SX2P Gas Strut Modification Issue?

.

An excellent reaction, Pete

it demonstrates the merit of of a forum.

I don’t recall seeing any authoritative documentation about the ‘reverse running problem’ of ballscrews but I feel sure it must be out there. ... If anyone happens to find anything, it would probably merit a new forum thread for convenient reference.

MichaelG.

It seems to me that the issue is that the head is acting like one of those sprung woodpecker on a pole toys. I'm not sure of the simplest of fixes for this but the toys do rely on friction and a force rotating the sleeve relative to the axis of the pole to work. Once there is some upward force from the tool on the workpiece to balance this downwards force the friction at the column reduces and the head drops.

The use of gas struts and springs cause issues because they do not give a linear response of force over distance compressed or wound up. There are two ways to have a more linear response, one is to use a pneumatic piston with a constant supply of air at a fixed pressure from a regulator, the other is with a counterweight and pulley system.

I think I would try rigging up a temporary weight and pulley system with the upward force taken around the middle of the head to see what effect that had without the springs or gas struts being used. It is easy to adjust the counterweight and gib strip to experiment.

Martin C

The woodpecker thing is a narrow collar though - it's like the mastic gun principle. The mill head has a much longer area of contact.

I put the sprung load as close to the column as I could, to minimise the lever effect as much as practically possible, but it still sticks.

Sod it. I'm going to saw the rack and offset each half. That appears to be the simplest way of giving positive motion from the feed wheels, with continuous pinion contact both up and down. If it doesn't work I'll get another rack and live with it until something better turns up.

That looks like an excellent solution to the problem, very nicely done also.

Just to clarify, I have no experience of this happening, just something I read on here from someone who had tried it, be interesting to her more from anyone who might have experienced this.

Pete,

I don’t think this helps enough to deserve a new forum thread ... but it’s a start: **LINK**

http://www.machmo.com/tools/ball-screw-back-driving-holding-torque-calculator/

Big problem seems to be that you need to input % efficiency [which seems like the big unknown]

MichaelG.

.

Does anyone have some ‘typical’ numbers we could try in this calculator ?

.

Edit: __ Just found this : https://www.youtube.com/watch?v=pCD2HXjiWRQ

Edited By Michael Gilligan on 04/03/2021 17:30:45

Stop looking for complicated answers to a simple problem 5 months on and the problem not solved.

Time to learn how to learn how to back off the cutter lead angle to 90 deg for machining caste iron, bronzes, and brasses.

Only my biased opinion based on a lifetime of engineering after all it's your time and money.

Eric

Edited By oldvelo on 04/03/2021 17:32:19

Edited By oldvelo on 04/03/2021 17:32:33

I had similar problems to the OP with my milling machine. Specifically, the head had a habit of making its way down into or up out of the work depending on how tough the going was. Of course, it was OK if I locked the z axis but this is not an option for CNC operation.

My chosen solution was a heavy counterweight (album photo) of sufficient size to ensure that the leadscrew was always loaded in the upwards direction. The application point of the counterweight load was estimated to minimise the couple, and therefore friction, experienced by the z axis slide. My analysis of the mechanics suggests that this application point is not in line with the leadscrew. To do this properley one would need to know the weight of the counterweight, the weight of the head and the position of its centre of gravity. Rather a tall order in the case of the latter two measurements but perhaps doable on a small machine.

To be clear, the aim would be :

weight of head x distance CG of head to leadscrew CL = weight of counterweight x dist. application point to leadscrew CL

Offered as food for thought.

Chris

In MEW 251 Phil Dawes showed his approach to using a counterweight to balance the head of a Sieg SX2P. Might be worth you having a look at that article to see how he did it.

Neil

The way I see it, balancing the head isn’t the fundamental issue, it’s balancing *and* eliminating sticking *and* eliminating backlash. In fact balancing is probably the least important of the three. Addressing one Z-axis issue but not the others seems pointless, since the one(s) remaining will eventually get you.

I understand about the couple on the head and position on the counterbalance link to both reduce sticking and balance the head, but as soon as you put cutting loads on the head, the theoretical link position will change slightly (if I understand correctly).

I think the most relevant improvement in this case would be to eliminate backlash - or reduce it to a sensible level. You can see from the video I posted that it’s quite severe - about 5-10mm. Combined with the stick-slip it’s asking for trouble.

If the head’s not balanced, it just gives a bit more load on the hand wheels (the head won’t drop anyway with fine feed engaged). The gas strut is fine - in fact almost perfect notwithstanding the offset load it gives.

If there’s any sticking, it’ll be overcome by the action of positively moving the head with the hand levers. The worst that can happen I assume is that there would be a momentary increase in load. I doubt this would be felt to any significant degree by the time the worm drive has increased the handwheel torque on the pinion.

