Lathe Packing :(

I followed a procedure I got from the 7x12 forum, briefly put a round bar in a chuck (3 jaw is ok) clock it near the chuck rotating the lathe by hand, average the difference between the high and low readings, move well down the bar and do it again adjust the packing until the results are the same. The bar has to be straight and the same diameter for this particular version of their procedure to work.

Initial I was about 0.15mm out at the chuck and 8 inches down the bar about 0.27 out, I spent about 11/2 hours adjusting the packing and retesting each time and I came out almost exactly as I went in almost no change!!

I know the lathe is past it's best

From my previous post :-

I'm confident that a skilled and experienced engineer could sort it out, there is some wear in bed it's looser near the chuck, the tail stock can move from side to side slightly + so does not always clamp centrally , there is some damage to the tool slide and the bearing clamp behind the chuck has been welded, it must have snapped at some time, but it adjusts now.

If somebody could explain (simply casuse I'm new) what went wrong and how I should proceed I would be greatfull.

Many Thanks Alan

Reply to
ora
Loading thread data ...

The only accurate way to sort the headstock axis is to chuck a bar and take off a skim, then measure it at two points along the length if it is the same diameter all the way along then it's sorted, if not with a bit of trig you can calculate how much to move it by, or put a clock in the tool post and adjust the headstock untill the clock shows good.

Although why you mention packings or shims is beyond me, unless the head is no longer parallel to the bed due to wear, in which case I would doubt adjustment would compensate as the bed would not be likely to be worn equally along it's length.

Reply to
Simon Cochran

For this method to work the bar does not have to be straight. It also doesn't need to be the same diameter although it's easier if it is.

The clock needs to be mounted on the cross-slide and it's position must not shift as the cross-slide is moved down the bar. Also the average referred to is the sum of the min and max readings divided by 2, ie if the clock readings are 2.30 and 2.60 then the average is 2.45. At the other end the readings may be 1.50 and 3.40 in which case the average is 2.45 and the headstock spindle is perfectly aligned with the bed in the orientation that the readings are taken.

Apologies if this is what you have done. But in this case shimming the headstock appropriately is bound to (as far as I can see) affect the readings.

Pat

Reply to
Patrick

On Sat, 19 May 2007 19:30:20 GMT, "Simon Cochran" top posted:

One assumes that he was referring to shimming the bed to twist it into alignment. This is normal practice for all lathes apart from a few, such as Hardinge's that use a floating bead.

Mark Rand RTFM

Reply to
Mark Rand

**One assumes that he was referring to shimming the bed to twist it into alignment. This is normal practice for all lathes apart from a few**

That is exactly what I was trying to do, but I now realise that I did it wrong!! I had the clock mounted on the bed not the cross slide.

My clock is an electronic device, I put it on the bar rotated the bar to find the minimum (most -negative number) zero'ed the clock rotated the bar recorded the maximum number then divided that by 2. Then I re- located the clock down the bar and did the same again.

I'll do it again mounting the clock off the cross slide but as I move the cross slide down the bed I will let the clock run down the rod but without re-zeroing it.

Many Thanks Alan

Reply to
ora

I'm getting a bit old, but I prefer analgoue clocks for "tuning" things (tightening the bed clamps to true a lathe is a good example) and digital for measurements.

Reply to
Steve W

In which case it sounds as though the rod you are using is not of uniform diameter?

Be careful to take the algebraic average, unless one reading is negative there should be no 'difference'or subtraction involved. For this reason it is best to position the clock so that all readings are positive.

Pat

Reply to
Patrick

There's a fundamental problem with the method you are adopting - it assumes that the chuck will clamp the bar so that the axis of the bar runs parallel to the axis of the spindle. Big (and often false!) assumption there. So with this method, any error you measure could be due to a misaligned headstock, or an inaccurate chuck, or a bit of each.

As mentioned by another poster, you need to take any inaccuracy of the chuck out of the equation by taking a skim across the bar and then measuring any resultant taper to tell you how well/badly alligned the spindle is.

Regards, T>

Reply to
Tony Jeffree

I think the method he refers to is known as 'Rollies Dads Method' in which case you don't need an accurate chuck or even a straight bar. The method eliminates these offsets when the average of the min/max readings are taken. The method relies on the clock remaining in a fixed position relative to the carriage as the carriage is moved down the bed.

The obvious benefit of the method is that no cut is taken, although it seems to need a degree in very difficult sums to figure out that the method actually works!

Pat

Reply to
Patrick

But if no cut is taken how can you be sure that in use the theory is correct ? There is a big difference in presenting a spring loaded needle to the work and getting a tool to push against the work under load. Theory is well and good but at the end of the day lathes are made to wang big hot chips all over the shop, dog and surrounding areas but not necessarily in that order.

