Interim dual extruder proposal

While RRP's multi-extruder kit should be very nice, with (as far as we can tell) at least three extruders and possibly support for mixing, there is no news on when it will be available. There is the 4-nozzle Kraken head, but that is expensive and needs a lot of cables. So I'm looking at making an interim 2-nozzle head. My current thoughts are:

* Replace the aluminium heatsink block with a wider one. It would have 2 threaded holes for nozzles near the middle. The screw holes for attaching the hot end to the nozzle mount would be moved outside the screw holes used to hold all the hot end parts together (heatsink, fan etc.).

* Print a new wider nozzle mount with the screw holes in the new position.

* Fit 2 nozzles to the heatsink block, each with its own heater block, cartridge heater and thermistor. If one of the heater blocks is fitted upside down then I think there is plenty of room for this.

* Add a second extruder assembly and Bowden tube to feed the second nozzle.

* Use the Duet expansion board to run the second extruder stepper motor.

* The Duet expansion board can also drive the second heater and thermistor. However, another 6 wires to the hot end will be needed, assuming the heater wires are doubled up again. Including 4 wires for the z-probe, that's 18 wires in total from the Duet and the expansion board to the hot end. As I already have a microcontroller on the hot end, I'm thinking of doing a variant of my differential IR sensor/fan control/lighting board to monitor 2 thermistors and control 2 heaters. If I run each heater at max 50% PWM and run them in antiphase, then the maximum current will be no more than it is with a single heater. So no power supply worries, and no need for more than 4 conductors in total to provide 12V power. I would use one of the Duet serial ports to send commands to the hot end board and data back. Total 8 wires (or possibly 7) to the hot end board instead of 18. This will allow me to use just the original main loom and dispense with the sensor loom.

Possible issues I foresee:

* Is the 40x20mm fan and heatsink sufficient to handle the heat from 2 nozzles? I'm fairly confident it will be when printing PLA, but it might be marginal when printing ABS. A larger fan might be needed.

* Can I get the 2 nozzles at the same height, to within better than 0.05mm? I guess this may be an issue for the Kraken as well. Once it is assembled, the easiest way to adjust it might be be to print different nozzle mounts with the screw holes for attaching it to the x-carriage at slightly different heights, or with slots.

* When changing colours, will there be a problem with the not-in-use nozzle oozing filament? I guess I could move the head off the edge of the bed while I cool one nozzle down and heat the other up.

* When changing colours, how can I get the new nozzle to start extruding cleanly? At the start of a print, printing the skirt does this - but what about a colour change in the middle of the print?

* What else haven't I thought of?

Comments and suggestions welcome!

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

A bit too complicated IMHO, would a delay in the color switching be a problem?, if not, my vision is more like shifting the filament on the cold side and keep the hot end as is

from the "Dual Extruders" tread
[forums.reprap.org]

Erik

Hi DC, I've been thinking of making a dual extruder myself.
Here was my plan, but I haven't had time to work on it due to other matters.
1) Replace the Heater Block, with a wider block, but unlike your idea, have a single heater. (PLA from a single supplier I've found is close enough to use a common Temp.
2) Replace the orginal Nozzle Mount with a wider one, to allow the fitting of the wider Heater Block.
3) Replicate the Extruder Feeder, as two feeders would be needed.
4) Use a AND gate, and a Darlington array to switch between Feeder stepper motors, rather than use an expansion board... (the AND Gate would be linked to the Duet OutPuts to the Feeder, AND a signal Pin on the Duet to select which
feeder was active). Maybe you might be able to pick the Bones out of this idea, the Pros and Cons and if it will work. The other way would be to use an Expansion Board, but I was hoping to keep costs down.
Basicaly that's it....

The Heater Block could have two Nozzles fitted, or be drilled to allow mixing. I think the mixing option should wait until the twin nozzle approach was tested.

Well that was my plan... Which I had hoped to have running by now, but, lifes rich pattern and all that, and being 400 miles away from the Miller doesn't help.

It would be interesting to hear your views as to the AND gate Idea. I would have thought it would work, but your much better at that sort of thing I think.
In the mean time, I did see a small stand alone stepper controller available on one of the sites... Maker Bot I think, that would be another way of cracking the nut.
I think they were selling them as replacements for their machines at about £20 a shot, complete with heatsink.

Kim

Quote
dc42
While RRP's multi-extruder kit should be very nice, with (as far as we can tell) at least three extruders and possibly support for mixing, there is no news on when it will be available. There is the 4-nozzle Kraken head, but that is expensive and needs a lot of cables. So I'm looking at making an interim 2-nozzle head. My current thoughts are:

* Replace the aluminium heatsink block with a wider one. It would have 2 threaded holes for nozzles near the middle. The screw holes for attaching the hot end to the nozzle mount would be moved outside the screw holes used to hold all the hot end parts together (heatsink, fan etc.).

* Print a new wider nozzle mount with the screw holes in the new position.

* Fit 2 nozzles to the heatsink block, each with its own heater block, cartridge heater and thermistor. If one of the heater blocks is fitted upside down then I think there is plenty of room for this.

* Add a second extruder assembly and Bowden tube to feed the second nozzle.

* Use the Duet expansion board to run the second extruder stepper motor.

* The Duet expansion board can also drive the second heater and thermistor. However, another 6 wires to the hot end will be needed, assuming the heater wires are doubled up again. Including 4 wires for the z-probe, that's 18 wires in total from the Duet and the expansion board to the hot end. As I already have a microcontroller on the hot end, I'm thinking of doing a variant of my differential IR sensor/fan control/lighting board to monitor 2 thermistors and control 2 heaters. If I run each heater at max 50% PWM and run them in antiphase, then the maximum current will be no more than it is with a single heater. So no power supply worries, and no need for more than 4 conductors in total to provide 12V power. I would use one of the Duet serial ports to send commands to the hot end board and data back. Total 8 wires (or possibly 7) to the hot end board instead of 18. This will allow me to use just the original main loom and dispense with the sensor loom.

Possible issues I foresee:

* Is the 40x20mm fan and heatsink sufficient to handle the heat from 2 nozzles? I'm fairly confident it will be when printing PLA, but it might be marginal when printing ABS. A larger fan might be needed.

* Can I get the 2 nozzles at the same height, to within better than 0.05mm? I guess this may be an issue for the Kraken as well. Once it is assembled, the easiest way to adjust it might be be to print different nozzle mounts with the screw holes for attaching it to the x-carriage at slightly different heights, or with slots.

* When changing colours, will there be a problem with the not-in-use nozzle oozing filament? I guess I could move the head off the edge of the bed while I cool one nozzle down and heat the other up.

* When changing colours, how can I get the new nozzle to start extruding cleanly? At the start of a print, printing the skirt does this - but what about a colour change in the middle of the print?

* What else haven't I thought of?

Comments and suggestions welcome!

---

Please send me a PM if you have suggestions, or problems with Big Blue 360.
I won't see comments in threads, as I move around to much.
Working Link to Big Blue 360 Complete

Hi Kim,

Thanks for your comments.

The problem with your and-gate idea is that the stepper motor windings are driven in both polarities. So the switch needs to be bidirectional. To stand any chance of getting to to work at all, you would need at least 4 darlingtons and 4 diodes. Alternatively, a DPDT relay should work. However, the Duet expansion board will drive 4 more extruders, so it will be ready for when RRP release their upgrade kit. At GBP40 inc VAT, I think it is good value.

I wanted separate heater controls for the two nozzles because I've observed that an unused nozzle bleeds filament, even if the filament has been retracted by 5mm.

The single nozzle with mixing sounds nicest, but I think it needs a specially drilled piece of PTFE or PEEK to make it work.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I like the idea from Dr Adrian Bowyer so that the hot end stays the same but the filament is changed, this can be done by retracting the old and forwarding the next colour via his "train track idea"
I have provisionally started to redesign the feeder so that the paths for each filament are identical and to provide mounting via the existing hot end attachment screws.

This could be the cheapest and potentially the best solution, no oozing from not in use nozzles, no change to power requirements, no expansion boards required etc.

The only drawback that I can see is retracting the filament cleanly without causing any tails or blobs

Even if tails resulted, as long as they were not too long then on change of filament an extra amount would need to be extruded away from the printed part until normal flow was attained.

Whatever we do we need to make sure that the setup and maintenance is minimal.

Time is very limited at the moment and I will not be around for the next month so I haven't progressed very far, It will be interesting to see what you all come up with.

It will be exciting having multi colours.

Thank you for all your efforts
Paul

---

appjaws - Core XYUV Duet Ethernet Duex5
firmware 3.1.1 Web Interface 3.1.1
Ormerod 1-converted to laser engraver, Duet wifi
OpenSCAD version 2020.07
slic3r-1.3.0, Simplify3D 4.1.2, Cura-4.4.1

Hmmmm I was thinking the expansion board was going to be about £100+ for some reason, so I agree it makes alot of sense to go that route.
Also the mixing idea would be great as you wouldn't be stuck with two colours, and have a choice of a blend if you wanted it.
You've got me thinking about the heater block/nozzle arrangement thou... I did start thinking that feeding the two filaments in from either end might work, but then
it's likely that the filament going in one side would tend to try to go up the other side, as well as the nozzle. So I think the way to go is a shortend J Nozzle, which goes into
the Heater Block. Then the top of the Heater Block then having a screw in 'Y' connection. The arms of the 'Y' could be lined with plain PTFE tubing, and the Heat Block would
be a Heated Mixing Pot before the Plastic comes out the Nozzle. Provided the 'Leg' of the 'Y' was "in the heated" area, I would have thought it would work a treat.

A shortend J Nozzle would have to have an internal chamfer (lead in) for the plastic, a shoulder would cause problems, but that's nothing to turn up in a Lathe.
The 'Y' could even be a regular Brass 1/8" gas fitting (or smaller if there is such a thing), with the 'arm' of the 'Y' carefully drilled to accept a suitable piece of PTFE, and the 'arms' clamped to a heat sink maybe.
But there again, if we accepted that the arms could be filled with a mix of molten plastic as well as solid filament, do we really need a HeatSink at all?
Provided the PTFE Feeder tube was a fairly close fit to filament nominal size, I can't see it making a differance, as to where the actual point of melting is, as if it's above the HeaterBlock.

---

Please send me a PM if you have suggestions, or problems with Big Blue 360.
I won't see comments in threads, as I move around to much.
Working Link to Big Blue 360 Complete

One of the things I like the most about dual extruders is the ability to print all complicated support with the water soluble PVA - that wouldn't obviously work with a single nozzle.

I've seen videos of a 3 extruder setup with a single nozzle, and it looks very interesting, but what is worth noting is that the colors don't *mix* per se, but rather come out like some brands of toothpaste!
This gives a very interesting effect on your print in that the color is different when looking at it from different directions.

My thinking on this runs as follows:
1. Single hot end with 'train track' junction for the two filaments
2. Spring loaded extruder already shown to work
3. Two sides of the hobbed drive are available

So we require:
Two Bowden tubes leading to the new Y junction at the hot end
Single extruder, with some magic DC firmware to reverse direction depending on whether the A or B side is selected
Pivoted dual filament body which holds either the A or the B filament against the hobbed drive, and the B or the A filament pressed against a brake - plus a slot for an eccentric drive
New eccentric drive to hold the pivoted filament body against one side or the other - probably with a spring involved.

Maybe some variation on a two position switch could be used - they seem to be spring loaded
Greg

Edited 1 time(s). Last edit at 06/08/2014 04:45AM by GregL.

---

Ormerod #17

Quote
GregL
Pivoted dual filament body which holds either the A or the B filament against the hobbed drive, and the B or the A filament pressed against a brake - plus a slot for an eccentric drive
New eccentric drive to hold the pivoted filament body against one side or the other - probably with a spring involved.

How would the eccentric drive be controlled? If you're using another stepper motor for that, you might as well just have a whole separate extruder. The stepper motor is the biggest cost, weight and space of the extruder-feeder assembly anyway.

I agree that another stepper might just as well be used as a separate extruder.
Something smaller should be sufficient - like a two position spring loaded actuator.

I was rather hoping DC, or someone else, would come back and say ' we can use ...'

---

Ormerod #17

I like the idea of a device to push the filament against the hobbed insert, or not. I find that after my printer has sat for a few hours with the filament in the same place, the slightest movement tends to snap the filament inside the extruder. Relaxing the pressure pushing the filament against the hobbed insert might avoid this.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42
I like the idea of a device to push the filament against the hobbed insert, or not. I find that after my printer has sat for a few hours with the filament in the same place, the slightest movement tends to snap the filament inside the extruder. Relaxing the pressure pushing the filament against the hobbed insert might avoid this.

