The main frame.

Decided to use a different corner fastening instead of the plates.  Once built the sides will be boxed in with clear acrylic sheets.  This will stiffen everything up (if it isn't stiff enough anyway) and provide a fume cabinet for printing ABS.  The frame will be constructed from V-Slot rails some of which will take the wheels of the carriages.  The Y carriages run on the lower top horizontal rails and the Z carriages on the inner vertical rails.   V-Slot extrusion is actually cheaper than other sorts.

This top  view diagram shows the principle - this actually refers to my biggest 3D printer project but the same principles apply.  This Mini printer is simply a smaller version of the bigger one.  The orange square with the red border is the print bed, the rail sticking out from under each side guides the bed and holds it in the right horizontal position.  The ends of that rail are attached to two carriages that run on the side uprights which form the Z rails.  The print bed moves up and down to form the Z axis.

The X axis is comprised of the horizontal rail shown above the bed, which the X carriage runs on.  At the ends of the X rail are the two Y carriages which run on the two front to back rails.  XY motion is controlled by cords running over pulleys and attached to drums on a pair of stepper motors.  This drive arrangement for the X and Y axes is called Core-XY.

The X carriage partly equipped.  Needs hotend heater and various cables.

A couple more photos.  X carriage and rail just on the table and top view of both X and Y carriages.

Preparing for ordering the frame parts - I think I already have most of the working parts..  Been looking into which sizes of clear acrylic sheet are available to complete the box as the frame parts can can be ordered cut to length.  Ideally the depth would be 350mm but so far I've only found 300mm or 400mm.  The width works out fine at 400mm.

Looking at the possibility of 300mm depth, this would make the Y rails 260mm long (300mm - 2x20mm for the uprights).  Y range required is 200mm less a margin round the edges of the pr int bed that doesn't get heated.  This gives about 190 as the minimum Y axis range and the maximum width of the Y carriages as 260 - 190 = 70mm.  That might be alright if the Y carriage wheels could go right up to the ends of the Y rails.  The back is alright with the pulleys but the front requires space for the drum on the motor shaft making the minimum overall width 80mm without adding extra pulleys.  Can't be done!

Found some 350mm x 500 mm acrylic sheets, cost more than the 400mm x 500mm but if that's the size I want sobeit.  Anyway, I'll design the whole printer before committing.

Going for a footprint of 350mm x 400mm and the overall height will be 500mm. 

List of parts required :-

1. V-Slot Linear Rail - 20x20x500mm -- 4 off
2. V-Slot Linear Rail - 20x20mm - Cut To Size 310mm --- 4 off
3. V-Slot Linear Rail - 20x20mm - Cut To Size 360mm -- 4 off
4. V-Slot Linear Rail - 20x20mm - Cut To Size 350mm with tapped holes  -- 1 off
5. Inside Hidden Corner  -- 16 off

There might be more but I can order anything else later.

The above parts have arrived so I can start assembling the frame.

Mostly assembled the frame but didn't order the uprights for the X rails and accompanying corner brackets.  I think I am missing a couple of grub screws from the brackets I have but I guess it's possible that they escaped and landed on the floor somewhere.

The other parts have arrived and I have finished assembling the frame.  Another small advance   ATM I'm just doing little bits to several projects.

Found a massive design boo-boo   The wheels on the bottom of the Y carriages run in the underside groove of the Y rails and as I've designed the frame would run into the Z rails!!  The Z rails will need fitting to the outside of the Y rails and bottom frame rails as in the design of my Giant printer.  I may need to buy some more brackets.  The printing height will be reduced by 40mm but I think this will still give a build height of something like 220mm which is alright.

Had another idea - keep the current framework but add new Y rails above what were supposed to be the Y rails leaving space between for the wheels.

Been continuing with construction of the printer.  Printed the Y carriages and assembled both them and the X carriage with extruder and Z probe.  It looks as if the Y carriages would sit fine on the rails without bottom wheels as long as the print speed was kept fairly slow but with very fast changes in direction and rapid acceleration/deceleration the Y wheels could lift due to the larger mass above the Z rail.

I don't really know if this is a problem and I think I'll try the printer without bottom Y carriage wheels.  It would be no problem to add extra rails later if required.  I think that with the Y drive being by the pulleys above the rail the masses of the motor and the X rail may cancel out

I've had another idea for the Z wheels that keep the print bed located - I could use the corner upright rails.