So I’m left with swapping the rack for a screw, making a spit/preloaded pinion, or experimenting with splitting the rack. The latter seems the cheapest, simplest and least risk option. Having said that, I can find no examples online, which is concerning, but there we go. The nearest example I’ve been given is a split and offset rack in a mountain railway system, but that’s not to reduce backlash, it’s to ensure consistent pinion engagement. The usual method for precision rack drives us to split the pinion, but this is a far more involved modification. I can understand why this is a preferred option, since it doesn’t mean messing about with adjusting potentially long lengths of rack.

Just save me wading back through every post I keep seeing the mention of "sticking" are you saying that the head actually sticks to the column..?

It’s a ‘jerky’ column movement: stick-slip. It’s intermittent. If I rapidly turn the hand levers down, there’s often a distinct thud-thud-thud as the head moves down. Do it slower and occasionally it’ll stick, and wont unstick until all the backlash has been taken up and the pinion forces it free.I’ve tried adjusting the gib strip and lubricants, but I think short of scraping the slides (which I’ve never done before), and fettling the gib strip the issue will remain.

It’s strange because I’ve checked that the head isn’t moving around, with a DTI, and I can push the head up and down by hand quite freely with the rack removed. This suggests the gas strut is pretty much spot-on in terms of balance.

probably something to do with how the pinion is applying a lever load to the head, or maybe the gib strip is not contacting all along it’s length, or the screw detents aren’t right, or a billion other things.

All this balancing out the weight of the head stops it sitting firmly down on the screw, as soon as you make the head "weightless" then you get backlash and the sticktion is not overcome by the weight of the head so you have to push/pull with the leadscrew to overcome it. strut on the side also twists the head on the column so you may balance out the weight but the head itself is not balanced centrally.

If the head was "heavy" then that would have a similar effect to you pulling it down with your hand in the video.

I'll stick with my unweighted X3 and a strong hand to wind it up

Edited By JasonB on 05/03/2021 08:38:38

Backlash-free rack and pinion

An improved rack and pinion has been developed that eliminates backlash without using split gears. It is amenable to easy retrofitting of existing systems.

Might be worth having a look

Understood Jason, I also assumed that, but as I said previously, even without the gas strut fitted, the sticking is still there. That's why I tried an adjustable strut and de-rated it to force the pinion down into contact under the weight of the head - still the same stick-slip. I then removed the strut altogether, assuming it would cure it, but it was still there.

I then went to spring loading the head down, as near to the column as I could get it, and it still sticks intermittently unless you put about 35kg on the dial, and even then you only get a limited amount of downward travel before the spring load runs out.

Thanks Pete.

Looks basically like another way of increasing engagement with a spring.

I tried incrementally shimming the rack out already, but even when it got the the stage of unacceptable binding and rough operation, there was still backlash. So there didn't seem much point in continuing down that path.

Edited By Dr_GMJN on 05/03/2021 08:54:02

.

I would posit that everything else you are trying is futile [perhaps even counter-productive] unless and until you have established smooth sliding of the head on the column.

MichaelG.

So, when it sticks when moved down with the fine feed, it immediately un-sticks once the pinion contacts a rack tooth (ie once the lash is taken up). So I'm assuming that if there's no lash, or a very small amount, the movement will be smoother. So instead of a 5mm drop, I'll have perhaps a 0.1mm drop. I can live with that.

exactly!

If your fine adjustment does nothing when adding cut, then the head drops 5 mm then the rack /pinion interface is strange,. Perhaps the pinion is moving in its bore The sx2 is a cheap light weight machine. Built to keep the costing down. When I converted mine to cnc I found that the z head dovetail casting was warped, probably from machining it when the casting was a bit green. It took a lot of work to get it so that with the column off the machine the head could slide up and down without any twisting. The test was done with a 3/4” silver steel bar in an R8 collet and the head was pushed along the column by hand. I can see that this could be a source of your problems as well.

I think the twisted casting was causing it to rock and this would be a course of the sticking.  I found it by putting the head on the surface plate on the working face of the dovetail.  From memory it I think needed about 15 thou packing under one corner to stop the rocking.

the idea that a ball screw nut will allow the head to fall is bunkum. The nut has to rotate for it to move! Fitting any z feed screw would solve the problem as you will have total feed control and sticking issues will be resolved unless there is serious clearance issues within the z slide arrangements as I had . Fitting a motor to any feed screw would make positioning impossible without the associated electronics and/or computer control. A hard wheel would do the trick nicely, so you end up with an arrangement as in one of the photos above and like the down feed arrangement on the Emco FB2 mill

I could never understand the original spring arrangement as the lever arm resulted in a change of upward force as the head is moved up and down all this and the z feed system was scrapped In my conversion which works very very well without any form of upward thrust. When jogging up and down with Mach 3 I need 1-4 x 0.01 increments in order to obtain a reversal of the head movement . This has not manifested itself as any problem while in use.

fit a full feed screw and dump all the z feed arrangements which I always considered to have too much lag and interface clearances to be effective. Your problems seem to confirm this

Edited By Zan on 05/03/2021 10:06:29