Thread: Squaring my lathe

Originally Posted by ftkalcevic

Just a quick question...

After aligning my spindle parallel to the way, I found that I now face convex.

I'm about to realign the saddle - should I aim for dead square or slightly concave? If concave, how much?

Thanks,
Frank

Some more detail about the lathe size would help. I do not know a whole lot, but I can tell you that it should face concave by a small amount so anything that is faced will sit flat. It is no good facing something flat, then finding it is slightly convex and won't sit flat.

I am sure that a more accurate answer will be provided in time, but lathe details would assist in this process.

Dean

The lathe is a CQ6123B (Herless), or a generic 9x20 lathe.

To turn a parallel bar, I would always have to place the workpiece between centers - I could never just mount it in a chuck, no matter how short it was. It wasn't until I replaced the headstock bearings and aligned the headstock that I found the problem.

https://dl.dropboxusercontent.com/u/...ne%20Tools.pdf

What is your skill level and machine tool repair experience?

How would you go about "realigning the saddle"

Hi there,

Before you even consider touching the saddle, I'd be interested to know what test you performed in order to get the headstock parallel, what test you performed to measure the concave from the cross slide. Also how/if you levelled your lathe and to what degree of accuracy before you conducted those tests.

If you are a seasoned/experienced machinist then apologies for the basic questions.

Cheers,

Simon

I am a rank amateur at machine tool repair. After I finish the lathe, I will have completed 1 project

I attended the 2nd Melbourne scraping class with the aim to readjust the lathe, but every time I looked at starting the project I just baulked. I get intimidated by the 3 surfaces - the vee and the flat. I did machine a small test piece that had the same profile as the saddle and tried to scrape that to fit, but the vee is very small and had trouble 1) bluing it, and 2) squeezing the scraper in the small vee. How I scrape to fit AND align is beyond me.

I levelled the lathe using a machine level running on the saddle the length of the way, adjusting and shimming the tailstock end until it appeared level with minimal twist. My lathe is mounted to a big steel framed bench (purpose built), but that sits on a wooden floor. If I move my weight from the front to the back of the lathe, it registers on the machine levels, so the leveling is as good as I could get it trying to keep my feet in the same spot.

I used "Rollie's Dad's Method" to align the head stock. After it was aligned, turning parts mounted in the headstock came out parallel - unlike previously where I had to hit it with a file or wet and dry to reduce to size.

Because I was super confident my lathe was aligned properly, I decided to take a skimming cut of the face plate. That is when I found it was cutting convex. My straight edge test wouldn't lay flat.

I have also done a cnc conversion on the lathe. When the saddle changes direction, there is a definite wobble of the saddle, implying either a poor fit on the way, or the crappy "gibbs" holding the saddle down.

My plan of attack...

1. double check the head is still aligned
2. mount a straight edge in the chuck and align that so it is perpendicular to the way (use DTI on one end, turn spindle 180 degrees and DTI again)
3. Make some adjusting plates to align the saddle - these will be mounted where the gibbs are with set screws pushing into the side of the lathe to turn the saddle
4. Adjust the saddle squareness by running the DTI on the cross slide against the straight edge.
5. Use moglice to make the saddle fit.
6. I am also planning on putting in oil groves, and press oiler buttons. I've read moglice can be damaged easily by swarf getting drawn underneath, so I need to also set up decent way wipers. While on the topic of way wipers, is this a real website - http://waywipers.com.au/ ? It looks incomplete and I don't want to email a potential scammer.
7. When I know the saddle is riding the way correctly, I'm going scrape or file the surface the saddle gibs contact. I have already made tapered gibbs but they don't work because the irregular distance between saddle and gibb bearing surface.

I have already made a test piece to see how the moglice works. I found it tends ooze everywhere and sticks very well to all surfaces - use lots of mould release.

Frank

Originally Posted by ftkalcevic

While on the topic of way wipers, is this a real website - http://waywipers.com.au/ ? It looks incomplete and I don't want to email a potential scammer

Yes surprisingly it is real... The domain is owned by a company called Enkosi Pty Ltd http://www.enkosi.com.au/ who list http://www.enkosi.com.au/pmachprot.html

I do not really know where to start... But I will say this...

If you want to put moglice on the saddle, you will have to mill it down, then build it back up to correct height...

Whilst you have mentioned the cross slide misalignment... Have you considered the lead screw alignment and the rack gear meshing?

I assume moglice is going to need a minimum thickness...

That's quite a lot of work to perform for a novice. Some of the tests you describe maybe flawed too. Some other members here are more skilled/experienced with such matters. I'm not sure how reliable it is to place the level on the saddle and move the saddle up and back. I was under the impression it needs to sit on the actual ways, either the V or the flats but certainly not put on top of the V since this is not necessarily machined flat to any level of precision. If your floor moves so much that it affects the level of your lathe then I suggest you start with the floor BEFORE anything on your machine. It could be that if the steel frame bench is solid enough, that the actual lathe still stays flat (does not twist) but the entire bench moves. This should not affect your readings but how would you know? It would be a shame to go to all the trouble of aligning your lathe and actually removing metal from precision surfaces only to find that you were aligning everything to a twisted bed.

