Glad it worked out for you. With the quality of parts available lately it makes sense to check everything the way you did and end up with the cheapest solution to boot.

Well, I've measured the play using a dti as suggested, with the bearing being a tapered type I measured it as close to the same plane as the bearing as possible. I measured the play at 0.006" so adding the required preload amount 0.001-0.002", 0.007-8" (or 0.2mm) was the magic number. I had a spare spacer 0.004" smaller, so that took care of half of the amount, I then removed the rest with 120/180 grit glass paper, measuring with the dti.
As I still wasn't certain the play was from the bearing, so I assembled the bearing in the hub WITHOUT the spacer so the races were held fast (not to full torque) then felt for play, of which there wasn't any, thereby eliminating the stub axel etc (if you think otherwise please share)
Anyway, spacer replaced hub torqued the play has gone, I'm over the moon that the faffing has equated to a result, a cheap one at that.
Thankyou for the help and information!!

I have done a few hubs over the years. Have saved all spacers removed. I also found a length of steel tube correct inside and outside dia. can part for custom spacers. Made a jig to hold hubs in vise so I can fit races and set up hubs on bench. Have a couple of Mk 1 door glasses with emery paper glued to both sides for lapping spacers. Also made a set of shims .001 though .010 for testing. Have seen plenty of hubs with worn race seats. Some have dimples in that area and some have signs of Loctite used. I imagine for personal use they could be used. I replace as I must stand behind what I do. You can buy hubs and they aren't that expensive. Recently been machining seal spacers for the hubs. They fit between race and seal  to locate seal lip correctly. The inside ones are a little thicker than the outer ones. Steve (CTR)

If you have new bearings correctly fitted and still have play then you have a wear issue in the hub or drive flange. You have two choices replace the worn items or custom fit a different spacer. Since this appears to be a recurring problem I would go with replacement of hub or drive flange or both

When installing the new races, I used a brass drift and carefully tapped it down, as the race travelled over the part of the hub that although was thoroughly cleaned wasn't as well finished as the actual seats it was slow progress, but when on the last 3/4 inch or so a couple of taps and it was in, comparatively easy. The race would then not move, or rotate by hand. If the race should have resisted until the moment it was seated then this may well be where the problem lies. The play appeared in a 12 month/1000 mile period. The old  bearing didn't look worn (to my untrained eye anyway) and the play remains with the new bearing fitted,  so I am wondered if there is a problem else where.

Just to clarify. Outer race (cup) of tapered bearing is an interference fit (pressed in). Inner race bore to  shaft is loose fit (.001 loose?) should slip in but not have radial movement.

The bearing races shouldn't drop in.  They should require the use of a press or a brass drift to drive them down to their seat.  If they dropped in, you have identified an out of tolerance problem with the hub.  This could be a manufacturing problem or the result of damage during a previous bearing replacement.

Hi, we're talkin taper roller bearings, It makes sense now that you were previously on about ball bearings. There was no sign of wear or blueing of the seats, the inner race perhaps dropped in a little easier than expected, but i don't believe it to be of concern. I have a dti on order, I will try to measure the play (rocking, not axiel) on the same plane as the bearing and go from there. Thanks for the help.

Mechanic is too nice of a term Doug.

If the play is 100% axial and you were using ball bearings then it would not matter where the measurement was taken.  However, since you don't know if the hub is worn and you may have tapered roller bearings, you may find more reliable and repeatable measurements by placing the dial indicator as close to the center axis of the hub as possible as Malsal suggested.

I don't want to re-read all of this thread from the starts so I apologize if you have already answered the following.  When you replaced the bearings did you inspect the inside of your existing hub for damage in the bearing seats?  I have heard of people whose cars have had the bearings "spin" and ruin the seats.  I was also surprised when working on a friend's Mini to find a previous mechanic had "dimpled" the bearing seats using a poorly aimed drift.  The dimpling prevented the bearings from fully seating.

Hi islandblue, thanks for your reply. I think I grasp what you are trying to say, but wouldn't the play get exaggerated the further out the pick up point i.e outer edge of the disc vs most inner point on the disc? Cheers

Hi Yorkspv - sorry, I don't frequent the board very often anymore, and happened to drop by and see your question on this old post.   Which spot you pick on the rotor to touch off the indicator is not critical.  Wherever your adjustable arm will allow the dial indicator to reach.  Any movement will appear on the dial, as long as you start with a spacer that is slightly too wide.   The key is to make sure the magnetic base is secured tightly to the hub (or some other stationary surface that is not the rotor) using cable ties, clamps or which ever means prevents movement.  This ensures that all indicator needle movement is from the rotor. 

I would work on the spacer first before buying a hub assembly. Measure the play as close to the center as possible.

There is a decent amount of play, along with a knock. I can't imagine it would get through an mot. There was play before I changed the bearing, hence why I changed it. I may just change the hub and see if that sorts it. But with all these spuriously sized bearing spacers who knows!.....

How much "play" are we talking here? If it's a small amount just felt not a big old rock then I suggest as long as it feels good when rotated not notchy or noisy in anyway you leave it alone, a little play is far preferable to having the bearing too tight ie too much pre-load as this will destroy the bearing and likely the hub as well.......

Yes. 60 lb. ft.

Hi islandblue, could you please explain where you are 'picking up' on the brake rotor? I would have guessed at the most inner part of the rotor surface? I am struggling with hub play which has persisted after replacing the bearing. Thanks

3 months later and about 2000 miles and my right hub again had play this weekend, and I got 1/2 turn of the axle nut to tighten it back up. 

This time I am going to throw parts at it.  The bearings I used previously to replace the bad ones were matched to races and were lightly used and turned smoothly, but they were lightly used.  

The races had not ever spun in the hubs, and to ensure they would not, I made a few punch dimples in the hubs and applied Loctite bearing retainer compound to the hub and race before pressing in the races.  Everything was dry assembled to ensure .002 preload.

So even though I cannot see what could possibly change by replacing the swivel hub with new, I am going to replace that along with the bearings.   Drive flanges have barely 5000 miles on them.

The left side was set up the same way, and it is still tight.

This is a frustrating problem that has been going on for years.  

I use a dial indicator.  DRMINI explained the method to me a while back.

I cable tie my magnetic indicator base around the base of the hub on the front side of it just below the steering arm mounting bolts, and position the indicator so it picks up off of the front of the brake rotor.

The trick is to start off with too wide of a spacer so you have some play to measure.  Torque the hub nut to spec with a flat washer under the nut (not the cone washer), and then push and pull the brake rotor and read the needle sweep on the indicator.  If for instance it is .015", then you need a spacer that is .016 - .017" thinner than what is in there to give you .001 - .002" preload.

With the proper spacer in the hub and the nut torqued to spec, you will not feel any play, or read any play on the indicator.  The spacer prevents the bearings from pushing too hard into the races and prematurely failing.

Can you explain how you are measuring the preload? Are you using a dial guage? or?? Thanks