The time honoured way is to first level the bed as accurately as possible in two planes with a decent level and not one of those s**te things from B&Q. Then get a decent sized bar sticking out the chuck about 4 to 8" depending on lathe size and relieve the centre section leaving two lands, one stuck out in midair and one next to the chuck. Then a light cut is taken in one pass and the two diameters measured. From this a bit of tweaking on the holding down bolts and a recut will get it turning parallel. This deals with the headstock bed alignment, the tailstock is then next and another issue altogether.

-- Regards,

John Stevenson Nottingham, England.

Visit the new Model Engineering adverts page at:-

formatting link

Reply to
John Stevenson

Pat -

Re-reading the OP, I think the problem was in the description - he says "average the difference between the high and low readings..." where what I believe is intended is "take the average of the high and low readings". I would agree that with the latter definition, you do indeed lose any inaccuracy in the chuck or in the straightness of the bar. However, the bar does need to be of a consistent cross section along its length, and if that cross section is anything other than circular, have no "twist" at all.

Regards, Tony

Reply to
Tony Jeffree

On or around Sun, 20 May 2007 11:43:05 +0100, Tony Jeffree enlightened us thusly:

It'd certainly be false on mine. The spindle and chuck body run true but things held in it don't. 8" Pratt Burnerd 3-jaw in this case.

Tried a replacement set of jaws, which were better, and tried stripping, cleaning and lubing the chuck which made it better, but it's still not right. I also tried clamping a large bearing shell from a taper roller bearing on the outer edge of the jaws and then skimming down inner edge of the jaws with a carbide-tip boring tool. That also made it better, but it's still not really true, just good enough for most of what I do with it.

I know that there's a point about 3-jaws are never perfect, but this one seems less perfect than most.

Anyone with any other ideas is welcome to share them - a new 8" chuck is not really and option right now, even if I can find one that fits an L-zero.

Reply to
Austin Shackles

Please don't tell me to f--- off but you seem to be contradicting yourself :)

On the one hand you're proposing a 'light cut' to set the lathe to cut parallel and on the other hand seem to be saying that such methods are not good because the lathe is intended to 'wang big hot chips all over the shop'.

I don't think it matters too much which method is adopted. It seems the real difficulty is figuring out what is actually being measured and how to correct it.

Pat

Reply to
Patrick

Yes, his description is off. But then the description in the original text for RDM is ambiguous to say the least - it's not surprising that people have difficulty understanding it and consequently applying it.

Pat

Reply to
Patrick

"Austin Shackles" wrote in message news: snipped-for-privacy@4ax.com...

How about chucking a piece of rod. Clock the rod close to the chuck and take three readings at points where each jaw is opposite the clock. That should tell you which two of the three jaws need some metal removed and exactly how much. Then remove the jaws and lap the measured amount off.

Pat

Reply to
Patrick

It is a variation of Rolly's Dad's method that I'm attempting to use, at the moment I have all the hold down bolts slackened off and it will be Friday or later before I can get back to it. I'm trying to replace the compound/top slide as it's the only bit with significant damage rather than wear and use.

Thanks Again Alan

Reply to
ora

On or around Mon, 21 May 2007 16:19:07 +0100, "Patrick" enlightened us thusly:

that might work. apart from the ability to work the jaws when not in the chuck, they're hard, and I doubt I could machine them accurately enough.

I might live with it until I can get a new chuck.

Reply to
Austin Shackles

The inner edges of the jaws of my 3-jaw are concave - would lapping work?

I had thought of mounting a small grindstone on the toolpost, and grinding the inner chuck edges while the lathe was runing slowly.

Problems then might include, a) the jaws are not in compression and b) while it may (see a) be accurate at one opening size, the scroll may be inaccurate.

Though if you used collets (rotationally located against eg a mark on the chuck body) and a very teeny grindstone this might be a way to really accurise them - but again, the lack of approriate compression during the grinding operation could be a problem.

Reply to
Peter Fairbrother

On or around Mon, 21 May 2007 21:56:00 +0100, Peter Fairbrother enlightened us thusly:

I think this might be what went wrong with my plan, which appeared to work otherwise - gripping a ring on the outside of the jaws is not the same as gripping something on the inside.

I have seen a picture of a cunning device which claims to allow this sort of work to be done, but I remain sceptical.

Probably the real answer is to replace the chuck - chances are the scroll on this one is worn and so it doesn't track properly.

I see J&L have "Toolmex" chucks in 2 grades. anyone know if they're any good?

Reply to
Austin Shackles

You need a sort of spider-with-a-hole; a ring with three legs of equal width projecting inwards. Close the chuck jaws so that the tapered sections on the jaw faces grip the legs. If you get the length of the legs right you can grind the centre faces of the jaws while they are being forced outwards by the legs against the scroll; and if you get the width of the legs right you can go right into the bore.

The spider does not have to be fancy - three strips of aluminium plate rivetted to a ring will do.

Reply to
_

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.