Member postings for Colin LLoyd

Unfortunately, I only began to transcribe his war memoirs after he died. Or else I would have asked him the same question - along with many others that arose during my attempt to read his difficult hand-writing.

My father also noted that this exercise separated those "engineers" with only machine training (e.g. lathes and milling machines) from those that could turn their hand to anything. He didn't know what the things were until he examined one of the early locomotives of the Canadian National Railway at Halifax, Nova Scotia after his ship's (HMS Malaya) crossing of the Atlantic with gold bullion on board destined for safe keeping in Fort Knox (27th Feb - 4th Mar. 1940).

Speaking of lathes - he also told a tale that illustrates the British attitude to war. During one of the many Mediterranean actions that HMS Malaya was involved in, my father went ashore and found an abandoned German lathe. He got the rest of the engine room artificers to row to shore, load the lathe onto the boat with some difficulty and row back to the ship. With difficulty, the lathe was winched up the side of the battleship where the ship's Captain was waiting (hidden). As my father put his head over the side, the Captain said "Lloyd - is that lathe yours? The reply was "No". "Lloyd - you will return it to where you found it". So different from the Nazi attitude to personal items.

At the beginning of the 2nd WW, my father, a young and newly qualified Toolmaker & Fitter left Archdale's Engineering Works in Worcester to join the Royal Navy. The entrance practical exam at Portsmouth Dockyard for Engine Room Artificers included making a Block, Strap, Gib and Cotter in 42 hrs using only a hammer, chisel and 3 files using no machinery except for drilling holes. I've attached a photo I've found of this type of test piece.

My questions are: what are, or were, Block, Strap, Gib and Cotter and what were they used for. And why were they used to test entrants to the Royal Navy at the start of the 2nd World War?

Just my pennyworth - and it may be a repeat of something I've said elsewhere.

For CAD design I use OpenSCAD as a) it is free and b) it is more geared to engineering use as opposed to something like Inkscape or Blender which are more artistic based. In OpenSCAD you define your object through mathematical interactions of primary shapes such as circles, cylinders, cubes, plates etc. These are combined through logical processes such as difference, intersection, etc. Because the items are created using engineering units, items such as nylon gear wheels, journal bearings etc. can be created. These can then be exported through a number of different protocols including STL for import into your preferred printing program which converts these STL (for example) files to the G-code necessary to move and operate the 3D printer. Examples are easily seen - just search for OpennSCAD Engineering examples. And for anyone who fears having to write the program to create these items, OpenSCAD has extensive libraries where programs can be downloaded, studied, copied or modified - so easing the design stage and the learning curve.

For Printing I use Repetier-Host which includes several different conversion sub-programs for converting G-code to 3D printing output such as Cura and Slic3r. Again, Repetier-Host is free and available for LINUX, Windows-aargh, and Mac-for the rich people.

Having nearly completed my Mill stops "a la Cox" with slight modifications to accommodate my linear digital scale units - I'm now in the DIY market for stops for my Amadeal CJ18A mini-lathe - which is a Real Bull variant of the common Chinese lathe. Can someone point me in the direction of an efficient design within (or without) the pages of MEW - that I can go back to (with my new digital subscription). The Lathe bed has a truncated "V" as the locating position - but that is just a fabrication issue - it's the ease of use and efficiency design that I'm looking for - how easy to re-position coupled with position security. And I'm guessing - looking at my lathe - that a stop that turns off the leadscrew feed is not practical or even sensible. So, as a beginner, I'm again guessing that lathe stops are only used with manual feed of the tool.

Thanks for all the good methods. Generally it comes down to whether you are fabricating one item or there is the possibility of having to make many. Belt and rotary sanders I'm quite happy with and have table setups using DIY belt and rotary sanders running with the sanding surfaces vertical when creating the complex curves that my electric guitar body builds require - but the time and effort to set these up is not worth it for just a couple of items. Similarly KWIL's idea is good - except that I haven't got a ball turning attachment and currently see no reason to either get one or make one - this may change.

