arnoldb

I haven't been very active on the forum of late; work life got seriously in the way of hobby time for the last 6 months, and things are not letting up.  Not all bad though; at least I'm fortunate enough to really like my job and the challenges getting thrown my way  - said with a nod of the head to Zee who appears to be in a similar position.

My dad (swarfmaker) forwarded this link on - and it might be of interest. 

Any aspiring engine builder out there without a workshop and with some paper, glue and a hobby knife ?
https://www.youtube.com/watch?v=Pf3jnP6AbM0

Admittedly, a cobbled-together CNC laser engraver/cutter built from old CD/DVD drives and an Arduino or Raspberry Pi might help 

Kind regards, Arnold

Hi Bob

What kind of projects plans are you interested in getting?  Something simple, like making bits of tooling, or bigger projects like engines?

Can't have people just sweeping up 

Kind regards, Arnold

Nice going Dave 

It's difficult to pick a favourite engine from my collection, but Fancy is definitely one of the top three, and I always enjoy seeing another one made!

Kind regards, Arnold

Good going Tony - persistence pays 

Not many parts left to make from what I can see.

Kind regards, Arnold

Thanks Jo, Ian & Don 

Ian, I forgot about the lighter wheel knurl - and I have a whole bunch of old Bics lying around...  Actually I must just pull my finger from "in there" and make a proper knurl; I've got loads of silver steel lying around and it shouldn't take too long to make one.

Today I made the cranking handle from some aluminium and the counter weight from stainless steel, as I wanted the counterweight to "stay out of the way" in use:


Next followed some finishing touches.  The front of the bracket was rounded over at the bottom, the feedscrew was trimmed to proper length on both sides, and all corners were generously knocked off with a file.  I stamped numbers on the micrometer dial as well.  Then I took a family photo of all the parts:


Everything was assembled again, and adjusted with a bit more care than last week - backlash is down to 0.01mm which I'm pretty pleased about; the feed nut isn't even split yet for adjustment; I'll do that at a later point when things have loosened up again:


Some turning tests followed, and I'm very pleased with the results.  The direct readings works a charm; I arbitrarily turned a bit of rod down, measured it at 10.27mm with a micrometer, dialled in 0.27mm, turned and measured 10.00mm on the dot with the mic.  The slightly lower feedrate compared to the standard feedscrew works a treat for parting off as well; a couple of parting tests gave a nice "feel" and extremely good results.

Now it's time to clean and service the machines for a change, and pick an engine project - it's a long time since I built an engine...

Kind regards, Arnold

A bit late - Nicely done Colin 

Looking forward to the Minnie.

Kind regards, Arnold

Thanks Don.  It actually went quicker than I'd thought it would - Indexing off the gear train is a lot quicker than using the dividing head, and there's no need to count turns 

A bit more done this weekend.  On Saturday I started off with the handwheel from some 60mm aluminium stock.  The cut end was very carefully faced with extremely light cuts at low speed to "just" get a flat face , as it was sticking out of the chuck quite a bit.  Then center-driled to get support for the tailstock center, and after that I could lay into it and get some turning and knurling done.  The curvy bit was step-turned and smoothed out with a file, and the boss on the face matches up closely with the recess turned into the micrometer collar earlier on:


It was sawn off from the parent stock, chucked in reverse, and faced.  A 16mm diameter recess was bored half-way through it, and the rest was drilled and tapped M6 to match the threads on the feedscrew.  The recess is big enough to allow an M6 nut to be screwed into the recess and tightened with a socket to lock the handwheel to the the feedscrew.  A quick assembly check - the vernier collar isn't mounted yet:


Today I did a bit more - though there's not many photos to show it.  The handwheel was drilled and tapped for a handle and counterweight.  The vernier collar's mounting holes were drilled and tapped.  I scrounged a nice stiff 12mm OD, 7mm ID spring from a "retired" multi-function printer, and shortened that by trial and error to provide a firm-but-easy-to-adjust pressure on the micrometer collar with things assembled.  I still need to finish the crank handle and counterweight, and there's a fair amount of finishing and cosmetic work left to do, but things were finally at a point where I could assemble the lot on the lathe and test it:


- Everything works smoothly and feels really good when operated.  I mounted my trusted 0.01mm Mutitoyo dial indicator and checked things out:

Total backlash, without any "fine" adjustment is down to less than 0.05mm - that's good with me.

