Member postings for Martin Connelly

Just be aware that there is at least one Mach3 wizard that has an error. I rewrote the facing wizard because in some cases it moves the tool past the minimum X value. I think it was when going from the centre out to the edge, not too bad it you are cleaning off completely but pretty bad if you are leaving anything in the middle such as going from say X=3 to X=15. It may however chip a carbide tip if it contacts the rising side of the workpiece past X=0. I think I may have posted the new script on here already.

Martin C

PS You can use the turning wizard to do the same thing as facing from the centre out. Sometime though, when cutting plastic that results in stringy waste, the facing wizard is a better option to produce short strings instead of long ones.

Edited By Martin Connelly on 18/09/2023 18:07:03

If I was going to drill 24 holes in the face of a backplate I would set the x and y zero on the centre of the backplate then after drilling the first hole use G68 to rotate the coordinate system 15°, drill the next hole, rotate another 15° and repeat until 24 holes have been drilled.

(set up for first hole)

(code to drill hole)

G68 A0 B0 R15

(code to drill hole)

G68 A0 B0 R30

(code to drill hole)

etc

Martin C

Baz, sounds much like using a Scotchbrite type of wheel in a Dremel. The bits that fly off them is best kept out of the workshop.

Martin C

The new arbour has probably got centre holes at each end. Is there any way for you to set it up between centres, such as in a lathe, and check if it is bent? I ask this as the usual reason for changing arbors when I was at work is because someone swung a large radial drill around and clouted a vice with a drill chuck. The neck of the arbor was the weak spot that bent as a result. I would assume a new arbor without any obvious bruising is not going to be bent and that the spindle is likely the source of your runout, especially since you have tried different positions and got the same result.

The other thing to do is when you measure the runout on the end of the arbor mark the high spot on the arbor and the spindle Then see if it moves with the arbor when the arbor is tried in a different position or stays in the same place on the spindle.

Martin C

Classics scholars are the ones who tried to insist you should not split infinitives and so tried to apply this Latin rule to English. That was the reason for people complaining that "To boldly go" at the beginning of Star Trek was wrong and should be "To go boldly". However since it is English and not Latin we can use it as said and thumb our noses at Latin scholars.

Martin C

You could try some Little Hoggers. There was an earlier thread on this Thread 120179

Martin C

Ady1, I think the reason converting lathes to CNC does not get a lot of coverage is that there is such a large variety of lathes that it is hard to be definitive in how to do it. I think it is also likely that there is a lesser demand for lathe conversion than there is with milling. Converting a milling machine works on 3 axes but lathes are usually just 2 so there is a smaller gain in utility. Also a milling machine does not need a spindle encoder or speed sensor unless you go in for rigid tapping, they can do thread milling without a spindle sensor.

I do agree with John Haine though, if you are going as far as putting steppers or servos on the two lathe axes and a spindle sensor (I think the ELS uses a spindle encoder) why not go for CNC. With a used laptop, a suitable interface board and some software, which can be free, you can do all that the ELS does and so much more.

Clough42 has removed his compound slide and put in a solid tool post to improve rigidity but has now produced a video showing how he modified the compound slide mounting for better rigidity just so he could remount it to do a short taper. CNC takes care of tapers with ease so the compound slide becomes redundant.

Martin C

I swapped out an old brass fitting for a new one on Sunday. I have not been hosed down but the fitting at the end of the hose was constantly dribbling and wet feet was often the result. The problem was the plastic compression ring had split and there was always a point where the hose was not sealing as a result. I think the warm weather has allowed the hose to soften and this makes potential leaks into actual leaks. I did not bin the old fitting so a copper sleeve may be worth trying with it.

Martin C

Grub screw is probably anti-rotation. I think the side pocket may have been for an indexing system for a tool post that located in the circular pocket.

I think I would see if I could get an M20 x 1.5 or M20 x 1.75 tap. Make a new tool post stud to suit this M20 thread, and tap the slide to M20. If the circular pocket is no longer in use make a collar to secure the stud in place and put a couple of screws through the slide from below into the collar in the pocket.

Martin C

I grew up in a Victorian mid-terrace. It had a consumer unit with two (rewireable) fuses, lights and sockets. My grandparents had bought the house in the late 1930s and it was then wired up by my uncle. It only had power sockets on the ground floor so if you wanted power for anything other than lights in the bedrooms these adapters were the only way and we used the type with pull cords. Two of the ground floor rooms only had one socket and I remember them being changed over from the old round pin sockets to the current type G. I don't think there were power sockets in all the bedrooms when it was sold in 2003.

Martin C

Emgee, the problem with two Vs is poor repeatability of positioning when fitting the tool holder to the block. It's a bit like the boat in a lantern tool post. It doesn't self level under the pressure of the clamping screw due to friction. The same is true for refitting something in a 3 jaw chuck. Friction stops the part from fully self aligning in the jaws. A tool holder mounted on two Vs is the same, pulling the holder into the block with the piston does not cause the holder to always take up the same position because of friction and because absolute perfection in manufacturing cannot be achieved, there are manufacturing tolerances in the holders and the mounting block. It is interesting to look at the design Harold Hall came up with for a home made QCT. It follows similar lines to what I am suggesting regarding a single V. Looking from above it follows the similar tooling practice for locating parts with two holes that use a round dowel for position and a diamond dowel for rotation. In the case of a tool holder as it is pulled in it rotates round the pivoting point of the V (or a dowel in HH's design) and up against a hard stop.

Martin C

Clive, doesn't your third paragraph prove my point? The two lines of contact on opposing sides of two Vs is the same as two lines of contact on one V. By having them separated by a relatively large distance and clearances between the tool holder and central block the tool holder can take up different positions, much like putting a part in a lathe chuck and not getting repeatability. Using one V and a positive fixed stop at the other end gives better repeatability. Even with one V the weak point in the setup is likely to be the small cross section of the piston pulling in the tool holder either side of the cam hole. Since my stop on the unused V is a Loktited grub screw I can adjust the position of the holder so that the pull in cam over-centres (toggles) to lock the holder in place and does not just rely on friction of the cam on the plunger. It always bothered me that the pull in cam on the various bought tool holders I have all ended up with the cam in a different position when tightened up and only one or two over-centred the cam.

The only good reason for two Vs on a tool holder is so that it can be used on multiple faces of the centre block without changing the basic geometry. Manufacturers do not need to make left and right handed holders and stockists do not need to hold two "hands" with the chance of people ordering the wrong one.

I agree that the error from slightly different mounting positions is going to be small but if you are using CNC that small error is unwanted as tool tables become less accurate mainly along the Z axis. Also for someone making their own holders the process of making two accurate parallel Vs is the hardest part. If it can be reduced to just making one V it is much easier.

Martin C

I have made some of these holders with one V. I think two Vs is over constraining and as a result allows the holders to take up different positions when put on the mount. Due to machining tolerances the two Vs can never be identically spaced from one holder to the next or to the mounting block. This means that either V may be the one that controls the position of the holder when it is mounted or it may be the two outer faces or the two inner faces and it is impossible to predict which it will be. The single V works as well as the twin V for rigidity and has better repeatability of position.

Martin C

This is just speculation.

If they are classed by weight then the lower limit for one of the classes may be 150g. To be on the safe side of legality they aim for a minimum of 152.5g. That means 4 pears must weigh at least 610g.

There is also the issue of things like pre-packaged fresh meat where they aim for an average pack weight. This average is indicated by an E symbol next to the pack weight. You did not mention if the pack of four pears has this E symbol next to the weight. If the limits of acceptable weight for four pears are 600g and 620g then 610g E may be put on the packs of pears.

End of speculation.

I notice that if I buy something like a jar of large pickled cucumbers/gherkins they often have a piece that is cut off a whole item to meet the weight or fill the jar. With a fresh pear that is not something you can do.

Martin C

You mentioned no sign of an average as I was composing this.

Edited By Martin Connelly on 05/09/2023 08:59:44

It looks like a capacitor start, capacitor run motor. Some of these type of motors have limited numbers of starts per hour or require some minimum time between starts so not shutting off the motor when at pressure may be due to this to ensure it does not overheat. The other possibility is that the motor will struggle to start if it is trying to work against existing pressure in the tank on restart. I have a compressor that bleeds off air until the motor is up to speed to allow it to start in a reasonably short time.

Martin C

Runout needs to be the same at different distances from the headstock, not just at one place, if there is any swash in the backplate there could be different runout at different positions along the workpiece. The runout also needs to be consistent, did you try different positions for mounting the collet chuck? Does the runout change if the collet chuck is removed and replaced?

I have OEM "in the spindle" collets so I have never bothered to check the runout, it seems to be OK for my needs but if you are making your own collet chuck these are things to consider and check. If you have a consistent runout judicious use of a piece of fine abrasive may remove most of it. If it changes with mounting rotation find the best position and mark it in some way for repeatability.

As for what is acceptable, what do you aim for when dialling in a 4 jaw chuck? This will give an idea of what is expected of a collet system. In the end acceptable values of runout are those that meet your needs.

Martin C

Sorry, missed the bit where it mentioned a spreadsheet. I have had family visiting this week and was just skimming through posts to catch up. However having done a quick spreadsheet for10 blocks it is clear that you could add a column that counts how many blocks are used for each combination. 10 blocks results in 1024 combinations and there are likely to be a few that give the required result. A column that also shows the number of blocks used will help pick the one(s) that require the smallest number of blocks. Sorting on the resulting total for the blocks and then the number of blocks allows you to quickly identify a combination that best suits needs.

Martin C

There is another way to do this type of there/not there calculation, that is to use a spreadsheet in the "sledgehammer to crack a nut" way.

List all the blocks across the top row

In the columns below, in the first column going down put 0 1 0 1 0 1 .....

In the second column put 0 0 1 1 0 0 1 1.....

In the third column put 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 ...

Do this pattern for all the blocks increasing the repetition of 1s and 0s in powers of 2 for each column until you get down to a row that is all 1s

Finally add a calculation in the cells of another column that uses absolute positions for the sizes and relative positions for the 1s and 0s to add together the values of the there (1s) blocks for a total thickness.

If you add a column numbered 0 to whatever the bottom row is (eg 7, 15, 31...) you can sort the table into total size and unsort it back to original order as required.

It is probably easier to do than to explain but is easier for a non-programmer to understand and expand to their own needs. If we had the option I would create an example, it would only take a few minutes, and post it in the forum.

Martin C

You can set a ramp down time that actively brakes the motor within some limits. One is that you may need an external resistor to take some load. Another is that some trip on overload if ramp down time is set shorter than the VFD can handle. This trip stops the control of the ramp down speed so it then coasts to a stop. More likely to happen if you have a large rotating mass starting from a high RPM. Setting a ramp down time does noticeably reduce coasting to a stop time but it will not stop within one revolution from typical speeds (not using back gear for example).

Martin C

I would be looking at what Tony suggested. When I put a ballscrew on my mill X axis I got one that was a little too short to go all the way along the table based on the fact that at the extreme travel of the table the shorter ballscrew would still be sticking out beyond the ball nut. As I was going to be machining something for the (in my case) right hand end anyway it did not make sense to pay for more length of leadscrew than was necessary and machining a hardened ballscrew did not appeal.. This was for a ballscrew with pre-machined ends so I made an extension to fit on the machined stub of one end.

Martin C