View Full Version : BUILD LOG: 8x4 Router build thread

23-05-2011, 03:05 PM
Hi All,

Here are the various design reviews that this machine has gone through since I started designing it in February this year.

As you can see many ideas tried out, some kept, others discarded, often due to cheaper/more rigid solutions, original ideas stemmed from the build it yourself on-line one and Joe's JGRO, then morphed through Joe's 4x4 and the R&P 4x4 through to it's current state.

The budget including all electronics is around 1500.00, I am keeping a spreadsheet that tracks the current component costs for all parts down to screws, etc and includes postage, currently well within budget (original budget started at about 800.00 but did not include any costs for bolts, screws, etc, these soon mount up.

Early designs


4x2 linear Z

4x2 profile double X

4x2 profile sides

4x2 profile sides linear Z

8x4 version 1

4x2 profile sides linear Z version 2

8x4 version 2

8x4 version 2 -Y axis

8x4 version 2 - X and Z axes

8x4 version 2 - Z and B axes

8x4 version 2 - Y and Z axes

C and B alternate idea

Comments welcome - design changes ahead of parts purchase only costs the development/drawing time.


blackburn mark
24-05-2011, 01:32 PM
its a good idea to do lots of drawings and mull over them at leasure :smile:

I predict from your drawings that you will have lots of fexibilty troubles resuling in a really disapointing experience (unless your cutting polystyrene)

you NEED as much ridgidity as you can pratically get without resorting to cast iron or stone:smile: lol
tacticaly you should be thinking about minimising flex EG your Y axis disecting the width of your machine and not the length and your Z axis travel seems abitious.
... imagine your Z axis at the bottom of its travel, if you put some pressure on it in the XY plane i imagine it would flex millimeters rather than tenths of millimeters:sad:

hope im not putting you off to much but get the job right and you will be doing cartwheels down the street when its done.... get it wrong and it will break your heart having to re-work the bugger... good luck:smile:

24-05-2011, 02:43 PM
Thank you for your comments and insights.

I agree that a long Y rather than a long X would be more rigid, it is also more expensive (extra rack). I will have another play with it and try swapping it round again and re-cost. Should be cheaper than last time, now that I have dropped profile sides for the Y in favour of Unistrut.

Yes, the Z axis is ambitious, and I have similar concerns, but needs to be if I have any hope of doing anything in the rotational axes, I am also hoping that the X axis is stiff enough to prevent too much flex with the distance between the bearings, it does however have over 400mm of Z travel.

The design is also adjustable in that the bed can be moved up and down manually - probably taking a couple of hours so I can start with it high up and minimal Z travel and then drop it down as I become more ambitious and work out any flex issues.

I also realise that the weakest stress/flex points for my B and C axes will be the joint between the axle and the turntable.

An A/B axis attached to the bed of the machine would obviously be more rigid than attempting to tack it on the end of the Z axis (I have designs for this as well, just not drawn up, although my more rectangular trunion design along with a tail stock would work equally well here).

A spindle would also be less moving mass and torque than a router but I have the router and it was a fraction of the cost of a spindle and VFD.

I am also designing a smaller one for PCB and other small more precision stuff, about A4 size bed, I will post up a thread for this one as well, I am looking to fabricate a spindle for this one from a model aeroplane motor and an ER collet chuck, not looking to do heavy cuts so hoping this will be up to the job (not expensive and easy to upgrade/modify if it is not), Proxxon spindle would be the next choice if this idea fails.


24-05-2011, 03:59 PM
I have to agree with mark here...

I predict from your drawings that you will have lots of fexibilty troubles resuling in a really disapointing experience (unless your cutting polystyrene)

Yes! It's interesting though - you've gone for very expensive aluminium extrusion which is good for rigidity and yet very cheap, and in my opinion a bit dodgy, bearings. I think you'd be better off investing that money in proper linear bearings from eBay (look up linearmotionbearings seller and send him a message with the sizes you want - may be pleasantly surprised). What your doing on the X axis relies on the steel strip having a constant width and doesn't account for wear. It would be better with the bearings sprung. There will be a lot of force on each bearing, and due to the small contact area in your design a lot of pressure.

your Z axis travel seems abitious.
... imagine your Z axis at the bottom of its travel, if you put some pressure on it in the XY plane i imagine it would flex millimeters rather than tenths of millimeters:sad:

Again, yes very ambitious... I went for 400mm Z travel on my machine using 20mm thick aluminium plate for the Z axis. It still flexes, not too much but enough to stop it machining metals. A lot of my flex probably comes from the gantry so I'm going to lower that.

It's not really worth thinking about the B/C axis untill the other bits flexing are sorted as it's only going to add more flex/overhang to the design. That's why I've left my B/C axis design for now, still planning it but I'll only make it once I've done the steel frame for my machine. The following drawing shows the general idea/shape. The 'C' shaped bit that holds the spindle will be made with something thinner with a gap between and possibly welded to make it strong enough. I've calculated for the final belts I'll have to use 25mm wide HTD to get it strong enough. Two belt reduction on both axis with 3nm motors. Bearings will all be tapered roller bearings ... etc:


The centre distance of your belts looks too small for the ratio. You should aim for 6 teeth engaged on the smaller pulley.

26-05-2011, 08:00 AM
I replied to the first response shortly after it came in, but my post is still not showing after 2 days.

I appreciate that new members posts are vetted for the first few occasions, but having to wait 2 days for a post to appear is a bit much.

I agree that a shorter X in preference of a longer Y would be more rigid, and this is how it was originally designed. The idea was dropped because of the increased cost. However I have since dropped the profile sections on the Y axis so it should be cheaper again - I am mid re-design on this again to see how it looks and costs out.

I also agree about the potential flexibility of the Z axis, which also has around 400mm of travel, especially with the B/C ideas that could have all sorts of slop in them - a more stable B/C would be a lathe type design on the bed that could either use one of my trunion designs or simply be a chuck geared down to a stepper with a tailstock.

I am also going back to some homebrew versions of the 'CNC router parts' carriages and some thicker steel for them to run on 6mm instead of 3mm, still redesigning the X/Z carriage interface.

Image shows the carriages on the alternate Y axis design.