Kind regards, Arnold

Thanks Jo - maybe I should mill a pocket on the top to make sure tools stay there when put down...  Seeing as I've used the front end of the cross slide for keeping drills lying around, I'm sure the flat top will see a lot of use 

Nearly a month without shop time...  At least a bit today.

I had a bit of time to think things though, and decided to use a "Metal Removed" scale, and doing some quick calculations, decided on a 0.05mm scale.  On a 38mm diameter dial - for which I have stock, that leaves divisions that's nicely spaced as well - at about 1.5mm apart for the 80 divisions needed on the dial.  Making a vernier collar to take things down to 0.01mm based on the dial's divisions is also easy; that runs to "100 divisions" at 1.2mm spacing, so also easy to make and see.

First off, the vernier collar - just some 40mm aluminium bar faced and turned down to 38mm OD:


Then I set up the lathe's banjo with a 60 tooth gear on the spindle side, driving a 30 tooth gear coupled to a 50 tooth gear to use as the index, along with the crude detente I made for a similar job at Tel's suggestion a while ago.  The 50 tooth gear even still has the permanent marker markings I made on it  :


I used a very sharp 60^o threading toolbit mounted side-ways to cut the markings about 0.1mm deep.  In theory, only 5 vernier markings is needed for the layout I chose, but I cut 10, with a long marking in the middle for 0.  When I made the scale for my rotary table, I used a similar layout, and have found it extremely useful for working on either an in-feed or out-feed:


Next I bored out a hole in the center - this is to clear all the working bits of the cross slide once mounted to the cross slide bracket.  The engravings also got a tidy-up with a bit of sandpaper and then Scotch-Brite to dull the finish a bit - in use, I don't want the scale to be shiny, as that makes it more difficult to read under the shop lights:


The lot was taken to the mill, mounted on the rotary table (convenient to mount the lathe chuck in my shop), the center of the hole located, and the table rotated to get the zero-line in line with the mill spindle with the X axis at zero and using a small drill to match up to the line.  Just me being lazy to get the workpiece oriented so that the mounting holes would come out fairly square to the zero line on the collar. Mounting holes drilled at 3mm - after this they were counter bored to 6mm to allow room for M3 cap screws to go below the surface:

The vernier collar was then parted off. 

The same bit of stock was then turned down to 38mm to about 6mm away from the chuck, and the rest toward the chuck just skimmed to clean it up.  Then a 40 tooth gear was marked with permanent marker to show the gaps where the 0/1/2/3 mm marks must go (the 0/2 and 1/3 are in the same location) - those have the extra dots on the gear hub, and the 0.5mm spacings are the two without markings on the gear hub.  The 0.1mm markings would come at every second gap in the gear starting from the One markings, and the 0.5mm markings from every second gap starting from the gap just after the One markings:


Actually, it's easier to do than describe - the 40 tooth marked gear was mounted in place of the 50 tooth one shown earlier, the detente adjusted, and the motions followed.  All done here:

The 0.5mm marks are a tad too short  - that's what one get by just eyeballing things.

The center of the workpiece was bored to 12mm about 15mm deep - this is for space for the pressure spring I want to use- and then drilled right through 6mm to allow for the feedscrew threads.  The index markings was cleaned up, and excess stock sawn off on the bandsaw.  Then it was mounted back on the lathe - this time using the chuck's outside jaws and a bit of aluminium can strip for protection.  There's two reasons I used the outside jaws: 1.  The "curve" on the inside jaws when chucking up like this is bigger than the workpiece, and together with the can strip, it provides good clamping force without damaging the markings.  2.  My three jaw chuck's outside jaws are pretty accurate across the clamping range.  If that wasn't the case, I'd have used the 4-jaw and chucked up the workpiece in that.  The back of the workpiece was then turned down to get rid of the bandsaw markings, and then bored to make a friction surface / locating ring that will match up to the actual handwheel later on:


My scissor knurling tool cannot get close enough to the chuck to use in any of the previous operations, so I mounted the workpiece on a home-made expanding mandrel to get the needed distance and gave it a bit of a knurl with 0.5mm diamond knurls.  I'm well aware that no medals will be won for this job - according to "judging criteria" I've seen, it's too fine a knurl for this size of job, and the tops are not "well-formed and sharp"...  A straight knurl is actually what's required for this application, but I've not gotten around to get a wheel for that, and didn't feel like making one.  I could just have indexed a straight knurl similar to the scale engravings, but that's just too much work for a lazy sod like me.  This fine knurl, only partially done to depth is easy on the fingers in use, but gives a good grip even with oily fingers, so it's practical in use, and completely adequate for my needs:


For today's work, I ended up with this lot - not much, but extremely gratifying:


Kind regards, Arnold

Thanks for checkin in Don, Achim & Stuart.

Don, I can also only give you an amateur answer...

There's no need for adjustment of the steady while doing threading passes.  I only had to adjust it for the three cuts I used initially to turn the workpiece down to the thread's outer diameter.

There was very little metal raised along the cut line.  There's a couple of factors I found in the short time I've been cutting threads that seem to help making raised metal a non-issue (I did have problems with this when I started cutting threads):
a) I only use HSS or hardened silver steel/drill rod for my thread cutting tools, and always hone them to a very sharp edge - good enough to cut paper (or skin!) cleanly.
b) I'm never in a hurry when cutting threads - many passes with a small in-feed for each pass is the order of the day.  The biggest in-feed on the stainless leadscrew was 10 thou per pass, decreasing to about 5 thou per pass when about a third of the way and 1 thou per pass when I neared core diameter.  This light infeed prevents breaking the edges on the cutting tool, and also prevents "tearing" and raising of metal.  In the case of silver steel tools, it also prevents too much heat from building up and ruining the cutting edges.

For the tiny edges that do get raised - more like burrs - the fingers on the steady just push them flat again, and if the burrs are slightly harder to push down, they score tiny grooves in the brass fingers of the steady.  These are not much of a concern though; as the steady is following along at the thread pitch, the un-scored sections of the fingers keep the steady applying stability on the sections of the workpiece that matter.  In this case where I did a trapezoidal / acme thread, that's fine, as the tops of the threads remain with a flat section.  Cutting long lengths of V shaped threads might be a different mater, as once the crests of the thread start getting sharp, the burrs might cut away the "supporting" section, and some adjustment might be needed.  I haven't done that yet, so it's just speculation.

I just take a file to the ends of the steady fingers when the tips start looking too rough.  That's not much of a loss; with the low usage on a following steady it's not needed often.  In the case of my own particular unit, the fingers are still the original ones I made from brass house numbers I got at a garage sale

Quote from: StuartThose Zamak nuts have been a wear problem for a long time, they are now hardened steel and can only be got from the current supplier as a screw and nut part.

Yes...  And they're a bit dear at the current exchange rates!
Shrek very rarely "bites" me to the point of drawing blood - we have a great mutual trust relationship.


Eventually a bit of shop time today, and on to the mounting flange for the feed-nut.  A bit of 25x5mm flat bar was used for this.  The first run was a booboo - I forgot to add the necessary offset to the Y axis for the holes, and only after getting ready to drill the 12.5mm hole for the nut to fit in there realized this.  One of the dangers of not marking out, and just using the DRO scales to do things by...  I re-set things and did it a second time:


After cutting the piece from the stock and a bit of grinding and filing on it (nothing accurate - just by eye), I did a quick test to see if things fit as expected:

Not pretty, but it'll do.

