Restoring (hopefully) my new Smart and Brown SAB

I recently started a thread about using an MT3 ER32 collet In the lathe and seeing a lot of runout. This thread went down a rabbit hole and ended up in identifying that my lathe has 0.08mm spindle runout. I thought I'd bring that thread back in to this one as it's now more about fixing the problem on the lathe rather than the MT3 ER32 collet that originally started that that thread (https://www.hobby-machinist.com/threads/mt3-collet-chuck-in-lathe-runout.104810/).

Here are a photo of measuring the runout in the spindle both internally and externally (|0.08mm which probably isn't clear in the pictures):

9d43493262bb56a319e3098d398f3a2f.jpg

And the play in the spindle:

865bcee2c1611feda7fdb2e3417ffb77.jpg

IMG_0093.jpg

I used an MT3 tail stock alignment bar in the spindle to measure these results. I also had the indicators on the spindle and got pretty much the same measurement. This was able to be solved by tightening the cap on the back of the spindle.

This did not however fix the runout, so:

3fa81c04904c1e70cbb430acd33e59f4.jpg

The spindle is now out on and on the bench ready for some testing.

I got some really good feedback in the other thread about measuring spindle runout and spindle play, but I'll only copy across the last few posts as they inform where I'm going here:

The bearing on the picture is a trust bearing to absorb the axial forces. I assume you lathe has a bronze bearing, no ball bearing!

You can place the spindle on 2 V-blocks. Place the V-block at the bearing position. Put an indicator on top and turn the spindle by hand. There should be no run out.
If you don't have a v-block, measure the thickness of the shaft on multiple circular positions. It should be equal on all position.

To measure if the spindle is bent, you need 2 v-blocks. Place the spindle on the (fixated) v-blocks and rotate the spindle. Place the indicator on a ground surfce between the v-blocks. A run out indicates a bend spindle. The angle of the V-block is not important and you could make one your self from wood or aluminum.
If you have made the v-blocks, you can measure the inside run out again. If it is 0.08 mm, than the spindle is the cause.

You can put the spindle in the lathe and measure the runout at the chuck and far away from the chuck. The difference in the 2 values is an indication for the bent of the spindle.

Huub

So that's where I'm going from here. Measure the spindle on the bench and see what I get.
 
I have thought again about the inner and outer "run out" measurements:

If the spindle is OK, you would expect that when the TOP outer indicator is at his max, the BOTTOM inner indicator is also at his max. That would indicate that the taper bore is in the center of the spindle.
If you position the inner and outer indicator at the top of the spindle, the difference in inner/outer readings would indicate a difference in wall thickness. If the spindle bore is OK, the wall thickness should be equal around the spindle.

If the thickness of the spindle at the bearing position(s) is the same around the whole 360° than a "play free" bronze bearing will have no run out.

If there is a bronze bearing on the back of the spindle, you should check and adjust that side also.

You could mount the spindle without the gears. You can than put an indicator at a all ground surfaces between the bearings. If the spindle is bent, it would be bent outside the bearings.
 
I’ve done a bit of measuring on the spindle:

603aea8a6bd50fdf8e79b40e94429023.jpg


Measuring the above shows around 0.005 runout. That’s good. What’s no so good:

19e4941ccb6a86e03f6f1cb6abd8311f.jpg


This measures about 0.08 runout. I’m guessing this is where the runout comes from when mounted. These journals (??) run in the bronze bearings (??) in the head stock:

e5eb20ee4c45971e8364a48aa0d28c53.jpg


The rear journal (??) also has a similar runout. Im guessing this was in the past run without proper lubrication and has therefore ruined the journals on the spindle. Im really not sure what to do now. Im guessing there isn’t a fix I can do at home and this would be a case of getting it sent out to put in a working lathe and skim the surfaces of the journals?

Just to confirm the issues that I’ll have if I don’t fix this. I’ll still be able to do precise work, but if I have to remove the work piece and remount it, I’d need to remount in a 4 jaw chuck and centre it in? Otherwise remounting the work piece I’m even a collet would result in a ‘large’ concentricy error?


Sent from my iPhone using Tapatalk
 
Yes, if you need to remove and remount the work piece, you can use a 4 jaw chuck to adjust for the run out. There are adjustable 6 jaw chucks but those are pretty expensive.

You could make an adapter plate for this lathe. If you turn the registering diameter and face on this lathe, it will be running absolutely true. For this reason, I made my own collet holders for my mini lathe.
Such an adapterplate makes only sense, if the runout of the chuck is smaller than the run out of the spindle. So make a mark where the top run out position of the spindle is. Clamp a turned bar in the spindle and measure the run out. Losen the jaws a few turns and reposition the bar exactly at the same position and measure the run out again. Repeat this a few times. This will give an indication about the repeatability and run out of the chuck. To be sure, repeat the procedure using a lot thicker bar.
 
