My 6 input (5+1) mixing hot end

deckingman

@Dougal1957 said in My 6 input (5+1) mixing hot end:

@deckingman said in My 6 input (5+1) mixing hot end:

@arhi said in My 6 input (5+1) mixing hot end:

.......................
question, the acrylic "flow visualiser" with thermal probe - you made it or you purchased it ?

You haven't been paying attention have you? https://www.youtube.com/watch?v=Hai_zqtcauQ&t=145s

and

https://www.amazon.co.uk/gp/product/B078PFRG74/ref=ppx_yo_dt_b_asin_title_o08_s00?ie=UTF8&psc=1

I haven't watched it yet Ian but have you connected the Temp monitor to the Duet if not the Water-cooled PC guys do a sensor that is in fact a 10K thermistor which work quite well with the Duet.

You need to catch up too Doug Short answer (as you haven't been watching my videos), is no - that sensor just happened to be bundled with the flow meter so I bunged it on the printer because why not? I will be fitting a temperature sensor of some sort that I can connect to the Duet for monitoring purposes. No point in controlling the pump speed as it's almost silent, so I'm happy enough with the current arrangement which is to turn it thermostatically when the hot end temp>50 deg C. But It might be nice to take some action via a macro/conditional gcode if the coolant temperature exceeds some pre-determined value.

arhi

@deckingman said in My 6 input (5+1) mixing hot end:

You haven't been paying attention have you?

watched the video but missed that flow thingy the first time .. probbly something distracted me .. and since I use very different water arrangement (submerged pump, long hard pipes - ones ppl use for pluming that you heat weld, heat exchanger, pump, 10m away from the device .. ) I did not pay too much attention at your pluming work as is too different from mine .. you have it as a single portable device and I kinda created some plumming in my work room that handles watercooling different devices.. but this flow visualiser looks interesting so I'm def. getting one

deckingman

@arhi said in My 6 input (5+1) mixing hot end:

but this flow visualiser looks interesting so I'm def. getting one

I wouldn't have thought that it would be too difficult to print or otherwise make something similar. My initial idea was to make something like a paddle on an arm that would sit in the coolant pipework and get deflected be the flow. The arm would then trigger a microswitch which would light an LED to show that the flow was healthy. But then I saw that flow meter thing and it was cheap so..... It's just handy to have a visual re-assurance that all is well with the coolant circulation.

theruttmeister

@deckingman said

Copper has a much higher coefficient of thermal expansion than stainless steel, so one would expect the tubes to come loose at high temperature but they don't. Maybe they use some sort of high temperature adhesive but there is no sign of any residue.

If I knew the answer, then it would reasonably easy to replicate using thin wall 4mm ID stainless tube which would take a PTFE liner, instead of the 2mm that Slice Engineering use.

Copper and stainless steel have virtually identical coefficients.
If the copper side is doing it job correctly, the junction of the two metals doesn't actually heat up much beyond room temp.
Press-fitting stainless steel into copper is very easy, as stainless is much much harder than copper (which is basically cheese in comparison).

The trick is to design an assembly jig that both constrains the exterior of the tube and has an internal pin (as you say). That combination will prevent the tube from buckling or bending and it can be pressed in using an arbour press or whatever you have handy.

The biggest challenge is machining the copper, which is a pain.

Properly designed press-fits are a great way to build a hotend, you just end up spending a couple of hundred bucks on tooling up front.

dc42

@deckingman, it's great to see that you have the hot end working! Does it mix the colours better than the Diamond did?

Edit: I just realised you posted two videos yesterday, not one, and you showed the mixing in the first one.

deckingman

@dc42 Better yes, but still some way to go. Comparison and fuller explanation in the first video I linked to - this one https://www.youtube.com/watch?v=VQdoTJMsHGo

Dougal1957

Ian

Have a look at https://www.watercoolinguk.co.uk/p/XSPC-G14-Plug-with-10k-Sensor-Chrome_73938.html

deckingman

@Dougal1957 said in My 6 input (5+1) mixing hot end:

Ian

Have a look at https://www.watercoolinguk.co.uk/p/XSPC-G14-Plug-with-10k-Sensor-Chrome_73938.html

Cheers Doug. Not sure where I'd be able to fit that particular one. But waterproof NTC thermistors are not hard to find. Something like this might be a better option for me https://www.amazon.co.uk/SODIAL-thermistor-temperature-sensor-waterproof/dp/B01EIYCW5O

Dougal1957

@deckingman I used something like https://www.watercoolinguk.co.uk/p/BitsPower-T-adapter-3x-IG14-inches-matt-black_57666.html with a barb at each end is currently unused mail me your address and I'll send the set down to you.