I know how to do the simple hacksaw, file and finish methods (working in remote places in Africa, South America and Siberia taught me to use whatever was available to construct items) - but for my machine learning phase - I will see how I get on with the rotary table method - with just a test piece to see what the problems and solutions are.

MIchael G - silly me - should have said "stationary mill table" rather than "stationary end mill" - bit of a Railway Observer Einstein Relativity moment.

Mike Cox - thanks for assuring me that there is still room for good old physical workshop practices.

This question relates to Mike Cox's "Enhancements to the Sieg Super X1L Mill in November 2107 (No 261) edition of MEW and specifically to Fig. 1 of the part dealing with Axes stop bars.

Don't laugh - but I just wondered how you would machine the rounded end of this piece. Before I got a lathe and a milling machine, I would just have hacksawed bits off and then again the peaks of what was left - in a sort of workshop version of Pythagoras' approach to the value of Pi - and then use a file and emery paper to get the final curve and finish.

I could use a rotary table with the workpiece bolted through the hole (and raised above the rotary table by a thick washer) and just rotate the workpiece around a stationary end mill. But I'm open to better or more correct (in an engineering sense) methods.

I've got one of these to work with a 10mm dia x 10mm long high strength neodymium magnet (specification says these can lift 3.5kg) fitted to the spindle locking ring on top of the mill. As Howi says, even at 2500 rpm on my milling machine - the magnet doesn't move and I can have the sensor nearly 10 mm away from the magnet without loss of signal.

But I suspect from a cursory look at the subject of the Hall Effect and general magnetic field effects that a strong narrow vertical magnet would provide better precision for the sensor response - in that the effect would be a quick "on-off" as apposed to the "rise & fall" that the wide magnet would give. And again - I'm guessing - at high RPMs, with the magnet width being a substantial part of the overall circumference, the sensor might not have returned to it's "Off" state before the next revolution brings the magnet into its sphere of influence.

Perhaps someone with a better grasp of Hall effect proximity sensors than me could put me straight.

Just to add some facts to this part of the thread dealing with thrust bearings for the Z-axis screw rods. The Stepper motor - filament supply unit - heater unit - finned cooler - fan - and support bracket weighs 490g, the 2 y-axis traverse rods, I estimate, weigh together 990g and the remaining support structures for the longitudinal bearings and z-axis movement probably another 500g for a total of about 2 kg. Which seems to agree with spring balance measurements performed on one of the threaded rods. A hook was super-glued to the top surface of the rod and initially a 500g and then a 1000g spring balance was used to see what was required to raise the rod by a very small distance - discernible but much less than 1mm. This resistance relates to both the total weight of the z-axis structure and the deforming resistance of the grooved flexible coupling at the base of the rod. No movement was seen for 1000g. Using a 0-32kg (resolution spring balance seemed to indicate about a 1.5 kg lift for a discernible movement of the rod before the superglue attachment failed. What this seems to tell me is that the flexible coupling has little flexibility.

But I am still in two minds whether to add thrust bearings to the top of the rods - for) taking the load off the flexible bearings and against) I've seen no fault in my prints that might be explained by lack of flexibility in the couplings. As an alternative, removing the flexible couplings and injecting silicone rubber into the grooves may provide the mechanical support in the couplings while still providing the flexibility.

The proof of the pudding in this first layer adhesion debate is what the end product looks like and what the creator is happy with. I've tried diluted PVA and cannot get a consistently thin layer - but that could be my clumsiness. Which is why I use the hairspray; like acrylic paint - it provides an incredibly thin consistent layer which then creates a mirror-like finish to the underside of the bottom layer. Bed temperature can also play a part in this adhesion problem as well as affecting how solid (in a structural sense) a foundation is provided for the rest of the print.

I always use the initial nozzle "clear-out" and testing circum-navigation of the 3D print in Repetier Host to judge whether I need to adjust anything for that important first layer. There's nothing more frustrating than finding, after hours of printing, that because of a poor first layer, the whole print has been compromised.

