View Full Version : REBUILD: Rebuilding a SuperRelm Lathe

04-04-2010, 11:05 PM
A thread to document the potential rebuild of my SuperRelm lathe... I need to decide how much I want to do this instead of spending 400+ on a new minilathe or a 1960's Boxford or similar.

The SuperRelm, c1915, is a classic early C20th English flat-bed lathe with the heastock and bed cast as one piece. It is an 8.5 x 24, so considerably bigger capacity than most minilathes, with an MT3 spindle bored 20mm and an MT1 tailstock.


The rebuild will need to cover:

Sorting out the saddle, which binds about 32cm from the heastock end of the bed, but is loose enough to rock slightly about 5cm out;
Sorting out the top- and cross-slide leadscrews and nuts which are badly worn with a turn of backlash plus the cross-slide has 4mm of movement!
Sorting out the cross- and top-slide gibs to remove binding
Sorting out the plain spindle bearings to improve accuracy
Replacing the flat-belt pulleys with a v-belt (not sure about this one)
I think thats the order to do it as each will make the lathe more usable

Here is the saddle stripped down, and a close up of the dovetail and gib (with the vertical retaining plate removed). The saddle is disconnected from the apron so that I can assess how well it fits the dovetail. There is no rocking on the dovetail as far as I can tell.

1961 1962 1963

The saddle binds around 32cm from the headstock end of the bed and loosens up around 5cm out. I marked out every 2cm from the headstock end and measured the width of the bed using the saddle as a guide to keep the digital calipers square. The results show that the bed varies in width from 102.95mm to 103.36 with the low and high points as suspected. Its hard to tell on which face the wear is however a parallel placed on the rear face shows no gap while placed on the lip of the dovetail shows definite high spots. The dovetail face itself seems flat though (how would I check this?). The saddle has no relief at the point of the dovetail; I am wondering whether to mill one there so that the face of the dovetail takes the load and not the lip... or should I hone down the lip?

The steel gibstrip itself is banana shaped by about 0.5mm at the ends so that could do with a new one being made up. Would bronze or brass be the better option? I am tempted to machine a tapered gib and do away with the four gib screws. There is plenty of metal in the saddle to accomodate one. I would also incorporate a saddle lock which this doesn't have.

The top- and cross-slide leadscrews I shall replace by some 10x2G lefthand trapezoidal screw and make new nuts out of some 1" square brass bar that I have... I could buy a tap (at 36!) or use some spare leadscrew to make a tap (JohnS has kindly offered to do so and get it hardened, when and if I get some stock to him). I might make up some delrin nuts initially to make the lathe usable. The leadscrew will have a 6mm spigot turned on the end to join it to the screw extension and handle assembly - more on that later. There is a limited diameter to allow the top slide to traverse over the cross-slide screw extension which will run in a brass bearing but I want to find a way to incorporate two 8mm ID, 15mm OD thrust washers to control endfloat and also a micrometer dial calibrated 0 - 200.

As you can see the saddle has some t-slots and these will provide the means to mount a boring spindle to rebore the bearings in situ. Options are here to repour the bearing and rebore or to bore oversize and fit bronze sleeves, but I am not sure about the wear of the spindle on the bronze, which is harder than the plain babbit bearings....

Am I mad to contemplate this?

05-04-2010, 08:59 AM
If you are mad then its catching!
After all I think a hobby has no ultimate logic to it, I bet you will learn a lot from doing it!
On the cross slide how is the rear side held down as the gib looks to be verticle with nothing shown to hole it down to the bed!.

Go for it!


05-04-2010, 10:00 AM
see the threaded holes in the bottom of the saddle - there's a plate bolts to that to stop the saddle lifting. I'd probably make a new part that allows for a tapered gib there as well.

For the actual gib I reckon there's enough material to machine the gib face to a taper and then put a tapered gibstrip with a slot in it to accomodate the adjuster which will fit into the end of the saddle.

05-04-2010, 12:18 PM
Am I mad to contemplate this?Depends...

