Member postings for Andrew Johnston

Including my 4" build, complete with fumbles and how not to do it.

Andrew

No idea what the board is for, but the black blob looks a bit like COB - aka chip on board. The basic silicon chip is placed on the board and wire bonded to pads on the board. The whole thing is then covered by a blob of epoxy. It's a cheaper assembly option because the silicon chip does not need to be packaged before use.

Andrew

This is what Jason is referring to, an Ainjest high speed threading unit:

It has it's own half nuts and also has a repeatable trip mechanism. When threading to a shoulder, or internal threading, I leave 5 thou clearance. My lathe is imperial, and the Ainjest unit only works on imperial threads.

This is a M16x2 thread screwcut in the normal manner, using an insert, leaving the normal saddle half nuts engaged:

Andrew

Agreed; it would be better to forget the beam and bolt the lathe to the angles at each end of the frame. That way the structure resisting torsion is much larger and stiffer. Clearly the actual loads imposed by cutting forces haven't been thought through. This is the stand I made from 50mm square ERW tubing for my CNC mill:

Andrew

Must be something wrong with the tips or the technique? Touch wood I've never broken or chipped a threadking insert. I normally buy inserts from Cutwel. This is a 3/4" BSF thread screwcut with an insert:

Andrew

EN31 is similar in composition to gauge plate, minus the tungsten. It is hardened and tempered in a similar way. Like gauge plate it is formulated for oil quenching rather than water, at least for small sections. Hardening and tempering isn't that much different to silver steel in terms of temperatures. So if silver steel isn't suitable, or can't be hardened and tempered properly, I doubt EN31 will be any better.

Andrew

Plus one, there has to be something fundamentally amiss here. A picture taken from futher back would help, so we can see the whole picture, so to speak!

Just checked on my lathe (Harrison M300); when the cross slide nut is about to become disengaged the axis of the lathe spindle is roughly aligned with the centre of the 4-way toolpost. So a tool as shown would be well past the work centre.

Andrew

For each thread form there are two types of insert, partial profile and full profile. A partial profile insert will cut a range of tpi/pitches. The root profile will be set for the largest tpi (smallest pitch) and will be too sharp for lower tpi (larger pitch). So some tweaking of DOC may be needed. In addition the crest will not be formed properly. In contrast a full profile insert will cut a full profile thread, including root and crest, thread but is limited to a single tpi/pitch value. Of course there are also variants for internal/external and LH/RH.

I screwcut a lot of the threads on the lathe, mostly imperial but also some metric. I prefer to buy full profile inserts. It makes life easier and forms a better thread profile. I buy inserts as needed. In practice I cut a fairly limited range of tpi/pitch values. For instance the majority of Whitworth threads I cut are 40, 32 and 19tpi.

For reference, a partial profile insert for metric threads:

A full profile insert for metric thread, 2mm pitch:

Internal and external inserts for Whitworth 19tpi:

Andrew

Silver steel works fine for me when making cold chisels. So I suspect your hardening and/or tempering is off. Are you using an electric furnace or the heat 'n' hope method?

An 8mm diameter 250mm long rod is very slender. So irrespective of what steel you use it will be prone to buckling under load; basically a design flaw.

Andrew

I gave mine away as I never used them. To be fair I have a selection of accurate parallels made by the father of a friend who was a toolmaker.

Andrew

Malc: Thanks for the background information. All I can say is good luck, and I hope someone on here is able to help.

Andrew

If I understand the problem correctly the fingers will not allow the correct indexing to be achieved. I'm slightly surprised by that; 4/5 of a turn should be possible. If, for some reason it isn't, then simply index the 1/5 of a turn and move the fingers back at each index rather than forward. It helps to think if the indexing in terms of spaces rather than holes.

Andrew

An obvious question is why did the original strap snap, before making new ones from, presumably, the same castings.

A mill isn't essential. Many of the critical surfaces such as the sides, bore and grooves for the sheaves are done on the lathe. The rest can be done by hand or by making use of the lathe for light milling/drilling.

Andrew

I could, but am not that close to Wisbech, and I am too busy. Anyone contemplating this job would need more information, such as:

Are there proper engineering drawings?

Do you have material/castings or is this to be sourced?

Do the mating sheaves exist?

Is this a commercial job, and why are you unable to machine them yourself?

Andrew

Very likely, but I'm not in the market for one. I've never come close to needing a self-centring machine vice. I suspect the commercial ones are mainly used on CNC mill production runs rather than manual mills. If I was in the market for one then repeatability would be of the utmost importance. It's a bit like a 3-jaw self-centring lathe chuck. If it isn't self-centring......

Andrew

Looking at the picture posted by the OP the machine doesn't seem to have dials. Despite discussion of a nut to move the table I don't think it has one. It is stated that turning the screw thread, with the gear locked, moves the table, so the locked gear is in effect the nut. With the gear unlocked rotating the gear uses the screw thread as a rack.

It's a mistake to assume this is a hobby style manual mill, it isn't. It is a production mill, probably for a basic second op. It would have been set up by a tool setter and used by semi-skilled labour moving levers with distances, possibly only in X, set by stops.

Andrew

The calculations by Nick are wrong. For a start the gear looks to be helical. The pitch diameter D is:

D = N/(Pn x cos (A))

where Pn = normal diametrical pitch, ie, perpendicular to the teeth, N = number of teeth, A = helix angle of the teeth

The helix angle is going to be quite small so the cosine will be very close to unity. So we can simplify to:

D = N/Pn

We also know that the outside diameter, O, is:

O = D + 2/Pn since the addendum is 1/Pn

If we ignore the helix angle we get the formula for spur gears:

O = (N + 2)/Pn

or Pn = (N + 2)/O

if N is 66 and the OD is 4.1" then the diametrical pitch is:

Pn = (66 + 2)/4.1 equals ~16.6

Since pi is involved either the TPI of the worm, or the diametrical pitch of the gear, must be irrational. For convenience of making the gear the diametrical pitch of the gear is usually a standard value giving an irrational number for the TPI of the worm.

Industrial lathes, including mine, will cut the irrational TPI values corresponding to integer values of DP allowing worms to be made. Of course the TPI values on the lathe use an approximation for pi.

Andrew

Thanks for the elucidation.

If no accuracy figure is provided how does the user know that it is appropriate for his hobby application.

Andrew

A question I would have asked, is why was it in a clearance sale?

None of my machine vices have a swivel base. I've swivelled the vice to machine a part once in the last 20 years. So I'd take the swivelling base off; it just uses up height and reduces rigidity.

There are some (very expensive) self-centring vices available Self-centring Vices with accuracy figures. I'm not sure cheaper variants would have good repeatability.

A problem with self-centring vices is that one loses a fixed reference as both jaws move, and could twist, rather than one jaw being fixed.

Andrew

Only need one word - no.

The "replacement" motor is a totally different animal. It will need a different control board and most likely some mechanical changes to make it fit. I'd agree with Baz; I doubt the sale price would justify the time and money needed to change the motor, unless it is an exact like for like swap.

Andrew