Member postings for DC31k

https://www.model-engineer.co.uk/forums/postings.asp?th=45976

Can I make a suggestion? The resources available when you type your query into Google outnumber those here on this little site, good though it is, by several orders of magnitude.

Would you be able to copy and paste the following into Google?

bs0 dividing head manual

There are many sets of instructions, the US-Grizzly ones are normally good. If reading is not your thing, there are some good videos on YouTube.

It would be prudent to do an independent check on the numbers given in the division tables. There have been posts here that the ones supplied with the rotary table dividing plates are incorrect.

See post #39 here:

https://www.mig-welding.co.uk/forum/threads/d1-4-camlock-question.99683/page-2

That is very good. Is the connector providing the drive in lieu of a dog?

I have seen boring heads used horizontally and stationary in the tailstock for taper turning (saves offsetting and resetting the tailstock).

I think the same arrangement would work for between centres boring (i.e. it does not matter which end of the bar the boring head is). Is that correct?

The one Jason suggested is very nice.

If you search 'boring bar micrometer' and look at images, some of the results will give you some ideas. Any that include 'Van Norman' are not what you are looking for - it is the ones that look like a standard micrometer with the fixed anvil replaced with a vee block.

A vee-block with a micrometer barrel strapped perpendicular to the vertices of the two vees would work. Replace vee block with an old centre-square body and add micrometer.

Grind a notch in the cutting bit diametrically opposite the securing screw. Use a second screw with a tapered end engaging in the notch to advance the bit. Very limited range of movement with this though.

Use a normal micrometer across bit and bar after doing an initial cut and measure. If diameter of cutting bit is so large that micrometer anvil falls into the hole, use a gauge block to span the hole. This does mean you have to shorten the butt end of the cutting bit shown in the photo.

See also:

https://www.model-engineer.co.uk/forums/postings.asp?th=130528

https://bbs.homeshopmachinist.net/forum/general/37455-boring-bar#post721214

OK, found it. Search for Moore & Wright 462.

Edited By DC31k on 31/12/2020 10:38:52

The eBay link by peak4 is for a double-ended insert. This will not fit.

GTN inserts will not fit.

MGMN inserts will not fit as they are again double-ended.

Have a look at:

https://www.rdgtools.co.uk/acatalog/9PC-Turning---Boring---Parting-and-Grooving-Tool-Set-12mm-Shank-435344.html

I think the parting tool is the same. The photo is a bit small. Check the other shank sizes in the same range to see if anything more visible is shown.

If you can establish identity, the insert code is given on that page - it is not SCMT or WCMT. The two JCL15 references with I and E are, I guess, the internal and external screwcutting inserts. By a process of elimination, that leaves CK3.

Please ask someone else to critique this analysis before committing your cash.

See also: https://www.model-engineer.co.uk/forums/postings.asp?th=76422

Edited By DC31k on 30/12/2020 13:51:29

Can I mention a couple of ideas on the periphery of your request?

Have a look on the Dolphin site, particularly among the documentation for the milling
module. That describes a way to 'unwrap' the A-axis so the machine treats it as a linear
axis. With a bit of thinking and post-post-processing, that could work for a C-axis. I
think there is a program called G-code unwrapper that does the same thing.

Have a look on the madmodder site for some posts by Andrew Mawson. He has a Beaver CNC
lathe with C-axis capability. He wanted to mill a hexagon onto the end of round stock with
a live tool. His solution might give you some things to think about.

Have a look at the K-Flop controller. That is very configurable.

Have a good think about what you want to use the C-axis for. For instance, if you are
making a cam, the cut surface is just a series of (X, C) coordinate pairs with a bit of tool
positioning before and after.

It is not too difficult to use (spit!) Excel to generate the meat of the geometry and then
hand write the intro. and outro. Much easier in something like python, with the benefit
that python interfaces easily to dxf so some automation of input geometry and sanity
checking of output is possible (see also a program called CNC simulator)

If you want to mill a polygon on the end of stock, you just have to work out the equations
for one side, do a turn at the vertex and repeat for the other n-1 sides.

Hole patterns in the face is a simple index, peck and retract and repeat.

So my suggestion is to hand code the C-axis stuff. Once you reach the limits of this, then
go looking for a commercial solution.

I am not sure why the spindle zeroing before starting C-axis operations is causing you
distress. Say you are holding square stock and you want to notch it at the midpoint of
each side after turning the end of it round. Once the turning is finished, you switch to
C-mode, jog the axis so that the stock is in a known position (use a square off the lathe
bed, for example) and then zero that axis in the controller. Then load and run your C-axis
code. If you watch the Edge Precision YouTube videos, that is what he seems to do on a
multi-thousand dollar Mazak machine.

If you think about what you are asking, the machine would need to keep track of the
spindle position from the moment it is switched on and you would need to load the stock
into the chuck in a specific orientation. You do not set the DRO on your mill to 2mm
and then push the stock around on the table until the edge finder jumps. You clamp it down
anywhere, touch off and then type 2mm into the DRO.

On the spindle control, the only servo with which I am familiar is the Mitsubishi one. On
this, the motor has three modes of operation: position, speed and torque. For position
control, you feed it step and direction. For speed control, it needs a 0-10v analogue
signal. To switch between modes, you change the logic level of a couple of input pins on
the servo brainbox. This is easy to do in something like Mach3 and is an exact copy of
what it does for flood or mist coolant on/off.

So you need to be looking at the servo controller documents to see how it wants to be
mode-switched and then build something that will do that (possibly with a break out board
if it is something strange). The 270kHz for maximum spindle speed is likely to be a big
stumbling block. Mach3 will not do a lot more than 10% of this. As someone says above,
LinuxCNC is 110kHz. So look very hard to see if there is a way of controlling the spindle
speed by analogue methods. If not, you might have to build a dedicated circuit.

When you post them to him, how will you send him the hole in the middle of the insert, that he needs for securing it? The second 'T' in his TCMT is quite important...

The insert designation is using the ANSI (American) system. Inserts sold here use the ISO system.

There is a very useful chart here that has both languages on it:

http://www.carbidedepot.com/formulas-insert-d.htm

With inserts, the importance of the information conveyed decreases considerably the further to the right that you read. Some of what is shown on your cutter is likely a manufacturer-specific grade of carbide.

For buying replacements, start at the left and get the basic shape. Then, for your chosen manufacturer, look at their literature and choose one of their offerings to suit the material being cut.

What the sketch is depicting is true, but is it not mislabelled? Where it says 'initial setup', that should say 'final cut' or 'sparkout pass'. So you start with the wheel further from the pivot point than your final dimension and gradually advance until it hits a stop at the correct place. If you go beyond that point, as you say, you will end up with an undesirable result.

Conceptually, it is similar to using a vertical axis ball turner on a lathe. You turn the stock to the diameter of the ball, advance the cross slide until the pivot is under the workpiece centreline, then advance the cutter until it touches the OD of the stock. Note the cross slide reading or set a stop at that point. Then you can back off the cross slide and work inwards to your stop, sure that you will end up with the correct diameter ball.

I am having some difficulty understanding the above sentence.

When the thread being cut is a factor of the leadscrew pitch, a dial is not needed. An alternative way of saying that is "you can engage at any point the half nuts drop in".

If you choose to use a dial for these threads, you most certainly do not need to engage at the same point on the dial. You can engage at any point at all on the dial and at any different point on the next pass. Since the graduations on the dial are frequently less than the number of teeth on the gear, you can even engage at a point in between graduations with no ill effect. If the nut goes in, the threads will align.

Taking his 30t dial on a 3mm pitch screw as an example, Cleeve would call this a 90mm indicator. If you wait for the same point on the dial each time, you are effectively saying that all threads that are a factor of the leadscrew pitch have a minimum synchronisation distance of 90mm, whereas the true figure is 30 times lower than that at only 3mm. Say, for example, that the carriage is advancing at a rate of 1mm in 1 second. For the 'same point' method of working, you are waiting for 90 seconds whereas you only need to wait for 3 seconds. 87 seconds _each_ _pass_ are wasted.

Sorry, on re-reading above post, there is an error. For 2mm pitch on a 3mm leadscrew, the indicator IS required.

That is true but it is a case of marketing trumping engineering.

For a 3mm pitch leadscrew, assuming the necessary leadscrew gears are present, the following can be cut with no indicator at all:

0.2, 0.25, 0.3, 0.5, 0.6, 0.75, 1, 1.5, 2, 3.

The thread dial indicator is required only for 0.4, 1.25 and 2.5.

Any pitch that is a factor of the leadscrew pitch does not need an indicator.

As Thor says, the addition of a 28t gear (and dial) is the best 'bang for your buck' in adding missing pitches. Potentially, you could cut it on a blank sized for 30t then you do not need to adjust the meshing centre distance when changing gears. Alternatively, recut both gears on a 29t-size blank.

Edited By DC31k on 23/12/2020 14:45:07

Edited By DC31k on 23/12/2020 14:49:47

Look up 'Modern Machine Shop Practice' by Joshua Rose.

Also 'Modern Toolmaking Methods' by Franklin Jones.

You will not find GD&T inside them, but that will be about the only thing they are missing.

At the Internet Archive, almost anything with 'Treatise' in the title will assist.