I think an historically accurate rebuild, i.e. re scraping the ways to the original accuracy (if the lathe has an historic value) would be OK

Doing the rebuild because it would be fun to re-engineer and improve it, would be OK

But, rebuilding to get a better lathe, would be madness. There are a lot of lathes around in reasonable condition for not too much money.

The dovetail face itself seems flat though (how would I check this?).

You could probably make something like this:

05-04-2010, 12:46 PM
Thanks Bill..

I realise its not going to be better than it was to start (tho you could argue thats not true of the minilathes - as John (Bogs) has shown, theres a lot of fettling that can make them better than as shipped). I want a useable lathe without spending a lot of dosh. reality is that you need to spend a few 100 to get one 'off the shelf' thats any good. Its not just the money (I could afford 400 OK) its something about not throwing away something that is fundamentally sound (assuming that is true). I have a big problem with the throw-away culture. Plus I am keen to learn and I enjoy the challenge.

I dont think the lathe itself has much historical value, it was one of many companies that grew up in the early C20th, flourished during the 1st WW and then died in the depression of the late 1920s. Theres nothing new or special about the design or engineering. But it would be good to give it a new lease of life and, once fettled, I doubt it'll have less accuracy than a modern 7x14 would...

As far as checking the dovetail, part of the issue is trying to decide what reference point to use. There are three possible ones, the rear face or one of the two inner faces. What I cant tell is whether the inner faces are ground well enough to be a reference, and the rear face is the one thats likely to be worn.

I was thinking about the possibility of putting the bed on the mill and regrinding the rear face with a grinding tool held in the spindle. It would just be possible although tricky to mount... (but i do have a spare bed I could try with!)

05-04-2010, 09:02 PM
I doubt it'll have less accuracy than a modern 7x14 would...
Indeed it could be better; it looks more substantial than most of the cheap Chinese models around.

As far as checking the dovetail, part of the issue is trying to decide what reference point to use. There are three possible ones, the rear face or one of the two inner faces. What I cant tell is whether the inner faces are ground well enough to be a reference, and the rear face is the one thats likely to be worn.

I think I'd start with getting the top surface flat and as level as possible (taking the extreme ends as a reference). Have you got a good precision straight edge?

Does the tail stock run against the inner edges? If so, any sign of wear?

The inner edges will have been machined at the same time as the two reference edges, so if they aren't worn too badly, they should at least be straight and in line with the original spindle. If you use a smooth parallel up against them to average out any 'lumps and bumps' you should be able to get a pretty good idea how far the two outer reference edges are out.

If I were doing this by hand, I'd flatten and straighten the rear edge before tackling the dovetail. Use a right-angle and a gauge pin (or the shaft of a carbide milling bit) to check the dovetail.

If you know someone with a good sized milling machine or surface grinder, they probably align and skim all the edges in a couple of hours.

05-04-2010, 09:43 PM
well the inner edges are only used to reference the tailstock. I haven't checked them yet, thats my next task.

I have a box of 150mm x 3mm parallels and a pair of 240mmx20x35 parallels

I measured up the bed and the table of the mill... I could get the bed on the mill if I sat it on some 4" high blocks and shimmed it to get the top level, the legs would protrude out over the ends. I could reference a test bar held between centres to get the bed parallel with the table.. that would allow me to determine which faces are also paralell. I was thinking that if I was then to mount a 2" grinding wheel in a collet in the mill I could grind the rear surface parallel to the test bar, setting the head low enough to use minimal quill travel... then once I had done that I could reference everything else from there...

I am less sure about how to deal with the top surfaces of the bed... i certainly wouldnt want to use a facemill on them - the Z accuracy of the mill isnt good enough. If i mounted the bed on its side and used the ends to reference off that would tell me how far out the bed was. I dont have a long enough straightedge to do it otherwise. I wonder if the grinding wheel in the collet would work here tho as the quill travel is 110mm approx and the bed is 103mm wide.

I dont know anyone locally who could do this. I had a spindle reground locally but that cost me 40... this would be much more than that I reckon...

05-04-2010, 11:37 PM
You might pick up a few tips here: