Thread: RDG test bar for lathes

You place it in the spindle and run a dial indicator along it. You can just put a piece of round bar in the lathe and machine it, then measure. It's much cheaper.

With using a factory test bar, who knows if the MT in your spindle is really true, where machining a bar will tell what the lathe is aligned like and take out any of these variables.

Someone here bought one, maybe they could lend it out for a small fee if you really want to use one.

Dave

Hi Mike
Make a test bar yourself using the shaft from a mcpherson strut.
High finish steel and probably pick one from the wreckers for not very much.

Roger

Just bare in mind using something like that in a lathe spindle will most likely not yield very good results if you have to use a reducing bush in the headstock that will most likely introduce errors in itself...

If you stick a reducing bush into your headstock and then the test bar, and you get an error, how do you know where the error is? Is it in the reducing bush or the spindle taper?

It maybe would be handy for between centres alignment e.g., the tailstock and the spindle centre because the bar is centred at each end , at least you know that you are beginning the alignment with an accurate centred bar , and you are not guessing by using a home made jobby

Mike

Originally Posted by morrisman

These test bars seem to be rather short - at around 300 mm in length .

G'day Mike.

300mm long is almost too long for a #3 morse Morse taper mandrel. It will be prone to drooping which can be an issue depending on how you are using it. The large end of a #3 taper is less than 24mm, so you have about a 12.5: 1 ratio length to diameter. That will sag, better more expensive examples are actually hollow with a plug in the end for the center.

For a lot of tests that might not matter, any thing vertical for instance, like checking the radial run out of a drill or radial arm. If you were checking the vertical height difference between a head stock and tail stock, it would come into play.

In that EBay blurb, it seemed to suggest just putting it between centers, which would work, but the Morse part is unusable in that test. The Morse part is very handy for blueing up your tapered sockets. To see if they are bell mouthed etc.

Whilst I'd agree about turning a mandrel out of the head stock, these things really shine where you are checking tail stocks, and cant turn a test bar. There's all sorts of tests you can confirm with a test mandrel. Contact of the socket. Parallelism of the socket to the bed in both planes, and to the quill. Relative height to the head stock, etc.

Regards Phil.

For this you could make up a test bar (say 25mm) center drilled both ends with 2 collars on it just in from each end, then you take a cut off one collar and without moving the cross slide (lock it) move the saddle along and take a cut off the other one and measure. Adjust the tail stock as necessary and once it's true you set up is true. I use a nice sharp HSS bit for doing this.
You can reuse this bar once it's true both ends many times over.

The collars can be machined in, pressed on or welded/silver soldered on.
To protect the ends from a drop/bump you could bore the ends slightly and then center drill them, this would give them protection.

Dave

The question of how to use a test bar -
MSlatheinspection1.jpgMSlatheinspection2.jpg

These two are an extract from JIS standards I believe. The numbers on them tie up with Schlesinger and MTR, so I suspect that the world over everyone has agreed (probably copied Schlesinger as I think he started it all). As you can see from the diagrams, most of the tests are done with an indicator running along or around the bar.
I have a 2MT test bar, I think from RDG which I use occasionally. It's not hollow, so as Phil says it can droop. When ever I feel brave and try to use it I'm never sure what contribution is from the tailstock being out and what contribution is coming from the lathe bed. Until I get that ground/ scraped so I know it is flat (at the moment all I know is that it's not) there is probably not much point in trying to assess and correct tail stock errors.

Michael