Well, I seem to have dealt with my cross-slide, and have got the excessive dishing (0.35mm/0.0053” dip in the middle of an 80mm/3.15” workpiece) down to about 0.02mm/0.001”. As I said earlier, I didn’t want to get too near perfection for fear of going too far and ending up with convex facing cuts.
The job was done on my old Dore Westbury miller with a 60° dovetail cutter. Setting the saddle up wasn’t easy. The lathe is Chinese, so nothing is particularly parallel or square to anything else, and the only available datum faces were the two horizontal sliding surfaces on the top. I bought a couple of cheap Ebay feeler gauges to provide me with a variety of shims so I could get those sliding surfaces true. Here it is, with clamps everywhere I could fit one, including some hemming it in, because the “shims” were shiny and slippery, and it would have been a disaster if the job shifted.
Other problems encountered were:
1. Tramming the miller, which had to be done with the head at what would be working height (fairly low in this instance) in case its round column is not dead staight. In that case, altering its height after tramming would throw things out again. I adopted this rather nice technique
http://rick.sparber.org/TM.pdf .
2. The gib had conical dimples, no doubt made with a drill bit. The ends of the M4 adjusting screws weren't pointed or turned down in any way. As the dimples only roughly coincided with the screws, the gib had been twisting slightly as I tried to adjust it in the past. No wonder adjustment was always a compromise between no tight spots on the one hand and no free play on the other. To cure this, I flat-bottomed the dimples with a 4mm slot drill, faced the ends of the screws and turned the threads off their ends (happily, I have an old Perris baby lathe to use while the Chinese one was dismantled).
3. There are no saddle gibs bearing on the underside of the bed on my lathe. The side plates (those which normally carry the gib plate adjusting screws on Chinese saddles) were a close fit to the bed, so there was no discernible rock. In any case, unless I was turning something so large that the tool was no longer over the bed, or was using a back tool post, cutting forces would push the saddle down on to the bed, so the side plates would have nothing to do. But for some reason, after reassembly, the front plate was too tight on the bed, so I had to interpose about a thou of kitchen foil shim between it and the saddle. The rear plate was too loose, so about 2 thou had to be removed by draw filing. The result was pretty good; no discernible rock, and the saddle can be lugged along the bed quite easily, without resorting to the apron handwheel.
All in all, I’m happy with the result. The dishing has been reduced to what for me are acceptable limits, the carriage is nicely fitted to the bed, and the cross-slide gib problem has been eliminated.
My thanks to all who took an interest, particularly Pete (Miner); I hope his recovery continues. Likewise, I trust that Norm (Fergus) is over his pneumonia.
Andy