I've also found that the filament snaps if left, I might start retracting it after each use now.

---

Another RS Ormerod Mk1 meets the world

Retired now but I used to make....
CNC Machined Mk1 aluminium bed support plates for the Ormerod
CNC machined X-plates and ribs for Mk1 & Mk2 Ormerods
CNC machined bed support arms for the Mk2 Ormerod.
Dual Hot-End heatsink blocks.

Hmmm I sort of follow the drift of this, but not totaly sure of it, but if your wanting to Open and Close a sprung Feeder, then why not use a servo.
It wouldn't have to be big, as instead of an arm, you'd use a Cam.... That way it reduces the loading on the servo.. Or have it drive a screw to lift the sprung arm....
Even a micro servo can move quite a load if setup right, just won't move it far at the same time, without modding it.

Off Topic....I've even made Micro Radio Controlled Boats using Modied Servos (taken apart and stripped) as the Main Engines...
It's supprising what you can do with imagination.

If I keep the printer I might buy the Expansion Board and start tinkering.... Especialy if the weathers rubbish, and no good for flying.

---

Please send me a PM if you have suggestions, or problems with Big Blue 360.
I won't see comments in threads, as I move around to much.
Working Link to Big Blue 360 Complete

I have been planning to look into a multi-extruder upgrade for some time. I got so far as to buy the duet extension board and two extra extruder-kits. I also had a three-nozzle mounting-bar made. So basically all the hardware is ready to be mounted - just a bit of cables to be made.

Unfortunately I do not have the time for the software at the moment, so the project got stranded. However, I would love to participate in a multi-extruder project.

I have access to a thermal camera - incredible tool for thermal debugging :-)

Here's a very inspiring blog - rebuilding a Stratasys extruder. [haveblue.org]

Best regards,

Carsten

Quote
dc42
I like the idea of a device to push the filament against the hobbed insert, or not. I find that after my printer has sat for a few hours with the filament in the same place, the slightest movement tends to snap the filament inside the extruder. Relaxing the pressure pushing the filament against the hobbed insert might avoid this.

Has that happened to you with different kinds of filament? I have had a few different kinds (generally "whatever's cheapest" from a couple different resellers) left for a day or two, and never had filament snap in the extruder. Or might it be that the extruder is just a lot tighter or something?

Quote
jstck

Quote
dc42
I like the idea of a device to push the filament against the hobbed insert, or not. I find that after my printer has sat for a few hours with the filament in the same place, the slightest movement tends to snap the filament inside the extruder. Relaxing the pressure pushing the filament against the hobbed insert might avoid this.

Has that happened to you with different kinds of filament? I have had a few different kinds (generally "whatever's cheapest" from a couple different resellers) left for a day or two, and never had filament snap in the extruder. Or might it be that the extruder is just a lot tighter or something?

I've had it happen with 3 different spools of PLA, but they were all from the same supplier. I can't remember if it ever happened with the original spool supplied with the kit, because I used to assume that the cat was responsible until I saw it happen spontaneously.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hi all

Like every one else i have been working on a multiple head mod for the ormerod. One of the resons for my water cooled hot end was to overcome the size limitations of the fan and heat sink (btw for those working with abs a water cooled hot end needs no fan). My concept and dont laugh is to fix each hot end to a curved support so that when not extruding the surplus nozzles are up and away from the print, a bit like the multiple lenses on a compound microscope where you rotate the lense you need in to position.

The reason i think this would be benifitial is:

Only the active hot end can interfear with the print

Your total print size is not reduced by the distance between nozzles

There is still only one point of extrusion to have to cool (PLA)

One steper could be used for multiple extruders (this would require a stepper on the X carrige)

The hot ends not in use could rotate up behind a cover to both help maintain heat (less power) and prevent dribles

Challenges with this concept are:

The added waite of a servo and the extruder stepper on the X carige

The need to run a heating eliment and thermoresistor for each hot end

The need for accuracy when rotating the active hot end into position (eg some sort of snap to arangment)

I am not shore weather the software can be adapeted to only use one extruder, not require an offset for multiple heads and weather a PWM signal can be use to signal a swap between filements.