Do a search on this forum for Cross slide scraping. A member here (Stuart) hand scraped his cross slide. Only one component, yet the thread went for quite a few pages and it was quite a lot of work. There was much talk about techniques, alignment and measurement, much of which was quite technical (but very interesting) and certainly not for the faint hearted or novice. I fear you can easily do more harm than good attempting such a project without some experience under your belt. I mean no disrespect to you but it could be a possibility that your lathe (the way it is) is capable of more accurate work than you. This is the reason I have not touched my lathe, although I am reasonably happy with it, it can be improved. However, it is still more capable then me.

Also, obtain a copy of machine tool reconditioning. There are a few well known authors of such books. They are well worth a read before you start removing any metal. It describes all the proper tests to perform and gives the accepted tolerances for each test.

I hope I have helped.

Simon

Upon pondering it some more..

I could possibly think this may work...

Now the first part dealing with leveling and alignment is dependant on your lathe's headstock sitting on the bed on flat ways therefore it is adjustable... The reason why I mention this is what old Rollie's dad method totally fails to mention is it is mostly suited for US style lathes where the headstock is not adjustable as it sits and is scraped into a V way.. Most of your chinese/japanese/euro/british designed machines the headstock sits on flats and is adjustable...

Rollie's dad (and other methods that twist the bed) makes the assumption that the machine had had it's headstock aligned correctly at the factory in the first place... If it has not been aligned properly and use use a bed twisting method to align it, your tailstock will never remain aligned as you have twisted the bed...

Firstly grab your machine level again.. You will also need some small parallels..

now level the bed with the level on the parallels which themselves are sitting on the two flat ways...

Get the bed without twist... Do not worry about putting the level on the saddle..

Next we take the paper you have rollies dad's method on to the toilet and mail it away, never to be seen again...

OK, now we have a bed straight and twist free, well it should be...

Since you do not have a test bar for the spindle, chuck up some steel say 150mm sticking out of the chuck, and maybe 40mm diameter, and machine it so you have what looks like two collars on it about 25mm long.... At either end... Using the same setting on the cross slide, in fact just start the cut at one end and let it run through to the other, take a cut and measure the difference... Make sure the cut is light with a proper sharp HSS tool, otherwise you will deflect the bar and render the measurement useless...

Loosen the headstock retaining bolts and adjust the headstock until you get both collars within an appropriate size... Be prepared for much frustration..

Now do your cross slide alignment test...

Since you mentioned the machined had been CNC'd I take it there is no power cross feed, that is powered by the lead/feed screw.. That simplifies matters considerably..

So you can build up the bed with the moglice, then shim the apron/saddle interface to maintain lead/feed screw alignment..

Not sure where to start either.
First RDM isnt about headstock alignment.. its a way of making a worn lathe cut parallel(or at least close to it)
In itself it is not likely to be good enough for alignment.(though you never know your luck lol)


The headstock on this lathe mounts on a flat surface?

How worn is the lathe? I used my bed as the master for my carriage but my lathe is effectively new.

Stuart

Hi RC,

I assumed he used RDM to move the headstock... not twist the bed(which I guess really means he was just use "two collar method". but its still not really good enough by itself)

Originally Posted by .RC.

Next we take the paper you have rollies dad's method on to the toilet and mail it away, never to be seen again...

Originally Posted by .RC.

Since you do not have a test bar for the spindle, chuck up some steel say 150mm sticking out of the chuck, and maybe 40mm diameter, and machine it so you have what looks like two collars on it about 25mm long.... At either end... Using the same setting on the cross slide, in fact just start the cut at one end and let it run through to the other, take a cut and measure the difference... Make sure the cut is light with a proper sharp HSS tool, otherwise you will deflect the bar and render the measurement useless...

Couple of things I'd add here. Once the collars are turned mount a DTI on the tool post at center height and see what it has to say about the collars. Then do the same vertically(just to be sure)

Stuart

To add to what Simon said, there are 2 of us that have scraped out cross slides.
https://www.woodworkforums.com/f65/mars-lathe-crosslide-scraping-taper-gib-163993 mine is a mere 4 pages
https://www.woodworkforums.com/f65/cross-slide-scraping-152680 Stu's is 24......

What i would like to know is where the Moglice came from? For smaller lathes with easy to move parts i don't see a problem with Turcite, but since i have a 100kg plus saddle to scrape in i was wanting to use Moglice to save the handling issues.

Cheers,
Ew

Originally Posted by Ueee

What i would like to know is where the Moglice came from?

Ew.

The Australian agent for Diamant is this mob. http://www.dynaref.com.au/

They generally don't carry stock. Shelf life kills it. Its an order, pay and import thing.

Regards Phil.

Originally Posted by Stustoys

Hi RC,

I assumed he used RDM to move the headstock... not twist the bed(which I guess really means he was just use "two collar method". but its still not really good enough by itself)

The genuine RD method like the two collar test Hercus advocates, you shim under the tailstock end of the bed, to theoretically "untwist" the bed... The method probably works, but only under certain conditions that like all these methods never ever mention... So some poor bastard buys a machine, reads about the RD method then ends up with as much hair as Other Phil (steamwhisperer) as they tear it all out wondering what they are doing wrong because they cannot get anything right but rollies dad method is gospel..

Hercus themselves advocate when setting up the lathe doing a two collar test and shimming the tailstock... But that is because they assume their workers scraped the headstock into correct alignment in the first place and the Hercus headstock sits on the V's so it is not easily adjustable..