So I have my first try on this thread:
https://forum.duet3d.com/topic/16244/watercooling-heat-chamber-tool-changer-high-temp-3dp-build

But I'm abandon tool-changer design as I found myself not able to trust the tool-changer to have it run un-attended. And the printed locking mechanism is another concern.

So, instead of spending time and effort to improve reliability of tool-changer, I decide to go back to dual-X Cartesian design.

I just finish the design stage and waiting for 30mm extrusion to arrive during the weekend, then I can start building the 2nd attempt.

Titus

I been working on this since Dec,2019. I been asked to share the build, I think here is perfect place to do the build log.

Goal:
2 high-temp hotends capable to reach 500'c
Heat chamber (in future) at 80'c

At design concept stage I went through several try & error, mainly...

1. Dual H-bot <abandon on drawing board, due to kinematic support>
2. H-bot tool-changer <abandon after testing, due to racking>

At the end I went to common CoreXY with tool-changer.

--

Main components:

Duet2 Wifi + Duex5

5 linear rails are made by SURUGA Production, which are special 150'c high-temp-cert variant (and surprisingly just slightly more expensive then room-temp one).

E3D Titan Aqua watercooling hotend

E3D 300x200mm heated bed

Watercooling heat-exchange (dual 120mm fan) and D5 pump with 200ml reservoir

--

Since I want heat chamber to go 80'c, I purposely move all electric components outside the chamber zone, with 2 exceptions: extruder stepper and BLtouch.

Extruder stepper is the main reason why I went for E3D Titan Aqua watercooling hotend.

BLtouch, currently mount as part of T0. I'm going to move it to T2 as dedicated "utility tool", doing nothing except probing, and have it parked at enclosure-inside-enclosure to keep it cool.

--

Watercooling, I went super-oversize with dual 120mm fans heat-exchange. This is mainly because I've hard time to estimate how much heat I need to deal with. Therefore I go with the simplest solution: GO BIG!

--

Tool-changer, particularly the locking/unlcoking mechanism, turn out to be the most challenge part. The E3D-TC-style stepper-direct-gear-drive is not an option since I don't want to do another watercooling loop only for tool-changing stepper. After many try & error, I end up with old-school bicycle shift cable push-pull cable system.

And it drive a very simple spring + thrust bearing shaft, secure directly to the locker.

The tool parking mount, is using ISO30-tool-holder-style "ring", as I found it's much easier than E3D-TC-style design.

--

Z-axis, the simple 3 steppers so I can have auto-leveling. As right now it's simple mounting structure, but I should have enough space to put kinematic coupling in case heat-expansion become a problem.

Since it has 250mm Z travel, the free-rotating-end must be secured with bearing, otherwise you will have lovely "music" playing at different speed, as the leadscrew is long enough to vibrate and making music(noise).

--

Followings are what I haven't build/do:

Nozzle brush
Probably build directly under the parking location

Upgrade to 9mm belt
Probably with some toothed idler

Belt tension adjustment
Probably will build at left-hand-side frame

End-stop on C-axis
Right now the tool-changing C-axis doesn't have any end-stop. The "home" is preset by hand. A simple blocker should be enough for stall-homing, will findout.

Parts cooling
I don't think those high-temp filament need any parts cooling, but I'm building it so I can still print PLA/PETG etc. Again, a direct-fan-cooling is not a good option. The smallest 80'c-rated fan I can find is 60x60mm which is too big to mount. I probably will go with air-compressor parts cooling, challenge is finding the air compressor that take PWM signal for airflow control.

Electronic box
Right now all electronic was open-mounted. Need to build a box to safeguard those, plus some fuse to be safe.

BLtouch on Tool2
As said previously, I'm going to move BLtouch from T0 to T2, and have T2 parked in enclosure-inside-enclosure to keep it cool. Doing T2 is simple, find a way to allow auto opening/closing door is the challenge, as I don't want another stepper only for opening/closing the door.

Chamber heater
This should be straight forward, but I leave it at the very end so I can keep the frame open for anything I need to build, before seal the chamber.

--

Known problem...

The 5mm thickness of PETG is still not stiff enough for kinematic coupling on the tool-changer. Once the tool been locked, it sunk ~0.4mm. I'm going to make at least the main plate with aluminum, then see how it goes.