This was also the first chance I had at checking everything made so far in terms of proper alignment.  All my measurements for making the parts was done in an inaccurate way; measuring things across bolt and screw threads, and so on.  So I mounted the whole kit and caboodle to see.  The holes on the nut flange were a bit out, and I had to take a 3mm needle file to them to "move" them a bit, but other than that things were good to go without any issues:


To assemble the feed-nut, a quick silver soldering job was needed.  One big concern for this was adequate solder penetration for the job; I wanted a good, but small fillet on either side, with solder only applied from one side.  I made about 20 score marks around the flange part's hole with a triangular needle file to help with this.  The step on the bronze section was used to keep things square, and some whet flux applied, with excess wiped off.  It was then left for about 20 minutes to let some of the water from the flux evaporate.  I _think_ this also allows the flux to start doing it's job; I've had much better results when silver soldering if I leave things to stand for a while all fluxed up.  Then I heated up the lot to red-hot, playing more heat on the bottom than the top.  When the flux went completely melted and clear and started swirling around the joint, I gave the upper edge of the joint furthest from the flame a good feed of silver solder rod - about 3mm worth of the 1.6mm diameter rod I used, while concentrating most of the heat at the bottom of the workpiece.  Here all done, but still dull-red hot:


The solder did wick through properly, and while the joint is not quite as pretty as I'd have liked, it does appear to be sound.  A wee bit of cleaning later, this is what I had:


There's some more work to be done on the nut, but at that point, I couldn't resist a thorough check of things.  With everything once again mounted on the lathe, I can easily spin the cross slide back and forth by just turning the 6mm threaded part of the feed screw between two fingers.  There's a tiny bit of backlash I can detect manually.  The movement feels slightly rough when feeding out, but that should wear in. On the in-feed, it's as smooth as silk, and the nut still needs a bit of a split and backlash adjustment added.

Checking through my stock, I could not find something suitable for making the hand wheel and scales from; some shopping is required...

Kind regards, Arnold

Thanks for looking in Eric, both Daves and Stuart.

Like Dave mentioned, cutting the internal threads could be a bit of a challenge.  I could go for a similar boring bar - for this job it can be bigger than Dave's, or a bit of silver steel (drill rod) shaped and hardened, or even make a couple of successive taps...  For this job, the taps are overkill, so probably not.  The silver steel won out, as it's the easiest.

Stuart, thanks - I think you have a Super 7? - the commercial dials differ a bit...
You raise a valid point that I've been thinking about quite a bit; do I want a "direct in-feed" reading on the new dial, or a "what this cut will take away on workpiece diameter" reading...  On the ML7, I've gotten used to the direct reading - dialing in 10 thou would take off 20 thou on the workpiece diameter - and I've gotten used to mentally converting to "close to metric" on-the-fly for that - an in-feed of 20 thou would take of approximately 1mm diameter on a workpiece.  My small (cheap & cheerful Asian) lathe works on metal removed - but it's so-called "micrometer" dial is extremely coarsely marked out - 0.1mm steps.  Currently I'm tempted to go for the diameter reading, and a scale of 0.02mm/division.  Or even a vernier scale, but usually when I get down to a thou, it's time to break out the micrometer for accuracy...

Golly...  More than a month since I made any progress...  My work schedule has been hectic, and to top it, a bout of 'flu and some very cold weather kept me from the shop.  It's cold when Shrek insists on crawling inside my jacket:


On to some shop time I found this afternoon.  The feedscrew-in-making was mounted in the collet chuck; I love the collet chuck for this type of job, as it can grip onto threads without damaging them.  The unmachined side was turned down to leave a shoulder with enough room on it for some spanner flats, a step with a close fit (0.795mm) to the bearings (8mm ID), 0.5mm shorter than the overall distance needed between the outsides of the bearings, and the rest to 6mm. The 6mm section is still much too long; that will get shortened later:

After this, I used a tailstock die holder to thread the 6mm section to M6.

On to the feed nut.  To save on material, I'll make it from two parts, silver soldered together.  The "nut" section is just a cylinder turned from phosphor bronze - with a step on it to use as a guide to silver solder a mounting flange to it later:

The "slug" was just parted off close to the chuck after this.

As already mentioned, the silver steel won this round to make the internal threading tool from.  The nut has an inside diameter of 7mm, with a thread depth of 1.25mm leaving the total ID of the thread root at 9.5mm.  6mm silver steel would be _just_ too thick, so I used 5mm silver steel, and gave it a bash with a hammer in the big vise to form a head on it:

Unhardened silver steel is actually quite malleable, but it does work-harden quickly, so just a good couple of thuds with a big hammer works better than trying to tap it over with a small hammer.  Tongue in cheek: this is one mechanical job Jeremy Clarkson might just be able to do...