That is going to work   The two V-Slot rails that were originally designed to act as the Z rails now become thee sides of the bed.  I have tapped the ends for M5 screws and mounted V-Slot wheels on M5 bolts screwed into the ends of the rails.  One end fixed and the other with adjustment to fit the wheels to the upright rails.  The base of the bed is a piece of 6mm plywood bolted to the rails with 4 screws.  The holes for the screws for one of the rails have been elongated to allow for adjustment to make the wheels exactly fit into the grooves in the rails.

Popped out to help a friend with a problem but back now and taken a couple of photos showing the print bed frame and Z carriage.   I've placed the glass plate on the base to show the proportions.  On top of the wooden base will be a half inch thick layer of polyurethane foam insulation then the bed heater pad and finally the glass plate on top.  A 3D printed frame will hold it all in place and the heater pad against the glass plate. 

I could replace the 12v 220W heater pad with a mains voltage 300W heater pad with adhesive backing to stick to the glass plate but already having a non-adhesive heater pad and foam to press the heater pad against the glass there's probably little point.

Having sorted out the Z carriage, the next thing is to look into the Z drive.  (The XY drive is already designed.)  The Z drive will consist of four cords one to each corner attached to the wheel spacers, and going up and over pulleys to the back - that's pretty much as far as I've got.  The Z stepper motor will have a small pinion which will engage with a large spur gear that will connect to drums that wind up the cords to lift the Z carriage and print bed.  I have yet to decide how I run the cords to the drums.  I'm thinking probably more pulleys.

I have already tested the viability of cord pull-up and gravity pull-down for the Z carriage in my "GinaRep Titan" printer.  It works well and has been in almost constant use for some time now, producing good prints.

I have alternative idea to lots of pulleys - an M8 threaded rod along near the back (say 25mm from the back) could lift the rear side of the bed and the front would go over pulleys also onto the rod.  The rod would need to turn 16 revolutions to lift the bed the full distance and a 50mm pitch diameter spur gear could drive it with the same resolution as with my Titan printer.  I need to check all the movements for interference but the initial look seems promising.  I'll see if I can produce a sketch or set up a test rig.

With the above idea, care has to be taken that the offset due to winding along the shaft (or drum for that matter) causes a minimal error in the Z axis and that the bed remains level as it descends during printing.  This means that the geometry must be arranged so that the cord is at right angles to the shaft when Z = 0 and deviates from this as the bed descends, minimising the error since the error angle is minimised and the difference between the adjacent and hypotenuse (Z axis error) is minimised.  (The Z axis is represented by the adjacent side of the triangle and the actual cord length by the hypotenuse.)  I'm afraid those not conversant with geometry will just have to take this as read (or ignore it and think "I guess Gina knows what she is talking about but I don't" )

This seems to be workable but the length of the wound up cord sitting in the thread is 16mm whereas on a plain rod it would be only 8mm as the cord is 0.5mm diameter, which would be a lot better.  I have ordered a 400mm length of 8mm diameter brass rod (to match the ball bearings I have on order, though I guess plain bearings would be adequate) and could either tap it at the ends to take nuts or just use tight fitting 3D printed sleeves.  I can drill it to attach the cord.  ATM I have a 500mm length of M8 stainless steel studding that came from the old 3D printer kit.

Progressing with the Z drive - have two ball bearings mounted and one of the two pulleys.  Also, designed the gears to drive the axle.

Here are some photos of the Z drive shaft bearings and a method of adjusting the cords so the the bed is approximately level.  Cord adjustment is by slackening the screw and sliding the screw and backplate along the groove and tightening up when adjusted.  Exact levelling is dealt with by the auto-bed-levelling routine in the Marlin firmware.

I might be able to advance the Z drive without waiting for the 8mm x 400mm brass rod to arrive.  I've found a 440mm 10mm stainless steel bar that came from the Velleman kit and no longer used.  I won't attempt to drill this for the cords but I could make collars to clamp onto the bar.  I could use this with plain Nylon bearings rather than ball bearings.  Polished SS in Nylon bearings should be perfectly adequate for this application - I have found this combination very good in my clock projects.

With a 33t spur gear and 8t pinion on the stepper motor the Z resolution would be the same as my Titan printer.  I have been using 0.2mm layer height on the Titan quite satisfactorily so I think this drive for the Mini printer should be fine.  I have upgraded the Titan hotend system to a genuine E3D Volcano right through from the extruder (also E3D).  I have found a significant improvement in using genuine E3D parts over Chinese copies.

Z drive gear.