--

I hope this inspire others to do their owned high-temp 3DP

Titus

=====
NOTE (Jun18,2020)

I'd abandon this design and going back to dual X Cartesian. You can find my 2nd attempt here:

https://forum.duet3d.com/topic/17078/watercooling-heat-chamber-high-temp-3dp-build-2nd-attempt

Mechanical is most done now...

Y assembly (up side down)

Front view

Electronics and water cooling side cabine

Yes that's Misumi cardboard box as temp panel

And yes, this has 4 Z motors, as I figure it's easier/cheaper for me to get the same rigidity with 4 Z motors on each corner, than other 2 motors design which all require stiff Z frame for structure support.

I need to find a way to do Z calibration as I don't want to have BLtouch seat in the heat chamber. Maybe I will do a remove-able Z-probe bar to use flat top as reference point, we will see.

Until next time
Titus

Gauge indicator is super fast to use as Z frame leveling probe bar, then just use hand to rotate the Z motor leadscrew to make Z frame level on 4 corners.

And then, the 2 screw detachable BLtouch mount for G29 meshbed leveling.

Achieve 0.020mm mean error, HAHA!!!

Next... start the X/U calibration...

Until next time
Titus

@infiniteloop I got it works as you described. Thank you!!!

Some detail for future reference.

Duet [PANEL DUE 5V_EXT] ==(orange)== Relay [JD-VCC]
Duet [PANEL DUE 3.3V_EXT] ==(green)== Relay [VCC]
Duet [PANEL DUE GND] ==(black)== Relay [GND]
Duet [out3-] ==(yellow)== Relay [IN1]
Duet [out4-] ==(blue)== Relay [IN2]
Duet [out5-] ==(brown)== Relay [IN3]

config.g

; Laser mode
M452 C"out6" R255 F200 ; Enable Laser mode, on out6, with max intensity being 255, and a PWM frequency of 200
 
; FAN-as-IO setup
M950 P0 C"out3" ; Red Laser
M950 P1 C"out4" ; Laser Enable
M950 P2 C"out5" ; placeholder

No strange behavior during power on nor power lost. Everything looks good.

Note: the board next to relay is isolated PWN to 0-10V converter, in case someone wondering.

@Googliola I got this fixed by rewiring as what @droftarts suggested: H0 bed, H1 tool0, H2 tool1, H3 chamber.

Yep I tried the same "disable heater", no joy. Do the rewiring with bed>tool>chamber sequence. Seems the only way. Not idea but at least it works.

I got all the wiring semi-done now. Need to redo X/U with drag chain to make it cleaner.

Cooling loop is using customized alum distribution block I'd from other projects, Probably oversized but it seems works. Yes it has 3 loops, one reserved for Y motor later. Haven't run the pump yet.

After lots of try&error I managed to get 4 independent Z motor running G32 auto bed calibration with "probe bar". It's working but quite hassle to use as I also need to put 50mm block during the probing.

Basically I use M291 message with input-response to control step flow. Moving XY to Z1 motor location > place probe bar + block > probe > remove probe bar + block > Z2 again > Z3 again > Z4 again. It take quite a while to finish.

I will see if using traditional gauge indicator will make G32 operation easier. Since bed is quite ridge I don't think I need to use G32 frequently.

And then... put BLTouch on X for G29 meshbed, maybe make BLTouch detachable to avoid long-term high-temp chamber exposure.

Until next time
Titus

@Phaedrux said in Watercooling, heat-chamber, high-temp 3DP build (2nd attempt):

Would you be able to share your config set? It may prove helpful for new users trying to setup an IDEX.

Will do once I get it cleanup a bit, it's quite a mess right now

Titus

@infiniteloop I got it works as you described. Thank you!!!

Some detail for future reference.

Duet [PANEL DUE 5V_EXT] ==(orange)== Relay [JD-VCC]
Duet [PANEL DUE 3.3V_EXT] ==(green)== Relay [VCC]
Duet [PANEL DUE GND] ==(black)== Relay [GND]
Duet [out3-] ==(yellow)== Relay [IN1]
Duet [out4-] ==(blue)== Relay [IN2]
Duet [out5-] ==(brown)== Relay [IN3]

config.g

; Laser mode
M452 C"out6" R255 F200 ; Enable Laser mode, on out6, with max intensity being 255, and a PWM frequency of 200
 
; FAN-as-IO setup
M950 P0 C"out3" ; Red Laser
M950 P1 C"out4" ; Laser Enable
M950 P2 C"out5" ; placeholder

No strange behavior during power on nor power lost. Everything looks good.

