Bridgeport Manual to CNC conversion...

[John S]

Quill drive idea mk1,

gives a 16:1 reduction, takes up little space, mounts to existing threaded holes on left side of head.

Attachment 19498

Should allow both belts to be tightened by having only one moving shaft.

What about backlash in the rack and pinion ? ?

[Davek0974]

What about backlash in the rack and pinion ? ?

That is my main issue, its fairly tight but will have some BL in it. I was possibly thinking of some method of applying constant up force to the quill from where the existing drive is now, not sure how or what though.

Its a pain really - so close but not there yet, i know for a fact that the existing drive will fail and probably quite soon so doing nothing is not an option.

I have been looking at powering the knee and locking the quill but i still have my doubts about the knee ways and short life etc, plus the cost - ball-screw, bigger servo and drive, hardware, probably going to run about £600 just to do the knee, more if it needs a brake to stop it falling when turned off.

Then there is the speed issue - from what i have read, the knee is slow, peck drilling is going to suck as well as a few other jobs too - V-engraving would take ages due to the thousands of up/down moves used.

Just a bit baffled as to what to do really

[Davek0974]

This seems a pretty neat way of backlash control - its a brake cable from a motorbike, one end is mounted to quill and the other has an air cylinder on it with variable pressure. The guy that fitted it also fitted a very tidy gear reduction system in the aluminium casing on the left, 15:1 ratio, double reduction. He reports zero backlash and full quill travel :)



I think on cost alone, it has to be worth a try to run via the quill rack, powering the knee at any speed is going to take some time to save up for.

[Davek0974]

Measuring on the adaptor from the Z screw to the quill, near the quill, I have well over 0.5mm of flex - thats a bad figure as it equates to backlash at the tool, so doing nothing here is not an option.

So, looking at the options -

Power the quill via the left side of the pinion shaft, this has slop in it so will need some clever way of preloading the quill upwards - this could lead to rapid rack/pinion wear as it will need enough preload to counter cutting forces from pulling the tool downwards.

Power the knee - this is an option but will cost a lot as it needs a ball-screw, servo, drive plus full tear-down to fit it all.

Alter what i have got - looking at the picture below, is there any harm in removing the web of metal circled?? It would mean i can get a bigger connecting arm in there with much more contact below the bolt to resist the down-forces, without losing any Z travel.

[Davek0974]

After some thought I reckon the easiest option based on what i have going already is to modify the quill housing and build a better connector arm - this should give more rigidity and still maintain the full Z travel.


In the attached picture, on the left is the existing setup, on the right is my next version - as you can see, in the existing one a lot of twisting moment is placed on the bolt when the arm is pushed down (heavy lines).

In the new version, the twist is converted to an axial pull on the bolt and the force transferred to the quill body via the longer contact face below.

[Davek0974]

That was a long day!

Stripped the head, cut away the small web of cast at the bottom of the quill slot, altered the screw connector by welding on a 13mm thick bar underneath, the end of the bar was milled to the curvature of the quill before stripping, fitted up, refitted the Z drive and shimmed the connector, power head back on, motor back on, encoder back on, tested.

The connector now has an extra 25mm of contact below the quill bolt which is a massive improvement over the first attempt, this makes a big difference to the twisting effect on the quill bolt and is now far more rigid in the downward direction which is what matters.

Total backlash is now down from 0.6mm+ to 0.1mm and I also discovered that most of that backlash is caused by the BK screw block not clamping the screw correctly - this was always there but never spotted it before. I will strip that block and see whats wrong with it - either crap bearings or sleeves too short and the nut not reaching properly.

I don't think this one is getting any stiffer (ooh err!)

[Clive S]

It's always nice when a plan comes together

[Davek0974]

It's always nice when a plan comes together

Thanks, it wouldn't have worked without removing that little web of casting though:)

[Davek0974]

What can make the tool not follow the path correctly?

Did some work today and as seen here its clearly pulling away from the corners far too early - corner rounding?



Not quite so clear here...



Cutter was 2mm at 23000rpm, 1000mm/min, X&Y speed set at 3500mm/min and 350mm/s/s in motor tuning.

[JAZZCNC]

Your getting because of constant velocity and tight radius most likely caused by low accelleration. Try setting accelration higher and see if better.