Member postings for DC31k

Would you be able to point him to one that can do the _five_ divisions that he requires?

Your comment about 'easily made by machining' is rather chicken and egg in that you would need an indexer to make an indexer.

Three wire will not help if you do not know the form (vee-angle) of the thread you are cutting.

I can cut a Whitworth form thread of same nominal diameter and pitch as a UNC thread and make an identical three wire measurement on both.

Nor will three wire help you easily determine depth of cut (the title of this discussion). Without doing a lot of quite complicated maths, it will only tell you 'I need to go a bit further' or 'I have overshot the mark'.

It would be hard to sustain SOD's idea given that ER32 collets are available to buy in 1mm increments.

For interest, ER50 collets have a 2mm clamping range.

There are comprehensive diagrams of most thread forms on either Wikipedia or here:

https://www.gewinde-normen.de/en/index.html

They will tell you exactly the form of the root and tip.

If you have an HSS bit, grind it to a sharp point as suggested above. Measure the overall length if the bit. Grind or stone a small amount off the end. Re-measure the overall length. There is a relationship between the two measurements, the included angle of the bit and the length of the flat.

Exactly as suggested above, you do not need to see the flat you have ground, just establish a reliable technique using measuring instruments you already have to infer the length of the flat.

A little handwritten sign that says 'wait 30s before unplugging' is considerably cheaper, has no moving parts that might fail mechanically nor electrical parts that will emit smoke.

You could also locate the socket at ceiling level so by the time you have moved all the junk and fetched a ladder, the capacitors will have discharged.

It is possible to misinterpret your description above, so you might have to be more clear in the way you describe the connections.

In particular, you need to list clearly the relationship between the six wires.

I _think_ you are saying: red (A) and brown are connected; yellow (B) and black are connected; blue (C) and white are connected. However, my (or anyone else's) thinking is a particularly bad way to proceed. Certainty and clarity is required.

If this is the case, and you connect brown to terminal A (red), as you appear to be saying above, you are just connecting the two ends of the same coil together. No harm will come by doing so, but the motor will not work.

You have the delta triangle on the data plate. Each side of the triangle is one of the three coils. Each vertex of the triangle can be labelled sequentially A, B, C. So the 'start' of the first coil goes to 'A'. The 'end' of the first coil goes to 'B'. The start of the second coil goes to 'B'. The end of the second coil goes to 'C'. The start of the third coil goes to 'C'. The end of the third coil goes to 'A'. One each of the incoming phases goes to A, B, C.

Why?

The product to which he links is made by WEG, designed for industrial use. It will supply the size of motor listed in its specifcations.

You might have a point if he was considering buying a generic Chinese VFD, but that is not so here.

Is one of those gears doing the driving and the other being driven?

If so, measure the centreline distance of the shafts upon which they sit (when the backger is engaged). That number and the 68 total teeth in the train (or 34 teeth in the gap between shafts) will allow you to calculate pitch circle diameters

I agree on the DRO being over the top.

In some respects an internal thread is easier to pick up than an external one because the shank of an internal thread cutting bar is parallel to the axis of the thread. You just loosen the bar in its holder and slide it fore and aft, gently manipulating the cross slide outwards until the tool tip centres itself in the thread groove. This allows those who thread with an angled compound to continue to sleep at night.

Iscar make an external threading tool that presents the shank parallel to the thread axis, which allows the same method to be used for external threads. They call it type G, for gang tooled lathes. It is also available from overseas, with a suffix B.

https://www.iscar.com/eCatalog/Item.aspx?cat=3800002&fnum=357&mapp=TH&GFSTYP=M&srch=1

https://www.aliexpress.com/item/32924338528.html

...to purchase an inexpensive BA thread pitch gauge

https://www.ebay.co.uk/itm/352474288510

(from RDG but I could not find it on their website. Chronos also sell one. Maybe Proops as well).

If you lengthen the arm, you change the lever ratio between input (clevis in bottom of middle photo) and output.

That ratio change will persist along the operating rod and be seen by the other input to the drum.

It would be prudent to make sure that you have taken this into account before doing any modification.