The last point is my biggest hurdle since i have only a passing knowledge of programing.

(off topic is there a spell checking function on this fourum)

Quote
JMC
My concept and dont laugh is to fix each hot end to a curved support so that when not extruding the surplus nozzles are up and away from the print, a bit like the multiple lenses on a compound microscope where you rotate the lense you need in to position.

An interesting idea, but it is beyond my mechanical design skills to design something like that.

Quote

Challenges with this concept are:

...

The need to run a heating eliment and thermoresistor for each hot end

Alternatively, move the temperature controller to the hot end, as I proposed for my dual-nozzle concept at the start of this thread.

Quote

I am not shore weather the software can be adapeted to only use one extruder, not require an offset for multiple heads and weather a PWM signal can be use to signal a swap between filements.

Yes, the software could be adapted for use a single extruder for multiple filaments. There are plenty of spare outputs on the Duet that could be used for switching between filaments.

Quote

(off topic is there a spell checking function on this fourum)

I use Google Chrome as my browser, and it spell-checks my replies.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
VortyZA
One of the things I like the most about dual extruders is the ability to print all complicated support with the water soluble PVA - that wouldn't obviously work with a single nozzle.

I don't see why not. You would need a sacrificial column somewhere where a few mm can be extruded after changing material on each layer to clear the hotend of the previous material. If the slicer software cannot support such a column, I am sure it could be put into firmware - just add a command to notify the firmware of the place and size of two columns. The slicer software would need to print all of one material on each layer before switching to the other material, and the firmware could then cause the head to go to the relevant column and extrude a nozzle's worth of material after each change.

Dave
(#106)

Quote
dmould

Quote
VortyZA
One of the things I like the most about dual extruders is the ability to print all complicated support with the water soluble PVA - that wouldn't obviously work with a single nozzle.

I don't see why not. You would need a sacrificial column somewhere where a few mm can be extruded after changing material on each layer to clear the hotend of the previous material. If the slicer software cannot support such a column, I am sure it could be put into firmware - just add a command to notify the firmware of the place and size of two columns. The slicer software would need to print all of one material on each layer before switching to the other material, and the firmware could then cause the head to go to the relevant column and extrude a nozzle's worth of material after each change.

Dave
(#106)

I suppose you could do that, but I believe that the PVA prints at a different temperature to the PLA - could be wrong. That would make for a VERY long print!

I go for your proposed dual-extruder option, dc42, as a single nozzle version might cause mixing of materials as there's a chance that the oozing of one material might get mixed into the other. If you were to go for a quad-extruder, I'd guess that one of them would be for printing support/another material, and the remaining three might work for printing multi-color through a single nozzle.
If you end up making a kit for this, sign me up for one!

Hi Onno, thanks for your interest. If I go ahead with the plan to control the heaters on the hot end board, then when I order a PCB for it I have to order 10, so I can easily make a second controller board. I've started on the design work already. DaveK has kindly agreed to make me a prototype 2-nozzle heatsink block.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

The way the Ormerod has the nozzle mounted, might make aligning them a nightmare. The J-head type nozzles might be easier to get straight onto the x-carriage, but would require an overhaul of the nozzlemount and cooling. I'm not sure if you've seen the nozzle cooling on the mini-kossel (I happen to have one in front of me), but it sort of washes the J-head nozzle in air, at the same time cooling the filament extruded. It might be an alternative worth looking into

Hi All

My idea was to use the expansion board from Think3DPrint3D (Duex4) for the extra extruder drivers, as it was designed for.
Then to make a 'T' shaped mounting bracket to replace the X-carriage hotend mount.
There are already a few replacement parts, KimBrown and PhilipTheMobster have Thingiverse replacements for this mounting block.

That way I want to try and mount two of the existing hotends and fans etc back to back.
Already have the driver as I have printed a spring-loaded extruder as a replacement.

Alternatively, RepRapPro have already sorted the multi hotend problem on the Mendels, all thats required is to adapt the x-axis mounting to suit an Ormerod.
Print volume will obviously be smaller when using it in multi mode.

Still going to follow this thread to see if any better ideas emerge.