Setting up my lathe (4 years late) :)

I bought my SPG2109 lathe 4 years ago and just bolted it up and plugged it in and away i went. Accuracy has never been an issue, mostly i turn parts down by trial and error.

The bearings started to rumble so I fitted a set of Timken taper bearings. It went back together and works.

Whilst it sat there in bits waiting to the bearings to be delivered my attentions turned to the accuracy. Ordered a ground steel bar accurate to " h8 tolerance +0 / -0.033 ".

After fitting the bearings I tried a cut and it was terrible, I put an 12" bar in and tried moving it, i could detect a slight rock and nipped the bearings up a touch more to eliminate that. Phew!! that solved the issue.

The runout was fairly bad so I stripped and cleaned the chuck and took a slight skim of the spindle plate as I thought it maybe contacting the back of the chuck. It appeared to cut uniformly, so high spots or runout.

I turned a 20mm piece of aluminium bar down very slightly (close to the chuck). My 0.01mm dial gauge wiggled ever so slightly. Width of the needle at most.

But extend the bar 12" out in the same position and its now + and - 0.1mm out. If im reading the gauge correctly 0.01mm divisions and giving a total runout of 0.2mm at 12". The highpoint is exactly where number 3 jaw clamps and when at the 12 o'clock position. (standard 3 jaw chuck supplied with the lathe).

Although I faced the spindle plate I did not machine the locating ring, maybe thats an option? It does fit that very very snugly.

The cheapo lever style dial gauge decided to break at that point. The silly dovetail part snapped off the holder. Looks like I need to make a decent one.

Sorry for the long waffle guys. Am i going in the right direction here? What do I need to do next?

Thanks

Dave from Brum

I would leave it at that, 0.2mm TIR 12" out from a 3-jaw is good, they are not the most accurate thing for holding work. Probably bend in the aluminium bar does not help

You're assuming the rod is dead straight. It might be possible to check that depending on what equipment you possess. The h8 tolerance refers to the diameter (clearance fit / running fit, hole basis) but a different tolerance would apply for straightness. Doesn't sound as if they specified that tolerance?

Edited By DMB on 25/04/2016 12:15:14

Edited By DMB on 25/04/2016 12:16:16

I have a 1M length of aluminium bar also and that is... 20mm-30mm -0.08 / +0.13 .. But i dont think those are the numbers your looking for?

Plunger style dial gauge and the small finger style dial gauge both reading in 0.01 increments, I also

have some digital calipers and a 0-25mm micrometer.

Going to 3D print a decent holder for the dial gauges. That magnetic arm never seems to be in the right place. Limitations of a small lathe i guess.

I plan on trying to build a CNC router so i will want to cut my own threads over as much length as i can get onto the machine. So if i can get a bit more accuracy out of it then thats a good thing, but my knowledge and skills deter me. Although i can strip and engine and rebuild it changing the head bearings did have me crossing my fingers it went back together properly.

I did notice that rotating the bar in the chuck did vary the runout quite considerably.

Just inserted it close to the chuck and wound the tool in just enough to score a line, extended it out as far as i could and still reach with the tool and its clearly further away from the tool. So it would cut a taper with the smaller part being near the chuck. I wound in the tailstock and dead centre and it pulled it back towards the tool.

When taking a reading from the top though the runout is just that +/- 0.1mm.

Thanks.

OK, I maybe doing it all wrong but i tried the steel bar this time with the dial indicator attached firmly to the tool holder. No deviation at all (well as long as i dont heave on everything).

Close to the chuck it shows just a wiggle on the needle rotated the full 360. I set the chuck in a fixed position and set the dial to zero, I then moved the carriage as far to the right as i could and took another reading this was exactly 10.5inches away. Do you hate members that use metric and imperial? (sorry)

Lined chuck jaw 1 horizontal to the toolholder facing me as it would when cutting.

Jaw 1 zeroed at the chuck and reads -0.21mm @ 10.5" away

Middle of Jaws 1 --> 2 reads -0.26(8)

Jaw 2 reads -0.32

Middle of Jaws 2 --> 3 reads -0.32(5)

Jaw 3 reads -0.27

Middle of Jaws 3-->1 reads -0.20

These were taken with a ground steel bar and no tailstock, does this mean my headstock is not in alignment?

Apologies is this is meaningless, please show me the light

Thanks

I put my hand behind the headstock area and pulled slightly and the needle moved towards the zero mark.

Seems after 4 years its no longer totally flat on the base. Odd because the board i used was an old coffee table top which is made from some unknown substance thats not wood and not plastic.

Ive had a chunk of it in the garden since fitting and its exaxctly the same now as when i put it out, water and oil does not soal into it. So not the table part expanding. But this table sits on top of a wooden workbench and my fixings went through both, possible the lower table expanded and pulled the lathe down tighter? Also possible that it may shrink during the summer, if we actually get one that is.

I thought the headstock side had 2 mounts and one for the tail stock side, seems my memory is going and there is just the one fixing at either end.

Whats the best way to level a lathe with just 2 fixings?

Thanks.

Dial gauge at the 10.5" mark still which is 9" from the end of the bed, after unbolting the lathe i can grab the end of the bed and twist it 0.1mm by hand. Is that normal?

Edited By Dave Smith the 16th on 26/04/2016 15:03:58