Trying something...  Yippee   At the bottom of the photo is a foot steel rule (30cm) to show scale.

New power wiring diagram with 24v PSU.  12v for the water pump, fume extractor fan and LED lighting strip is obtained from the 24v supply via a buck converter.  The 24v PSU provides power to the Duet WiFi controller board which controls all printing functions including the hotend heater, parts cooling fan and heated bed via an SSR (Solid State Relay) from the 240v AC mains supply, plus water pump, fume extractor fan and LED lighting strip.  I would have liked to have controlled the water pump and extractor fan separately but the Duet only controls 3 fans (plus 2 always on).

Looking at a box for Duet WiFi control board.  Since I'm running the Z drive at 2.4A (the maximum the Duet allows) I think a cooling fan might be a good idea for the Duet board.  A 30mm or 40mm fan directly behind the Z driver chip should be sufficient.  It will need a design change for the box.

The Duet cooling fan can be run in parallel with the water pump and fume extraction fan.  If I had the remote power off function I originally thought of all these could have been "always on" except when the power was turned off.  To control them now they must be connected as a controlled fan and turned on at the start then turned off at the end of a print run.

I think I shall have two 3D printed boxes, one for the Duet control board with cooling fan and another to contain a mains switch, bed heater SSR and the buck converter.  May arrange the second box to cover the end of the 24v PSU where the terminals are and have all dangerous voltages safely enclosed.  The mains wires to the bed heater pad are well insulated and the aluminium bed plate is earthed to the frame (and frame connected to mains earth).  The parts enclosed by the second box are shown with a dashed red line.

Been reading about fan controls on the Duet WiFi control board.  Fan0 is used to control the Parts Cooling Fan as usual but the water pump, which corresponds to the hotend heatsink cooling fan can, with advantage, be run from Fan1 which defaults to a thermostatically controlled fan operated by the hotend temperature.  The result is that the fan will always be when the printer is switched on and will remain on when the hotend temperature if above the set value.  With subsequent print runs the fan comes on when the hotend reaches the set temperature in the fan control GCode.  The fume extraction fan can also be controlled by Fan 1 as fumes are only generated when the filament is hot.  That leaves the Duet cooling fan and the lighting with just one controlled fan left.

If the remaining Fan control is allocated to the lighting, that leaves the Duet board cooling fan which is required if the Z drive motor is on.  I'm reluctant to have the board cooling fan running continuously, though a larger fan running at reduced speed should be both quieter and longer lasting.  Anyway, I would rather have this cooling fan on whenever the Duet power is on to prevent the possibility of the Z drive being used without the fan running.  Another possibility might be to use a larger (and therefore more efficient) Nema 17 motor on the Z drive so that the Z motor driver could be run at less than 2A and not need the chip cooled.  I would prefer to reduce the motor current anyway regardless of cooling needs.

My board cooling fan is an 80mm PC type fan and is always on. it's so quiet I have to look at it to check it is running.

It's one of These

Thanks Dave   That looks like a nice fan - think I'll get one.  The Duet Wiki says to cool the back of the board rather than the front though.

I can scrounge a long Nema 17 motor from my Titan printer when I rebuild it.  It uses 60mm Nema 17s for X, Y and Z drives and these big motors are not needed for the XY drives.  Think I'll use a cooling fan anyway.

I think I'll arrange that air from the fan can also get round the bottom of the board and up across the front as well as the main flow over the back.  I've ordered one of those fans, Dave.  Due to arrive Monday.

Duet say the cooling should be from the rear of the board but none of the components on my board get more than slightly warm to the touch.

New power wiring diagram.

44 minutes ago, tekkydave said:

Duet say the cooling should be from the rear of the board but none of the components on my board get more than slightly warm to the touch.

You probably don't need a cooling fan at all then.  Just mount it vertically (as you have done) with plenty of room for air flow across the back.  This is what the WiKi says if you're not using more than 2A for any of the motor drivers.  My Z motor driver gets hot but not too hot to touch.  Just want to be sure the board doesn't fail after hours of use.

24v 15A PSU with cover containing bed heater SSR and buck converter for +12v plus mains ON/OFF switch.

Water reservoir.  I shall also print a lid for it to keep dust out.

Lid.

Printing the water reservoir.

Print not satisfactory due to problems with the support material for the pipes.  I can eliminate the top one by feeding the water tube through the lid.  I could eliminate both by feeding both tubes through the lid and using the syphon effect as I have been doing with a squash bottle but I was hoping to have the outlet at the bottom at least.  I'm using colourless PETG which gives an almost transparent wall but this stuff sticks so well to itself that the support wont break off.

OK - printing a new reservoir without inlet and outlet pipes and with a single perimeter wall for maximum transparency.  Pity it isn't possible to switch to Spiral Vase mode after printing the base with the fixing lugs.  At least it doesn't seem to be in Slic3r.  It would be nice to do away with the join where it jumps to the next layer.

I've had an idea which I might try tomorrow.

Well, the reservoir printed perfectly except for one thing - it leaks at the seam, where the layer step is

My idea worked   Hopefully a watertight container!  Used Spiral Vase mode but set the number of bottom layers to 6 corresponding to the thickness of the fixing lugs (3mm).  First layer 0.7mm subsequent layers 0.5mm so 6 layers = 3.2mm.

The design works but with a 0.5mm wall thickness my Titan printer is not quite accurate enough to produce a perfect wall and there was a leak a few centimetres up.  This would probably be cured with a greater extrusion width.

Here's a photo.  Not the full height as it ran out of filament.  It can be seen that most of the wall is pretty good but the lack of rigidity in the bed carriage can be seen in places and hence the need for the Titan re-build. See my Titan printer blog.

The cooling water pump has failed

Decided to print the reservoir in ABS and make it it three parts - the body and two pipes and solvent weld the parts together.  With the reservoir and the water block both having low and high connections I think there's a fair chance of the water circulating by gravity as in gravity fed hot water systems (the water block representing the boiler and the reservoir the hot water tank).  It will also have a lid to keep dust out and prevent anything else dropping in.

It leaks!!  And I don't mean out of the pipes - it leaks through the body of the tank!