Is there any reason the operating rod cannot have a joggle in it to snake its way past the obstruction? As long as you maintain the straight line (crow flies) distance between the two parts to which the rod is connected, the shape of the rod (providing it maintains sufficient compressive stiffness) does not matter too much.

Imagine a square of metal. Superglue it any old how to the end of a piece of bar. If each edge of the square in turn is kept parallel to the mill table below it, the square cut on the workpiece will be correct.

Keeping it parallel is fiddly (as the distance keeps changing) so it is easier to keep one edge of the square vertical using an engineer's square. Sure, the stock of the square will slide back and forth on the mill table each time you turn the part, but the blade will always be vertical.

It is a good technique to know. If you wanted a seven sided thing, you can just print one out on paper, glue it on the end of your stock and away you go.

A semi-finished version is available here:

https://d-gray-drafting-and-design.myshopify.com/products/indexing-plates-the-complete-kit

I have always liked the idea here:

https://www.youtube.com/watch?v=_qYdcolOhGQ

Edited By DC31k on 12/05/2023 13:10:12

A possibility is a rear toolpost and upside down tool. That could manage and direct the swarf in a way that is less labour intensive.

I wonder if multiple plunge cuts with a parting tool would remove material faster than multiple OD turning cuts.

With what you have drawn, how does it remain centralised on the hole initially?

---

As an aside, any method that makes the square hole other than as the first operation on the part has to consider the clocking of the square in relation to the rest of the part

MSC one here:

https://www.mscdirect.co.uk/DMB-66631B/SEARCH:CATEGORY/product.html

Alternative via Australia:

https://www.ebay.co.uk/itm/225149641607

Note that a 5mm square broach needs a pilot hole larger than 5mm diameter (eBay one says 5.2mm; DuMont's technical data will confirm the size they need). This may or may not be important to the part you are making.

At what stage does the square hole need to be made? I wonder if you could drill through 5mm, heat it up red hot and drive or (fly)press onto a piece of 5mm HSS, with the end tapered. Once you have the 5mm square hole, use it as a datum to make the rest of the part.

A proper blacksmith would not even drill the hole but would hot punch a through hole and then forge to the correct shape.

ISO turning tool codes are here:

https://www.cutwel.co.uk/blog/learn-the-turning-tool-iso-code-system

https://www.secotools.com/article/120340?language=en

You will see that the ninth field in the code is the cutting edge length. Any tool that takes the same shape insert (the second field of the code) with the same number in the ninth field will use the same size insert.

A screw thread is a linear device. I am puzzled by your use of a ratio in conjunction with its operation.

Its name, differential screw, describes its operation. 'Differential' implies 'difference', an additive or subtractive operation. It is not called a quotiential screw, implying a multiplicative or divisive operation.

If you do the maths, 1/30 differs by 1/32 by 1/480, so you are effectively making a 480 tpi thread on the adjuster.

You might ask Hemingway's engineering department what divisions you would need on a dial to convert that 480 tpi into a 1/10000" increment since 1/480 th of an inch is not an integer number of ten thousandths.

The closest integer is 21, and that is hard work to keep track of on a dial. If the dial happens to have 20 divisions, you will be roughly 5% in error every full turn.

I think you said in an earlier post that you had a Myford. Are you aware of any factory/manufacturer-produced drawing for the Myford lathe that shows this information? Sure, everybody knows what a Myford spindle nose looks like, but these are third party drawings not factory ones.

If you picked a random sample of lathe manufacturers*, how many of them would provide the drawing? For a standardised spindle (e.g L-taper or D1 camlock), the best you will get is a reference to the appropriate international standard, which you could then purchase from the standards body.

* my list, based on available online manuals: Schaublin, Colchester, Harrison, Monarch, Holbrook, Hardinge.

If such a drawing does appear, what will you do if it disagrees with your measurements?

If the drawing has any tolerances on it, what do you do if your register is outside those tolerances?

The piece of metal at the front of the lathe spindle has to be the ultimate arbiter of what you make. A piece of paper with lines and numbers on it will not change that.

An option is to make the register fairly loose, then with the securing bolts snug but not tight, tap the collet taper into the best concentricity you can measure. Then drill, ream and pin both parts and finally tighten the bolts.