Off to the bench grinder with that lot, and an initial form ground out with excess metal removed.  I couldn't care one iota about the job getting a bit hot while grinding; it got a lot more heat later on:


To get the angles, I used the original cutting bit as reference, grinding in the needed clearance angles on the bottom of the workpiece as well - not visible in this photo:


Off to the fire-bricks and the plumber's torch with a can of water handy.  I mentioned earlier that I didn't care too much with the grinding process if things got a bit hot.  This is where the real heat sets in.  IMHO, there's no need to harden the entire length of the toolbit; just the cutting tip section at the front needs to be hardened.  So I heated that to an orange colour, kept it there for about a minute, and dunked in the water:


Then it was off to the oilstone to give it a final sharpening - partly done here:


With the "slug" parted off earlier mounted in the collet chuck, I drilled it to 7mm, then using the tool made above, bored it to 7.5mm, using two passes at a very slow feed rate.
On to screw-cutting.  The first five passes was done quite coarsely at about 5 thou each, and that was followed by many (I didn't count, but most likely about 10 to 15) of finer ones.  About half-way through the threading process, I had this:

Not much to look at...

I ended up with a nut - the last checks were done by attempting to thread, and then threading the feedscrew into it.  I think I got things slightly on the loose side at the end, but the nut-in-making came out OK, though it's difficult to take a photo of the internal thread:


The nut easily screws along the threads on the feedscrew, and I left some room for rudimentary anti-backlash control on it.  Hopefully I can get back to things next weekend - for now, things were left at this:


Kind regards, Arnold

Coming along really nicely Colin  - the bore finish you got is excellent.

I know you just loosely assembled the engines in the second to last photo, but noticed that you have everything wrong-way around relative to the column...  Just thought I'd point that out - when it comes time to finish drilling the port hole into the bearing and final assembly, having things wrong way around will lead to a bit of heartache.  I've found that it's very good practice to always put things together in the intended positions when playing around even with loose assemblies or mock-ups, as one tends to "learn" the positions of parts rather quickly, and learning them the wrong way around can be a bit of a problem 

Like Bill said, counting turns is a good thing!

Just take some care when drilling the mounting holes - if you have the option of using a depth-stop on your mill, that's a great way to prevent going too far, but once again, the machine dials are your friend.

Looking forward to your next update 

Kind regards, Arnold

Thanks Bill 

The tapatalk was upgraded last weekend, but the upgrade changes some file permissions on the server which I forgot to check.  Apologies for the inconvenience and sluggish response everyone.

Kind regards, Arnold

Hi All

Can someone please test TapaTalk again and post here whether it now works or not?
I don't currently have a device to test TapaTalk with...

Kind regards, Arnold

Thanks Stuart.  I did consider that modification; I can get the bearings here.  I'm not keen on modifying any of the lathe's original bits though - any slip-up could be very costly, so I'd rather make new parts (and have fun while doing that) and keep the originals around in case my own parts don't work well.  Guess it comes from working in IT - I love to have a backup  .

Dave, thanks Buddy.  Follower rest: check - absolutely essential for this job.  Offset compound method - you score one more; I've used that in the past on coarser threads 
For those across the pond from Dave, a follower rest is also known as a "traveling steady".

Didn't have as much shop time as I'd like, but some progress was made.  The feed screw has to be as precise as I can possibly make it - and very smooth, so it was a bit of a challenge.  As already mentioned, for this job a traveling steady is essential, but one have to think about it's placement relative to the cutting bit.  If it's closer to the chuck than the tip of the cutting bit, and one cuts from right to left, on the first pass things are OK, but on subsequent cuts, the steady has to be adjusted, and will reach a point close to where the last cut stopped where there's a step in the stock that it cannot pass.  If multiple passes are needed to get the stock to size, one will be left with a succession of steps close to the lathe chuck.  This I didn't want - I wanted a clean edge there for this workpiece.  The other option was to have the steady lag the cutting edge by about 0.5mm.  For this to work, a short section has to be cut at each pass at the tailstock end with center support, the steady fingers to be adjusted to the new diameter, and then a complete pass taken toward the headstock, where the toolbit can then work up to a nice shoulder.  This is the option I went for.

