Test Bar in Harold Hall’s Lathework A Complete Course

[James Jenkins 1Participant @jamesjenkins1]

Hi all,

Has anyone else tried Harold Hall's test bar for checking the lathe is set parallel? For those who haven't read the book, it suggests mounting a 1" x 5" piece of mild steal in the jaws of a 3-jaw chuck and then taking light cuts, before measuring the diameter to ascertain if it is parallel. It's a great idea, but with tailstock support I was getting so much chatter and vibration that it seemed pointless measuring as any deviation was likely to come from this and the resulting poor finish on the far end, rather than the lathe bed.

Anyone tried it successfully?

James

Edited By James Jenkins 1 on 27/06/2018 18:49:09

[James Jenkins 1Participant @jamesjenkins1]

[Ian Skeldon 2Participant @ianskeldon2]

Hi James,

I am assuming that you have at least 3 or 4 inches of test bar inside the chuck/spindle and not just supporting it with a small amount of metal in the chuck? If so you should be fine to just take a couple light cuts without the tailstock, then mic it up to check for parallel.

If you can not hold enough material because of spindle internal diameter restrictions then use a smaller diameter. You don't need to use the tailstock until you want to align the tailstock. There is no point in aligning the tailstock until you are happy that your lathe is turning true.

[LambtonParticipant @lambton]

James,

Yes it works I use the similar method detailed in my Myford S7 handbook. You must not use tailstock support as it would give a false result. It is best to rough turn a reduced diameter section in the centre of the test bar then just take a light cut at both ends without altering anything. This saves a lot of time when actually making the test cuts. If you cannot take a light cut (just a few thou in this case) without chatter there is something wrong with your chuck, headstock or poorly adjusted slides etc.

Eric

[James Jenkins 1Participant @jamesjenkins1]

Hi both,

Thanks for your swift replies. Actually the book suggests only 1" or so being inside the chuck. I think I had closer to two, with an overall length of 6". Not sure my chuck would allow much more. I appreciate the need to not use a tailstock, due to this altering the readings.

I did change chucks and that helped, but still quite a bit of chatter. I have only just adjusted the headstock, so there is only 1" to 1 1/2" of play in that. Maybe I need to give it a go again tomorrow. Once done I want to do the tailstock alignment bar, which should be easier being supported from both ends.

James

[duncan webster 1Participant @duncanwebster1]

Test bars are an extravagance, taking a cut at the end of a bar stuck 5" out from the chuck is asking for deflection and inaccuracy. Google Rollie's Dad's Method. A bit more long winded, but just as accurate

[James Jenkins 1Participant @jamesjenkins1]

Hi Duncan,

Thanks for your message. I did a variation of that method, aligning a piece of 1" silver steel in the 4 jaw, then set the tailstock end (tailstock obviously not engaged) to the middle measurement (by rotating the chuck) and then ran the dial indicator along the saddle. Over the length 12" there was a little variation (she is 80 years old), but no more than 0.0005". I was just wanting to double check this with a cut really – only really because Harrold told me too!

James

[Ady1Participant @ady1]

On a hobby lathe if you want accuracy you must use the tailstock

Fixed centre is best but a well supported live centre can be good

once you get it right the lathe makes it look easy

but its actually you who makes it happen

GL

[HopperParticipant @hopper]

Too much slack in headstock bearings. It needs to be .0005″ or less for good results. My m type will turn within half a thou with smooth finish on this test.

Edited By Hopper on 28/06/2018 05:14:22

[Thor 🇳🇴Participant @thor]

Hi James,

I have used Harold's test bar method to align my lathes. I used a slightly larger diameter bar – 30mm – and used epoxy to glue two collars onto it. One close to the chuck, the other near the tailstock end. For testing if headstock is parallel to bedways the bar was not supported by the tailstock, and only light cuts taken. This worked well. Since the testbar has centre dilled holes at each end I can use the same testbar to test if the tailstock is in line with the centreline. This has worked well.

Duncan's mentioned the Rollie Dads method, that works as well.

Thor

Edited By Thor on 28/06/2018 09:28:26

[Peter G. ShawParticipant @peterg-shaw75338]

Might I suggest Tubal Cain's book Workholding in The Lathe (WPS 15) where in Chapter 9 he discusses all aspects of lathe alignment, including various methods of checking said alignment, Harold Hall's method being one of them.

Peter G. Shaw

[Neil WyattModerator @neilwyatt]

Haharold's method has the advantage of being a real world test, while Rollie's Dad's method is like the Schlesinger tests – it doesn't tell you what the lathe will actually achieve when you use it.

Neil

[HopperParticipant @hopper]

Also, I have found with the old M type, it does best with a very sharp knife tool with just a very small radius honed on the point with an oilstone. Use LH Sparey's suggested geometry of a lot of side rake on the top of the tool, about 20 – 25 degrees or so, giving a total "angle of keen-ness" between the leading face and top face of about 55 or 60 degrees. Sharp sharp sharp. Likewise, follow his suggestion of the small radius on the tip. I've found larger radiuses induce chatter if bearings are a bit worn etc.

[James Jenkins 1Participant @jamesjenkins1]

Hi all,

Thanks so much for all your replies, really appreciated. I worked on this yesterday and got the tail stock within 0.0005", although the surface finish wasn't all that exciting. I think there are three things go on with this:

1. The fine feed (from memory around 6&quot is a bit course

2. Possibly some vibration from treadling

3. Tool geometry – free hand sharpening is not allowing me to get this 100% right and I have just bought the Harold Hall Sharpening book with the intention of making the simple tool rest. What grade wheel do people recommend for HSS? Mine needs replacing and Harold goes through the differences, but doesn't actually recommend one.

James

[HopperParticipant @hopper]

Hand feed with the leadscrew handwheel will give finer feed. Bit of a circus act while pedalling at the same time though! You need to make the Duplex fine feed mechanism. Its a 2:1 reduction using two pulleys, one on the spindle, one on a stud where the reversing gear usually goes. I used toothed cam belt and pulleys with about 20 and 40 or 48 teeth respectively. The 20T gear is pinned to the larger gear and drives the rest of the change gears in the normal manner.

Grinding wheels. So long since I bought one I dont remember exact grades. Usually whatever grey wheels the local hardware store has in stock, Coarse one end, fine the other. Then a bench oilstone to rub the tool on and fine finish it afterwards. With a bit of practice you should be able to get a good tool done freehand. It's just putting angles on flat planes really.

[Author]

Posts