Note: the board next to relay is isolated PWN to 0-10V converter, in case someone wondering.

@T3P3Tony Still acting odd...

I now have...

No external 5V PSU, all power by Duet's 5V.
Duet [PANEL DUE 5V_EXT] === Relay board [VCC], which jumper connect to [JD-VCC] to power coil
Duet [GND] === 24V PSU [-V] === Relay [GND]
Duet [out3-] === Relay board [IN1]
Duet [out3-] === 4.7K Ohm === 24V PSU [-V]

And relay enabled as soon as power on.

I check the relay board's schematic, and I think it's neg-side-controlled. So having pull-down on the [IN1] actually ground the [IN1] and cause it enable instantly at powerup.

So if that the case, having pull-down on the [VCC] is actually kind correct?

@T3P3Tony I'm currently have 5V PSU connected as isolated supply to the relay coil

So on the signal side there is really no GND.

Is this going to cause some odd problem? Draining the board instead of pulldown signal?

@T3P3Tony I managed to get this work by...

[PANEL DUE 5V_EXT] === Relay board VCC
[PANEL DUE 5V_EXT] === 4.7K Ohm === [24V PSU -V]

If I do...
[out3-] === Relay board [IN1]
[out3-] === 4.7K Ohm === [24V PSU -V]
The relay will enable when powerup, and remain on when power lost for a long time (probably until PSU completely dry)

Did I do something odd?

Thanks in advance!

@T3P3Tony
You mean in series between Duet and relay board?

Add note...

I've 24V and 5V PSU -V connected together in the previous testing.

I also did following testing:

---

##Not using external 5V PSU##
[PANEL DUE 5V_EXT] === Relay board VCC
[out3-] === Relay board [IN1]

[PANEL DUE 5V_EXT] === Relay board JD-VCC (powering relay coil with Duet's onboard 5V)
[PANEL DUE GND] === Relay board [GND]

---

##Isolate 5V PSU##
Not connecting 24V and 5V PSU's -V together.

[PANEL DUE 5V_EXT] === Relay board VCC
[out3-] === Relay board [IN1]

External 5V PSU [+V] === Relay board JD-VCC
External 5V PSU [-V] === Relay board [GND]

---

In both scenario, at power lost, there is ~2.7V between [out3-] and [PANEL DUEL 5V_EXT], same as original setup.

I'm bench-testing my laser rig using Duet3 Mini5+.

I've...
[out6_buffer] === PWM-to-Analog converter for 0~10V laser power single

[PANEL DUE GND] === Relay board [GND]
[PANEL DUE 5V_EXT] === Relay board [VCC]
[out3-] === Relay board [IN1]
[out4-] === Relay board [IN2]
[out5-] === Relay board [IN3]
[External 5V] === Relay board [JD-VCC], to power the coil

And in config.g I've...

; Laser mode
M452 C"out6" R255 F200 ; Enable Laser mode, on out6, with max intensity being 255, and a PWM frequency of 200

; FAN-as-IO setup
M950 P0 C"out3" ; Red Laser
M950 P1 C"out4" ; Laser Enable
M950 P2 C"out5" ; placeholder

The bench-testing (not wiring to actual laser) seems ok. M42 P0 S1 enable IN1 relay, and M42 P0 S0 disable IN1 relay. P1/P2 also doing the same thing, which is good.

However, at power lost (ex: unplug Duet3), between [out3-] and [PANEL DUE 5V_EXT] will jump from 0V to ~2.7V and slowly bleeding away. While it's not enough to trigger relay, it's there. Same case for [out4-], [out5-].

Is there way to ensure that, at power lost, it stay at 0V ? As I plan to switch to MOSFET later.

---

(edit: better writing to avoid confusion)

Change the subject for better searching
Titus

@Phaedrux said in Watercooling, heat-chamber, high-temp 3DP build (2nd attempt):

Would you be able to share your config set? It may prove helpful for new users trying to setup an IDEX.

Will do once I get it cleanup a bit, it's quite a mess right now

Titus

During calibration testing I figured the temp X/U wiring is really too temp to use even with simple calibration. Solution: get it drag-chained

Until next time
Titus