I cut a bit of 10mm diameter stainless rod about 280mm long from stock, mounted that in the collet chuck on the lathe, and faced and center-drilled one end.  Then I extended it about 220mm from the chuck, with tailstock support from my small revolving center.  A section about 20mm long was turned down to 6.5mm (the root diameter of the thread I had to cut) - this left room enough to get the steady off the work, dial in a new cut, start it off for about 10mm, stop, adjust the steady on the new diameter, and continue the entire cut to the shoulder at the headstock end.  I turned the 10mm stainless down to an accurate 9mm in three steps - one rough to 9.2mm, a fairly fine pass at 9.1 mm, and a very slow and steady finishing pass on 9mm:


Next I sharpened up a HSS toolbit to a 30^o tip, taking care to keep a 15^o reference to one side of the toolbit.  As I'm a lazy rotter, this reference makes setting up the toolbit a breeze with just the aid of a square later.  Seeing as I'm cutting a metric lead screw thread, I might as well use the metric 30^o "trapezoidal" form rather than the 29^o ACME form (just in case anybody was wondering).  Putting some things I've learned along the way into practice (unbelievable as learning anything may seem  ), I then honed the toolbit on an oilstone - starting on the coarse side till the grinding marks were gone at the tip cutting edges, then on the fine side of the stone till all edges were sharp enough to cut paper cleanly.  With the toolbit's tip being dead sharp at this point, I stoned a flat on the tip - otherwise it will just break off on the first cut.  This flat also helps with the "root" of the thread, which is flat.
One of the most difficult things I've found to do in the shop is accurately measuring the width of the tip of a tool - the best option I've come up is "guesstimating" it using the 0.5mm scale bit part of a ruler with a magnifying glass - here my guesstimate is about 0.3mm:

(That was a darn tricky photo to take!)

Off to threading, and after two passes I noticed something wrong...  On both sides of the cutting groove, there was metal displacement rather than cutting happening, especially on the left hand side of the groove.  On the next photo this is clearly evident.  So I stopped and thought about things a bit; displacing rather than cutting stainless steel will work-harden it very quickly...  The simple fact is that I didn't grind enough clearance on either side of the V of the cutting tool.  With quite a large helix angle for this thread, that meant more metal displacement (or rubbing) on the left-hand side of the groove.  With things close to the start, I just took the QCTP holder off, carefully ground more clearance on the toolbit, and carried on:


After many passes, feeding in at 5 thou (0.127mm) per pass initially, and finishing off at 1 thou per pass (0.025mm), as well as using the compound as Dave suggested to add another 0.3mm width, and a quick run-over with a scotch-brite pad to tone down sharp edges, I ended up with this:


A close-up:


Must admit, I'm rather pleased with the result 

Kind regards, Arnold

It's been quite a while since I posted up on any of my antics in the workshop.  My work life has been absolutely bonkers for the last five months, and I just did not have time or energy left to do much.  Some other issues also detracted - one of which is the arrival of a couple of Raspberry Pi computers and some local people wanting me to showcase what can be done with those.  So some time went into getting intimately familiar with these little beasties - both on electronics and programming level.  I'll post more on that in due course.  This thread is somewhat related to the Pi projects though - there's some machining required, and all is not too well with my ML7 lathe.

My Myford ML7 is about 45 years old, and I'm the third owner as far as I'm aware.  She's starting to show some signs of wear - most notably the classical "loose carriage toward the headstock" bit, and, more alarmingly, a LOT of backlash on the cross slide.  There's no doubt a lot of wear on the lathe is down to me - the poor girl landed up with a complete newbie and I had to, and still, learn a lot of lessons along the way at the lathe's expense...

Of late, the backlash on the cross slide has become a real pain.  A quick check a while ago showed 1mm (40 thou) of backlash on the cross slide.  Part of that is due to wear on the "mazak" dial and the mechanisms behind it, but the feed nut is not too great either - at least 0.5mm (20 thou) of lash is in there after I took some measurements.

