SX2PG Head Tilt

[Mark Salzedo 1Participant @marksalzedo1]

Hello,

My SX2PG milling machine has got a right head tilt of about 4 thou. I’ve lived with it up till now. But I need to do something about it in order to do more precision work. See pictures 1 and 2 which show that as the head is moved down the error gets worse over a height of about 3-4 inches.

Google AI suggests that I remove the head, separate the spindle housing from the carriage (held by 4 – 8mm cap head screws) – see picture 3 and then shim the right hand side of the mating surface.

Before I take the head off again and separate the parts, is there anything I have overlooked.

This is the link to the images

Many thanks,

Mark

 

[JasonBModerator @jasonb]

Decide if it is the head tilted on the column or more likely the column is tilted relative to the table. If that is the case then loosen the bolts holding the column to the bed casting and shim with something like feeler gauge material. Both my Siegs are shimmed like that.

[Mark Salzedo 1Participant @marksalzedo1]

I’ve swept the table relative to the column, shimmed it, and run out is only about 1 thou. But I think the head is tilted relative to the column. Presumably this scenario is unusual with these machines? And it’s only 9 months old.

 

[JasonBModerator @jasonb]

I don’t see how the shimming you suggest would alter the readings you are getting? Loosening the 4 screws and seeing if you can slightly rotate the spindle housing relative to the part with the dovetail is more likely to see things move in the right direction. Then you would need to reshim the column to suit the adjusted head buy swinging the dti over the table.

[Mark Salzedo 1Participant @marksalzedo1]

Yes, that sounds more realistic. The problem is how much you slightly rotate it. To remove an 8 thou tilt is a very fine movement. But I guess it’s going to be trial and error.

[JasonBModerator @jasonb]

If you laid it on it’s side on the mill table, clock the rod so it is horizontal and then clock the dovetail until that too is horizontal, you should be able to get quite close without having to keep putting it back together and testing.

[rjenkinsgbParticipant @rjenkinsgb]

Can you borrow an engineers box level from somewhere? With a good level you can set the machine itself level, then set the spindle. A good one should have a scale with divisions of ten microns per metre.

[Mark Salzedo 1Participant @marksalzedo1]

Thank you Jason and rjenkinsqb (sorry, I don’t know your first name) for your suggestions. I do have an engineers level but wherever I shimmed – directly under the milling machine or under the column – it didn’t improve the head tilt to the right (by about 4 thou).

I think the head must be slightly off centre and by rotating the carriage section in relation to the spindle housing will correct the error. The problem is by how much do you turn it. Compensating for 4 thou is an infinitesimally small amount. I have placed some scribe lines around the spindle housing which sits just above the head as a baseline but this is not a very accurate way to do things.

Jason’s idea is theoretically sound. The head can be laid on its side on the milling table. But I will need to think how to do this due to its size and shape, and then to fit a test rod in the spindle and sweep the dovetail and rod.

The four cap screws holding the carriage to the spindle housing are quite easy to remove and the carriage can be rotated slightly to offset the inaccuracies.

There are some more images here.

[Mark Salzedo 1Participant @marksalzedo1]

As a by the way, Google AI suggested I drill four access holes at the back of the column through to where the four cap head screws are located, so that in future if there is slight misalignment all that needs to be done is to insert a hex wrench through each hole and realign without having to  disassemble the whole machine.

[ZanParticipant @zan]

I also had problems with alignment of my sx2p.  this was detailed in my series in ME starting 17 Aug 2018 for cnc conversion. There were two main problems

1 the female dovetail column to head mounting block was twisted, I put this down to a green casting being machinedtoo soon.  It rocked  badly on the surface plate. A lot of scraping on the slide side taught it some manners and the front was machined back parallel.

2 still no joy, I found an adjustment on the column base bolts caused totally erratic changes in spindle alignment.   On removal of the column, the seating in the base was terrible.  Not flat, had been attacked with a very rough file and the oil marks showed only limited contact. One mounting base had a relief filed in the centre leaving high pads at the ends, while the other had oil marks which indicated poor contact with the column. Tighten a bolt with no support under it would cause the column to tilt, twist, lean etc. see pic 2 p 303. Solved by using shims to set Y base dovetails totally level and machined flat on the Bridgeport.

it was a lot of work, only done because I had done a lot of work in the machine obtained all the ball screws, machined all the x mountings drilled and tapped numerous holes etc so I could not send the machine back. (The articles do not show the chronological  order in which the conversion was undertaken)  In the end  I ended up with only 0.02mm error on a 300mm long 15 mm sil steel bar in the spindle and to say the least I  was very pleased with this!

The cnc mill has had a lot of work, and is extremely accurate. I am very pleased with it.

[Mark Salzedo 1Participant @marksalzedo1]

Zan – thank you for your comments. I have sent you a private message.

[old martParticipant @oldmart]

One of the mills at the museum is a round column Taiwan made with no adjustments at all, so I didn’t bother to try thecking the tram until recently. I found that the spindle was slightly tilted in two axes and was thinking of shimming the base. We have some aluminium plates ground flat and I cut one and drilled the mounting holes and then sat it on the bed and skimmed the minimum ammount of the top with a flycutter that I had made. The plate was inverted before fitting it in place with slightly longer bolts. Most of the error is reduced, the maximum is about 0.0005″ over about 10″ travel which is good enough for government work.

You should get the X axis trued up before looking at any Y faults are addressed.

[old martParticipant @oldmart]

My results before and after fitting the plate are here:

[Author]

Posts