Yes, if you need to remove and remount the work piece, you can use a 4 jaw chuck to adjust for the run out. There are adjustable 6 jaw chucks but those are pretty expensive.

You could make an adapter plate for this lathe. If you turn the registering diameter and face on this lathe, it will be running absolutely true. For this reason, I made my own collet holders for my mini lathe.
Such an adapterplate makes only sense, if the runout of the chuck is smaller than the run out of the spindle. So make a mark where the top run out position of the spindle is. Clamp a turned bar in the spindle and measure the run out. Losen the jaws a few turns and reposition the bar exactly at the same position and measure the run out again. Repeat this a few times. This will give an indication about the repeatability and run out of the chuck. To be sure, repeat the procedure using a lot thicker bar.

Thanks Huub.

I have a good 4 jaw, so that’s a good start.

I see what you’re saying about the back plate. If it’s turned on the lathe, it’ll be aligned perfectly with the lathe. I got three chucks when I bought the lathe, each of which has a backing plate, but I’ve no idea if those back plates were made on the lathe itself, or on another lathe. I’ve also no way of telling how long ago those backing plates were made I.e they may have been made on the lathe, but so long ago that the lathe may not have had the 0.08mm runout at the time. Can I measure the runout on the backing plate by indicating the OD on it while mounted to the spindle? Or does it not work that way?


Sent from my iPhone using Tapatalk
 
All my chucks are mounted on a back plate (adapter plate). I turn the register on the backplate (chuck side) a bit to large and use the play to centre the chuck. That minimizes the runout.
My back plate, that is mounted by 3 bolts, is marked so it can be attached the same way it came of. I did turn the OD of the back plate so I can check the runout of the spindle and it reduces the unbalance of the spindle.
I faced the spindle when I got the lathe so that it runs true, all bits help.

You could/should check the OD of the backing plate when mounted on the spindle. It is hard to tell what the result will be, just try it. The results will give you something to think about.

Dialing in a part on the 4 jaw chuck is time consuming when you start doing it. In time it will go al lot faster. Some machinist use the 4 jaw chuck for all their jobs.
 
I’ve been a bit disheartened with the spindle runout over the last week and haven’t been too sure what to do, so while I have a think about it, I’ve got back to the ELS.

I’d had this mostly up and running before, but I’d broken the Clough42 board. I got a new one delivered a while ago so I got it back out and on the bench and all hooked up:

4a423b85557a677d07993c6adb9d071e.jpg


All seems to be working well. Now to try and get It mounted.


Sent from my iPhone using Tapatalk
 
I've not posted on this thread lately as I've not done much lately. I've had a bit of time over the last couple days so got two jobs done.

The cross feed power feed hasn't worked since I got the lathe, so I investigated it. The shaft that the clutch knob sits on was coming too far towards the apron casing and blocking the power feed selection knob from engaging with the cross feed pin whole. (I don't think that makes sense, but I have no pictures unfortunately). So I made a small bushing that forces the clutch shaft to be out of the way the power feed selection plate. That probably made no sense, but it's working now.

The lathe also didn't come with a z axis lock. I was doing some parting off recently and it seemed that the lack of locking the z axis was causing a problem, so I made a basic lock.

IMG_1082.jpg

Noting fancy, but it works. I tried some parting off now I have the lock in place and it seemed a lot better.
 
A couple of other jobs done now.

I fitted a touchDRO system to the lathe:

96a881ab210c56596da82e89372447ea.jpg


952795a52029491cd2bd51569b36ec59.jpg


I won’t show the bracket I made to hold that on. It’s not pretty, but works.

I also trimmed up the t-nut that came with the QCTP to use instead of the one that was already on the lathe. This meant I could tighten the QCTP centrally rather than having to use a whole I drilled and tapped on the old one thats near an the end of the t nut. I figured this would be more secure.

53d6f9cd4c2da0aca1aadc560d9e728b.jpg


642aba7d2c4fc79535a15241aee5ccd3.jpg


As you can see, I got my measurements wrong when trimming the t nut and the bolt hole is off centre. Stupid mistake, but I’m still learning….


Sent from my iPhone using Tapatalk
 
I will need to get a guard over the scales asap as I don’t think they deal well with having grit/dust/fluid on them.


Sent from my iPhone using Tapatalk
 
Back
Top