The easiest option to rectify at least part of the problem is to just replace the feed nut.  After getting pricing online for a new one, I picked myself up off the floor.  I can't remember if I fainted or went into a hysterical laughter session.  47 GbP for a feed nut, converted to a very weak Namibian Dollar and with the transport pricing still to be added is just ludicrous IMHO.

Next option - make one; it's not really a complicated part.  So I disassembled the cross slide's feed mechanisms, and while testing the nut by hand on the feed screw, noticed that it's slightly better fitting on the parts of the feed screw that has seen less use, so the feed screw itself is not that hot anymore either.

Coupled with the wear on the back of the dial and on the mounting bracket, everything is just plain worn.  Some bits are easy to fix with a new bush or spacer here and there, but by this time, I'd already decided to go the whole hog and re-make things to my taste.
A new 2mm pitch metric feed screw and -nut with some form of backlash adjustment, bearings instead of bushes, and while I'm at it, a resettable micrometer dial; one feature that I've wished for on many occasions.

Material selection isn't great; for the most part I'll use whatever I have lying around and the bearings were scrounged from an old printer.  I did splurge on a proper bit of phosphor bronze for the feed nut.  For the feed screw, I'll use stainless steel; it was either that or silver steel.  I'll see how that works out.

I scratched around in my own documentation and asked Google for the dimensions for the mounting bracket for the cross slide, but couldn't find it anywhere, so I set out taking the measurements off the cross slide itself.  This is one example where Marv's suggestion a couple of years ago to add a threaded hole to the height gauge arm came into it's own.  A 4mm cap screw conveniently allowed me to zero on the top of the cross slide, and then take the measurements below this:

To overcome the looseness in the feed nut, I carefully wrapped a couple of even layers of PTFE plumber's tape around the feed screw to tighten it up and keep it as central and straight as possible to take the measurement.

After taking more measurements, I had a bit of C-o-C with the needed dimensions - not super-accurate, but good enough:


A block of 80x60x20mm aluminium that I got from the offcut bin at one of the local metal suppliers a while ago volunteered to make up the beginnings of the new bracket after some marking out:


While it was a solid block, I hollowed out the underside for clearance for the top of the dovetail on the cross slide:


Next, some cut-outs on the sides to accommodate the mounting screws:


Finally the mounting holes - both about 0.5mm over-size.  This isn't sloppy work, but intentional, as the play in the holes are used to align the bracket with feed screw properly to the feed nut - the original mounting is built like this as well:


A quick trial-fit:


Checking from below that the dovetail clearance is good:


The bracket at that stage:


For the vertical bit of the bracket, I used some 60x10mm hot-rolled mild steel.  One long edge was used as a reference to mark the center point of the feed screw, and a hole was drilled under size and carefully bored to 16mm, and then a circle faced away on the flat side to remove the crud and make things nice and parallel.  This side will be mounted to the aluminium part made earlier:


How things will go together later, together with the two bearings I scrounged to fit in the bored hole:


To mount the plate to the aluminium part, I was torn between using two M5 or M6 screws or four M4 screws.  I didn't want the screw heads to encroach too much on the center of the bit, as some more things will need mounting there, and, should the bearings I have not work out, that there is enough space to install slightly larger thrust bearings.  So I opted for the four M4 screws - they are nicely out of the way, but will give about the same amount of strength as using two M6 screws.  So I worked out a bolt pattern that wouldn't be too much of an eye-sore, and drilled and tapped for M4.  A bit of clearance for the feed screw and retaining ring on the bearing was also added while I had everything clocked up on the mill:


The steel plate got the same treatment, except the holes were drilled for M4 clearance, and counter bored 7mm 4mm deep to recess standard M4 cap screws.  This is the front face showing - the turned back face is at the bottom, as that must mate very flatly with the aluminium bracket:


Assembled thus far:




There's a LOT of rounding over to be done on these parts - in use, they will be in close proximity to a lot of movement of hands and knuckles, and at this stage skin will get taken off.  There's more machining to do first though, so neatening it up will have to wait.  Next up is the new feed screw - which will be quite an interesting job, as I've not screw-cut this coarse a thread on stainless steel before...

Kind regards, Arnold