AVOR – a steel-framed medium-size router

[Neale]

The title? Medium-size – bigger than a bench-top machine but not as big as a “full-sheet” 8x4. Steel-framed – basic structure is steel box section throughout, including the gantry, no aluminium extrusions. AVOR? I don’t usually name my machines but it seems to be fashionable so here is “A Very Ordinary Router.”

I’ve put this in the Build Log section although this isn’t a build log – the machine is built already apart from some small details and is in fairly regular use. There are one or two points about this machine that are a little unusual, and I would also like to encourage others to look at steel as a build material.

There are some great build logs going on at the moment with some very nice machining, etc, going into them. Maybe this machine will demonstrate that a machine doesn’t have to look pretty to work! It’s great if it does, and there are some people who are skilled enough to build machines that are both good-looking and work well. But a machine does not have to look pretty to work well, so don’t let that aspect put you off building.

I must thank all those contributors to this forum whose ideas I have shamelessly copied or adapted, too many to name. However, any errors or problems that I describe are mine alone and not the fault of anyone else. I can make my own mistakes without needing any assistance

The Machine

I wanted to replace my first CNC router which was based heavily on the JGRO design (available via a Google search – plenty of examples around) and built in MDF. It had many problems which I wanted to avoid with the Mk2 although it did a useful job for me for around 4 years. Basic goals were a decent woodworking machine although the ability to machine aluminium would be an advantage. Not necessary, though, as I have a vertical mill anyway for metalwork (CNC conversion might happen one day ). Capacity is always a compromise - ability to work with quarter-sheets of material was a starting point, but the machine needed to fit in my garage/workshop so could not be very much bigger. I had already decided to use ballscrews and Hiwin rails. I started the project about 3 years ago (I’m not a fast worker) and at that time buying these from China was not as obvious a choice as it is now so I bought from CNC4YOU. Not as cheap as China but I have always had good service from them. They stocked standard sizes, so having worked out my minimum machine size, I bought the next stock size up for ballscrews and redrew the machine to suit. As a result, the overall dimensions are length 1800mm, width 1000mm, giving a cutting area of about 1550x750. The X rails overhang one end of the bed which allows me to machine work clamped to the end face of the frame.

I wanted a floor-mounted machine which could be moved around with the idea that longer pieces of work could overhang the bed if necessary by moving the machine out from the wall. For various reasons (cost, ease of working, and modification) I chose steel rather than aluminium. This is a CAD drawing of the machine as it shows the general structure rather better than a photograph (which is difficult in my workshop anyway). There is obviously a lot of detail missing from this picture; in particular, there is a lot more bracing of the structure in the finished machine.



Although I did a lot of the design in Fusion 360, this was more to work out overall dimensions and some of the details which is why the drawing has some odd little items (motor mounts, for example) but no ballscrews shown.



The photograph was taken during construction but you can see the bracing in place.

The structure is steel box. Most is 50x50x3 although you can see that there are some 100x50x3 sections supporting the side of the bed and the X rails to give extra stiffness for relatively little extra weight. Reading this forum, there seems to be a general view that anything less than 4mm wall thickness is needed, and 5mm might be better. I have not yet found 3mm to be a problem with one exception – it is not thick enough to take the M5 capscrews holding down the profile rails, or for similar attachment screws elsewhere.

An unusual feature (not unique so I’m making no claims here but I don’t remember seeing one described at the time I started my design) is the welded steel gantry. People worry about gantry mass and how it might affect performance. When I did the sums, I reckoned that the rotational inertia of the ballscrews was at least as important, if not more so, than the inertia of the gantry when it came to acceleration/deceleration (which is what often matters most – inertia isn’t very relevant to speed) even with the steel gantry. With suitable bracing it could be pretty strong even though I was using my 3mm wall thickness box. This design also meant that I could run the ballscrews through the uprights at the ends of the gantry which solved a lot of design issues. More about all this later.

Again, more details to come, but the drive is by NEMA 23 3Nm steppers, two for X, one for each of Y and Z. X ballscrews are 2005, Y and Z are 1605. Motors drive the ballscrews via belts and 1-1 pulleys. Rails are 20mm Hiwin.

The control box uses EM806 drivers running from 68V linear PSU for X and Y, and a recycled M752 for Z – less critical than X and Y, and not needing the stall detection which (in my view) is absolutely necessary for a dual-motor axis such as mine. There is a CSMIO-IP/M motion controller, plus Pilz safety relay and the usual odds and ends. Limit switches are inductive, NPN NC, running off 24V.

Spindle is the usual Chinese water-cooled 2.2KW/3HP choice driven by an HY VFD. These are the same units first used on my Mk1 router and are now about 5 years old, and have never given a problem. Cooling, when I bother to switch it on, is via a small caravan water pump recirculating water from a 10l bucket. In fact, to extend the life of the pump, I only run it on 12V when initially filling the system or to purge any air bubbles and most of the time it keeps a trickle going driven from 5V. However, for short cuts, I don’t bother to turn on the coolant although I do check the spindle temperature from time to time. Occasionally, the water seems to be a few degrees above ambient. I probably don’t drive the machine hard enough…