If you end up putting the leadscrew into the lathe to modify the end of it, think very carefully about the very long, unsupported bit that will be poking out the back of the headstock. Support it well (in a manner closely equivalent to having it in a fixed steady).

I think it is better to make the repair inside the large diameter part of the screw as there is more surface area (greater perimeter and longer depth) for the threadlocker to work. If you did want to thread it, you could drill and tap it 3/8" UNC say 2" deep and then bore and ream the first 1" to 1/2" diameter. That way, the loctited threads give security and the reamed bore gives concentricity. There is also no stress riser at the point where the diameters change.

With the left side, is that the side where the leadscrew is threaded for the thrust bearing? Turn it down maybe 1/16" then build it back up with weld, then skim back to truth. There's a description on here of how John Stevenson did the same with motor shafts.

Now for a real challenge: would it be possible to do all this work on the mill itself, without using a lathe? I think that on a Beaver, it would.

I would just cut off the smaller diameter piece flush with where the step in diameter is, drill and ream the leadscrew and use Loctite 648.

What torque is that part taking? Or are you in the habit of using a Stilsons with a four foot cheater bar on the handwheel? Threading both parts runs the risk of them not ending up concentric, unless you make the shorter bit oversize and finish turn it after locking it into place.

You should probably stone or file the keyway in the entire leadscrew while it is out to remove any burrs. Also have a good look at the drive key that runs in the keyway, and remake it if necessary. It is not the best design, having a keyway along a threaded shaft - the threads act as a cheese grater on the key.

Before you commit yourself to gluing something in, have a think of how you will secure the handwheel and dials to the leadscrew. If you have the necessary taper pin reamer, all well and good, but matching the existing holes in the dial and screw is not simple, especially with five feet of screw waggling about on the end of the piece on which you are working. It might be easier to do the fitting with the small diameter piece free of the main leadscrew. Then you can test assemble everything, make one simple length adjustment and glue it together.

Have a look at the part in the foreground of the lower photo, let us call it the dog assembly. Excessive end float in this can make engaging the power feed like changing gear in an Austin Maxi. A brass washer to shim it all back snug again transforms it to a Honda Accord.

There is enough depth below the teeth of the big one to turn the old teeth off plus a good bit extra and shrink or loctite or Dutch key a large-bored new one on.

The only thing to watch out for is if it is necessary for the teeth of the two to be timed/clocked to each other.

Easiest way to confirm centre distances if it is difficult to do a direct measurement is to turn up two discs that are the pitch diameter of the two gears and bored to the correct measurement. Install on their shafts. If they interfere, the PDs are too big. If there is an excessive gap, the PDs are too small.

Edited By DC31k on 20/12/2020 21:58:53

OK, the three dimensions do not agree.

A 42t gear of 2.164" OD is 20.3 DP, a non-standard size.

A 21t gear of 1.055" OD is 21.8 DP, a non-standard size.

63 teeth of gears on 1.5" shaft centres is 21 DP, again non-standard.

Have a really good measure of the three distances and see if you can come up with some numbers that will translate into standard gears.

The more standard they are, the easier it will be to find replacements. It is good that they only mesh with each other - that way the question of having to determine a pressure angle is irrelevant.

The grub screw in the worm should have a brass pad under it. The shaft should be smooth, not dimpled. The idea is that any bind up overcomes the friction between brass pad and steel shaft.

This might not be relevant to you, but have a really good look at the gears and sleeve between them you show in the plan view. On the later machine, the gears differ by one tooth (so at a casual glance, they look identical) and if you put the wrong one onto the leadscrew first, problems arise later on in the assembly process.

The centre distance for the 20DP size suggested above is 1.575" so if you can find a way of accurately measuring that, it would be good confirmation and make life much easier.

Edited By DC31k on 20/12/2020 19:29:29

You need to tell us the outside diameter and number of teeth on each gear.

Do the two gears only ever mesh with each other?

The centre distance between their two shafts would also help.

The bigger one looks like it is integral with another one. What diameter is the plain bit between the two?

Have a look at how the worm gear on the drive motor is attached to its spindle. This is supposed to be the weak point such that it slips if there is a bind up.

Short answer, as the others have said, is that it cannot be adjusted. If you want to alter it, you either have to change the length of the tailstock screw or change the length of the tooling inserted into the tailstock.

When you say it ejects at 1/4" projection of the barrel, is that 1/4" of the barrel projecting from the tailstock body or 1/4" reading on the tailstock scale? If the first one, that is normal. If you look at the barrel, it should be graduated. The graduations do not start at the end of the barrel. The ejection is a little less than the zero graduation.

It would also be prudent to obtain a proper drawing of a Morse taper shank and make sure any tooling that you have conforms to this drawing.