Member postings for Simon Williams 3

Hello Moderators, here's an oddity

According to the "latest forum posts" the last entry in this thread is from Brian. timed 11/12/2019 09:24:48.

But that isn't what I'm seeing above.

Uh?

Simon

So - Thank you guys for your kind and positve vibes, and it all turned out well if I do say so myself.

Here's the picture I took of the two gear clusters I made before sending them off to sunny Denmark:

As before, I have made the smaller gear on a stub, then a second gear cut as a plate wheel is soft soldered onto the stub. This time I made them out of EN32B, 'cos that's what I had.

Pleased it turned out well, and thanks for the encouragement off camera from Jesper and from Brian.

Season's greetings all

Simon

Good morning all,

Not sure case hardening is really about absorbing carbon into the surface layer of the parent at all. My understanding is that it is actually a nitriding process, which is why you don't need to quench the treated part to get a hard finish layer. Or do you need the carbon to capture the nitrogen into solution?

So bone meal has nitrogen in it, as does anything cyanide flavoured, but simple barbecue charcoal isn't (I think) going to do the whole chemical/metallurgical/structural alteration.

Any metallurgists?

Best rgds to all

Simon

Hello all, interesting question.

First observation is that you'll never get a pretty slot using a saw with set on it. So using hacksaw blades whether they're wavy set or a normal wood saw with side set ain't going to cut it, if you'll excuse the pun.

The slotting file is the right tool, and I seem to remember correspondence from Ketan a few months ago saying he'd bought some - inadvertently - and were they of interest. Might be worth a call. A slotting file is a thin file with no teeth on the faces, but teeth on the edges so it cuts sideways - or downwards if you hold it on edge.

A razor saw has also been mentioned - that's what the top saw in this pic' is:

xxx

That's an Exacto razor saw, and it will cut brass but the blades aren't very hard and you'll lose the teeth pretty quickly. It's really intended for cutting wood, particularly balsa or similar.

The bottom one is a jeweller's backsaw - if you can find a source of good quality blades it'll cut anything a hacksaw would, as it's designed for cutting 9 ct gold which is pretty hard. The blades are removable from the brass back, and usually double sided, but beware of cheap and nasty blades they won't last as well as the razor saw. My Dad was a silversmith, we used to have periodic expeditions shopping for this sort of stuff round the Hockley Hill area of Birmingham, but I've lost touch with this over the last thirty years and I don't know where to buy these blades. I remember the good blades were expensive, but the cheap ones were absolutely useless!

The backsaw willl give you a good groove to guide the slotting file and remove the majority of the metal, but the blade on this one is only 0.35 mm thick so you need parallel cuts or follow with a narrow file to get a slot which looks in proportion. You used to be able to buy blades of different thicknesses.

HTH Simon

Edit - MichaelG has - as oft before - pipped me to the post about ARC's slotting files.

Edited By Simon Williams 3 on 27/11/2019 14:36:30

Brian, Jesper, good evening

I've spent a happy few moments with my trusty calculator and come up with the following

Sorry it's a bit small, I've had all sorts of fun and games converting a table in Word into a format the forum can accept.

Obviously my table only shows the pitches where the 29/16 or 31/17 gear is relevant. The error figure in the last column is the pitch error in mm per revolution compared with the nominal "correct" number. Of course we're quibbling about microns here, and arguably the limiting factor is the wear in the lead screw and halfnuts. But that just explains why the traditional 127 tooth gear isn't necessary.

Using a 31/17 gear consistently gives a greater error than the 29/16 gear I chose. I can't take the credit for that as I didn't do the analysis at the time, though it looks like we came to the right conclusion nevertheless.

Best rgds, let's post this and see if it's legible!

Simon

xxxxx

My my my, thank you Brian for those warm words, JS did most of the thinking I just followed his instructions!

As ever your ingenuity in finding an alternative answer amuses me no end. I'm currently planning a house move so my playtime in my shed is somewhat curtailed, but I'll sit and work out what the error is for the useful range of threads doable on the Myford by using a 17/31 cluster as a comparison.

Your point about wrecking the internal gears of the gearbox is well taken - as is EGA's observation of the warning in the handbook. I'd just note that if someone is contemplating cutting coarse threads there is plenty here in the forum about using a mandrel handle to circumvent this problem.