Rod: You quoted - That's one of the G code files form the SD card (the Z micro switch holder). Need to sort out the initial adhesion though:

I use Sainsbury's Basic HairSpray tp provide adhesion to glass plate. Give the glass plate a wipe with Iso-propyl Alcohol and then a light spray with the hairspray. Never had any problems with adhesion - more the opposite -usually need to get a scapel blade under the object to release it. Don't try and release the object until the glass plate has cooled. The other reason for poor adhesion is if the nozzle is too far away from the glass plate for the initial layer. I just go to the 4 corners of the glass plate and then do a Z-axis "Home" with a sheet of normal thin paper beneath the nozzle. If I can pull the sheet out with slight resistance - this seems to work, otherwise adjust the sprung edge bolts on the bed

Neil - can you post the photo of your version of the bearings for the z-axis threaded rods again. Thingiverse shows many different versions - none of which appears to solve the problem of relieving pressure on the flexible couplings at the base of these threaded rods. I now have the B8 bearings and was just going to drill and tap vertical holes in the top of the threaded rods, then position the bearing on top of the plastic support and then use a bolt and washer to draw the rods up slightly. This seems to me to be the easiest solution to see whether the bearings make any difference rather than a disassemby of the top supports - although that's not a difficult job.

JasonB - please can we not use terms such as "jacking off" - but it is your method that I alluded to in my post above - glad to see I'm beginning to have the same ideas as the experts.

Actually I don't need to use long nuts. I have some scrap solid hardwood (from a previous project). I just need to make a thick shim that just fits behind the stud. Then unscrewing the nuts on the studs will give me about 5mm of "push" which is about the same length of the mandrel shoulder.

To John and Dave - thanks for that info. I will go with your suggestions - just needed to check what the experts did.

As the proud owner of a new 4 jaw chuck and 160mm faceplate for my Amadeal CJ18A mini-lathe - it looks like I might be changing chucks quite often. The 100mm chucks locate on a precise shoulder on the mandrel (very precise) so that it has been quite difficult to take the original supplied 3-jaw chuck off. I have used a thin flat-headed screwdriver and mallet to get between the chuck and mandrel shoulder surfaces - but that is hardly elegant and will, with time, begin to exhibit damage to the chuck and mandrel.

I was planning to mill 3 and 4 small flat-headed screwdriver-sized slots on the periphery of the 3-jaw and 4-jaw chucks so that I could use these to lever off the chucks.

Before I do that - is there some other way to easily remove the chucks. I thought about pairs of expanding triangle blocks that could be screwed together behind the chuck securing bolts and thus pushing the chuck bolts and attached chuck out in a controlled manner - but that seemed an over-engineered solution.

I noticed that Neil also added these thrust bearings but didn't notice much difference - and he has removed the picture showing them - so if I wanted to add these - how do you do it?

One thing I did find I had to adjust with the Geeetech Prusa i3 Printer and I suspect might need the same on similar clones is the "fit" of the Longitudinal bearings on the vertical z-axis. In contrast to the x-axis travel which moves the bed on four shorter longitudinal bearings and uses a toothed belt for movement, the z-axis bearings are 40mm long and any deviation from perfectly straight in the rods causes the bearings to stick especially as the vertical threaded transport rods allow very little accommodation in the bearings. The y-axis transport also has these long bearings but movement is accomplished, again, via a toothed belt and I suspect this more flexible transport has meant I don't have the same problem with the y-axis bearings.

I found that a simple fine emery paper sanding with the rods rotating in the lathe cured the problem. The threaded transport rods maintain the position that any slightly lower contact in the bearings might have compromised. And I see no problems in the resultant prints.

Sorry Iain - forgot to answer your first point. Yes - the dual head approach will allow the use of soluble support filament to be used. But support can also be achieved at the design stage - as I did for the LED supply box shown above with semicucular tops to the holes. But I could have designed the box with very thin "windows" within rectangular holest - which could then be just pressed or cut out afterwards.