhere in the area of the heat break, to the diameter of the PTFE liner. It wasn’t a case of molten filament being drawn up above the heat break by too much retraction, and then solidifying. Rather it was just a swelling of the filament. I seem to be getting it with PLA too but it doesn’t cause a problem. I have a hunch that it is temperature related so anything below about 200 degrees is OK but above that , problems occur. With PET-G (which has to be printed at a higher temperature than PLA), this “slug” (about 6mm) of swollen filament binds inside the tube and it takes an awful lot of extruder tension to get it moving. Once that “slug” has gone into the hot end, the filament moves freely until it sits and soaks at print temperature for half an hour or more. The longer it sits there, the worse it becomes. To make matters worse, using a high extruder tension on my Titan extruders can simple grind the filament away. This is because with a mixing hot end, we have to retract all the filaments at the same time – not just the one(s) that is\are being used. So taking the large “Aria” for example, there are about 5,400 retract\un-retract cycles. This means that by the time the final colour comes into play, the same section of filament has probably be retracted and un-retracted about 4,000 times. Bizarrely, PLA can withstand this sort of treatment as the hobbed bolt seems to cut small grooves in the filament which then acts like a rack and pinion. With PET-G, I found that the grooves got bigger and bigger and it just ground away the filament.

I mentioned earlier that to print the large Aria in PET-G required a “hands on” approach and this is why. Essentially I had to start with “normal” extruder tension when mixing the first two colours, then just before the next colour came into play, I had to increase the extruder tension on the “new” filament to overcome the partial blockage caused by the filament swelling inside the PTFE tube. Once the new filament was flowing freely and a few mm had been extruded, I could then back off the extruder tension and all would be well until the next filament came into play.

So it rather looks like the 5 colour suffers from “heat creep”. In fairness, it is possible that the issue is caused by the way I chose to install the hot end such that it is mounted between two parallel 2020 rails. Potentially this could restrict the flow of exhaust air as it exits the heat sink fins. On the other hand, the 3 colour version is mounted in exactly the same way and I don’t have any issues with that. So I’d be interested to hear if anyone else has experienced the same problem or otherwise.

My suspicion is that this is an inherent problem due to the design of the heat sinks. In order to fit 5 of them in the same circumference, the lower fins have been reduced in diameter compare to the 3 colour version. Here is a picture of the 5 colour to illustrate.

5colourDiamondOnMount

This next picture shows the 3 colour version for comparison.

fitThermistor

You can see how much bigger the fins at the bottom of the 3 colour version are compared to the 5 colour version  (the bottom being that part closes to the brass nozzle and not the way they are oriented in the pictures). This is likely to make the heat sink less efficient at conducting heat away and may explain why I have the problem with the filament swelling and jamming inside the tube.

I did do a crude and not very scientific experiment to try and measure the temperature by putting a thermocouple between the lower two fins with the hot end heated to 200 deg C and the fan running as normal. With the 5 colour version, I was reading around 66 deg C but with the 3 colour version it was around 42 deg C. It wasn’t very scientific because I cannot be sure how much of the thermocouple was in contact with the fins in each case, so it may not be a valid comparison but such a big difference seems to add weight to my theory.

Furthermore, it brings me back to the first part of this post where I mention that a 60 or 80 watt heater is recommended but that I have found that a 40 Watt cartridge is more than adequate. I can’t help but think that the reason such a large heater is specified is that problems with filament jamming have been encountered, and without fully understanding the reason for that jamming, it has been thought that a 40 watt heater is under powered.

Of course, I could be completely wrong so my next step is to try and prove it and hopefully find a solution. So my plan is to get more air flow over the heat sink fins – especially the lower ones. I have mentioned that I cannot find a higher flow rate 50mm fan so I am redesigning the mount, fan ducts and shrouds in order to fit a 60mm fan. I have two on order. One is 50cfm and the other a whopping 80 cfm.

Once I have these parts, and a bit more time, I will test and report my findings by posting on this blog.

I hope there is something in all the above ramblings that may be of use or interest to readers.

Ian