FYI Jesper has been in touch and we have agreed I will supply copies of the 16/29 and 17/30 cluster gears I made for myself directly to him. I'm looking forward to hearing his M62 x 1.25 thread has come out right on the second visit.

And just in passing, I only got involved in this whole saga because I wanted to cut a 20 mm x 2.0 pitch thread in a nut to fit the U2 collets on my Deckel grinder. I was intent on making a U2 variant of JS's ER32 collet blocks for milling a hexagon on something. I'd cut 13 TPI and got good results - the collet closed up and all was well, but my interest was piqued by having read about the 33 tooth gear idea, so I made one but couldn't fit it. JS explained I was barking up the wrong tree, and the rest is history.

I'll publish a error table shortly, it's only a few minutes work with a calculator.

Best rgds Simon

Richard -

Have you got a neutral connected to the switch gear - I think I'd expect the contactors to need a neutral to the coil, not be connected line to line. But if they are line to line the point above about being connected to the transformed input phases and not to the false phase is valid. What voltage is your converter actually applying, and is it three phase at all?

If you are not familiar with machine electrics that might sound like gobbledegook, so can you tell us where you are, maybe someone with the relevant expertise can assist "hands on".

Good luck, you need a competent sparks used to working with industrial machinery. Keep your fingers out of the birds nest, them wires bite!

To continue my saga:-

We've worked out how to play tunes on the gearbox output ratio to achieve a much wider variety of TPI or pitch settings that the original set by altering the input mandrel gear. It's only fair to say that this is the essence of Brian Wood's book, and far be it from me to steal his thunder.

The difficulty we've created for ourselves is that the input (mandrel) gear for the MK1 S7 QCGB is this all in one double gear. The Mk 2 lathe has a demountable driver gear so (within reason) you can fit any standard change wheel to it and play even more tunes on the TPI chart.

So the two compound gears I made were as follows Here is a picture of the standard 30/12 set on the left, and my 30/17 set on the right.

xxxx

I made it out of bronze because it is made in two parts - a 17 tooth gear cut on a stub shaft, then a 30 T wheel soft soldered onto the stub of the smaller gear. Besides , it looks pretty and I can recognise it easily in the drawer full of change wheels.

So the same principle applies to the gear I christened the 16 -1/2 tooth gear, which was a bit of a joke between me and the late John Stevenson. I don't have a photo of it, but it looks exactly the same as the bronze one above except the small end is cut for 16 teeth, and the larger one is cut for 29 teeth, but on a blank of the same outside diameter as a 30 tooth gear. This ensures that it meshes correctly with the existing tumbler gears - there's no adjustment on the centre distance - but runs at the speed of a 29 tooth gear - it's only got 29 teeth so it runs a bit faster (as a driven gear the smaller it is the faster it runs).

So far so good, but where do you buy one? Well, I only know one answer to that - simply there's the one in the picture in a drawer in my shed and I'll make a duplicate if you are interested. If so send me a PM and I'll explain further. I'm not too clear on the etiquette of touting for business openly on the forum - mod's please note I make no direct charge and all of the money goes to charity. It's such a specialist application I don't think anyone is making them commercially but if anyone with the technology fancies their chances be my guest. Indeed if Jesper knows anyone who can do this locally please have a go! Not sure if 3D printing technology is up to this, but some day someone will try it out.

So I hope that answers the questions remaining, if not let me know what's not clear and I'll try and fill in the gaps.

Good luck, and the arithmetic and more details are in the two threads referenced above.

Ooh, and Brian's book goes into much more detail, with pictures and set-up tables and I recommend it.

Best rgds Simon

Jesper - good evening. Firstly my apologies for not picking up your questions sooner, I've been out of touch for a day.

So, if I may I would like to summarise what we think we know, and answer your question about obtaining the requisite gear.

I believe we have established that you have the older type QCGB, that's what the photo's of the lathe indicate, and the fact that you have a 30/12 cluster gear in the gear train on the mandrel gear suggests the same. What we don't know for absolutely certain is whether the lathe cuts the threads as indicated on the gearbox label. It's just possible that there is a mismatch here, and the gearbox cuts either twice or half the threads per inch that is indicated. This of course is when you are cutting an imperial thread such as 20TPI. The label inside the gear cover sows a seed of doubt in my mind, and I would just like to be sure we know where we are starting from before telling you what to do next. Can you try cutting a thread with the gearbox set up for (say) 20TPI - it needs only be a scratch pass to prove the gear train - and confirm that it is 20 TPI that you get.