Doug

deckingman

It's been a while but here is a bit of an update is anyone is interested.

https://www.youtube.com/watch?v=Ck5wn1twh2M

o_lampe

@deckingman
looks very good, can't wait to see it print.
Maybe I'm wrong, but there seems to be no nozzle-tip? Aren't you afraid the block will scrape along the fresh print and collect more and more filament?

deckingman

@o_lampe said in My 6 input (5+1) mixing hot end:

@deckingman

....... Maybe I'm wrong, but there seems to be no nozzle-tip? Aren't you afraid the block will scrape along the fresh print and collect more and more filament?

That made me chuckle. I thought it was fairly obvious that the base had an M6 tapped hole in it to take any standard threaded nozzle (I just hadn't got around to screwing one in when I took those pictures).

deckingman

Part 16 is now on YouTube if anyone is interested
https://youtu.be/h90vsaQSSj4

T3P3Tony

@deckingman thanks for the update. Its obviously disappointing that your mixing chamber had the disadvantages you have highlighted. I feel for how much time and effort you have put into the research, and in letting us "follow along" with the blog and now video updates. I think you have done a really valuable thing in documenting this, and as you say it will provide a start point for anyone who wants to try and work around those disadvantages.

Things like a constant pressure feed with nozzle flow controlled by a metering valve, active mixing to have a smaller chamber etc have been talked about in the past but are really hard challenges to solve with a hot plastic under pressure!

dc42

@deckingman, you have achieved a lot, even if you don't think you have been successful overall.

Did you see the message I sent you on Slack? I think you might be able to use the 6th input like a plunger, to more rapidly alter the pressure just before the nozzle and so handle short moves better.

deckingman

@T3P3Tony @dc42 . Thanks. I'm not overly disappointed. If it was easy to do, someone would have done it already. As you say, knowing what doesn't work (and the reasons for it) is almost as valuable as knowing what does work. I've learned a lot along the way and any knowledge is good IMO.

Also, I've ended up with a multi-input hot end that works better than anything I can buy. It might not be good at mixing but it has 6 inputs so I can print 6 colour objects with ease, I can use any M6 threaded nozzle, I can change that nozzle quickly and easily, so I can print with "exotic" or abrasive filaments, it's smaller than a 5 colour Diamond and virtually silent. So despite the fact that I never fixed the mixing thing, I have gained some nice improvements.

@dc42 I hadn't seen your message but I have now. I did try using up to 20% clear filament in that 6th input, mostly in the hope that it might "mask" the incomplete mixing to some degree. I can't say that I noticed any improvement. I suspect the only way it would work as you describe is if the 6th input used a much higher percentage of the whole. But even then, although that 6th input is a straight path to the nozzle, it's still a "tube" of molten filament about 70 mm long, and I doubt that the filament exiting that tube would behave in the same "dynamic" way as the cold rigid filament entering the top of that tube in what is effectively a series of varying pulses.

T3P3Tony

@deckingman crazy idea. What if that filament (as clear filament, as solid as possible) was used as a plunger to start and stop the flow out the nozzle?

So the path would need modification so it only entered the hot block as close to the nozzle as possible. It would be used as a sacrificial plunger to block the main melt chamber ooze, then retracted when printing continued. Obviously it would need different logic from a standard extruder to control, probably something based on the amount of force needed to keep the nozzle mostly block as it melted away.

Taking it one step further, if it worked well, it might be able to be used as a metering valve and get over the resolution issues. There would be a head of static pressure in the melt chamber which is topped up by the incoming filament, but released by the metering filament movement.

Even if that all worked perfectly! The big issue of course is still purging.

T3P3Tony

a chemically compatible but higher melting point filament would even further reduce the amount of "contamination" that this sacrificial plunger added to the output. it would still need to be slightly soft at printing temperatures in order to get a good seal.

PaulHew

Ian, I watched the print of your mixing hotend.
Incredible and fascinating and no purge block. (I did listen to how you purge)

I thought your machine was incredible when I saw it at TCT a couple of years ago but the 5+1 takes the prize.
Clever man.

P.

o_lampe

@deckingman
..that was heartbreaking. I hope you'll find something else worth to chase.

Based on your description of disadvantages I envisioned a mixing hotend with only one transparent filament, where the dry ingredients that are used by the filamentmakers. (pigments) are mixed together just before they enter the melting zone. (Cyan, magenta, yellow, black would be perfect)

It would probably require a stirring rod as other have used before, but oozing and slow spongy reaction wouldn't be such a problem anymore.
Purging would be no big issue, since you'd mix the colour on demand. Only sharp colour changes would need a purge tower.

No idea how to prevent the filament from climbing into the dry-mixing chamber, tho...it'll stay a dry dream I'm afraid.