I chose 20 TPI because it is the same thing as 1.27 mm pitch - and if the M62 nut you were trying to make is only fairly short you may find that using a 20 TPI thread instead of 1.25 mm pitch gives acceptable results. After all the error is only 0,02 mm per revolution - so if your nut is 10 mm thick the thread error is still only +/- 0.08 mm.

However, we can do better. ASSUMING the gearbox IS the old one, the gear train you have identified starts with a 30 tooth gear with a 12 tooth gear married to it to make a compound gear. This gives the thread per inch selection as per the label on top of the gearbox.

If we change the 12 tooth gear for say 18 teeth we'll get a different set of TPI values for each position of the gearbox selector levers. In this example the gearbox output is speeded up by 50%, so our 20 TPI setting will now cut 13.3 (Edit - I got this wrong in my initial post - it's less TPI because the lead screw is running faster.  OOPS!)  TPI. Using other input gears gives us control over the threads on the workpiece over and above the "normal" (standard settings.

One such choice is a 17 tooth gear. If we change the 12 tooth gear to be 17 teeth, and set the gearbox levers in the positions for 24TPI we get 16.94TPI, which equates to 1.499 mm pitch. The arithmetic is simple pro-rata gear ratios, but work it out in TPI and then convert to mm pitch last.

So that's a nice accurate answer for 1.5 mm pitch. also 3mm, and 0.75 which is an oddity but it shows the principle. How do we go about getting 1.25 mm pitch.

Do the same arithmetic, and you find you need a 16-1/2 tooth gear. Now whatever I've said before, I haven't worked out a way of making a single non whole number gear. But it isn't a single gear we're working with , it's a compound pair. The other gear of the compound pair was always 30 teeth, giving a 1 : 1 ratio with the chuck, but it doesn't have to be.

16-1/2 : 30 is very close to 16 : 29, so if we change the input gear to be 29 teeth meshing with the tumbler gear, and now 16 teeth driving the otherwise standard gear train going down to the gearbox input we get jolly close to 1.25 pitch. And 2.5, also 0.625, another size we're likely never going to want!

Enough for a moment, this post is long enough already, I'll continue in a second instalment.

Edited By Simon Williams 3 on 25/11/2019 19:37:41

Ah ha!

Hello again Jesper, and thanks for the pictures. You obviously sussed out how to load them!

But...….

The plot thickens! Because the label you show from the inside of the gear cover isn't the right one! That label belongs with the standard Myford metric banjo kit, which was marketed to go with the NEWER gearbox, not the one you and I have got. The 2:1 gear reduction factor has to be taken into account.

So here is a picture of the label inside the cover of the gearbox I have.

If you have followed the links I showed earlier you will have realised that this is irrelevant, as using the 17 or 16-1/2 compound input gears sidesteps this problem.

HTH Simon

MichaelG - thanks for your contribution, that's a very neat solution to the problem. However one of the main advantages to using substitute gears on the mandrel and retaining the other (standard) gears taking the drive to the gearbox input gear is to retain the availability of the fine feed function.

Brian and I examined the pitch error introduced by using 17 and 16-1/2 tooth gears, I'm happy the error is insignificant and within the normal working tolerances of the single point screw cutting technique.  There is a summary table in one of the threads referenced which lists the theoretical error.

There's nothing not to like, apart from the intricacies of making these compound gear sets.

HTH Simon

Edited By Simon Williams 3 on 24/11/2019 13:07:45

NDIY -

The explanation is in the two threads I have referenced.

No, I haven't worked out how to make a half tooth - well not intentionally anyway. The key to this is that the gear cluster is either a 17 : 30 or alternatively 16-1/2 : 30. as a compound pair. You need both sets to achieve the whole range of standard metric pitches for the sort of sizes a Myford owner might encounter.

16-1/2.: 30 is almost exactly 16 : 29, so that's what I made. John Stevenson worked out for me how to make a 29 tooth gear on a 30 T blank so it would mesh correctly with the tumbler gears. Details are in the two threads referenced above.

Works a treat, and I'm fairly certain that's what Jesper needs to solve the pitch error problem he has encountered.

Rgds Simon

Good morning Jesper, hopefully Brian Wood will chip in shortly, can I suggest you need a copy of his book "Gearing of Lathes for Screwcutting" which covers this in detail and answers the questions you are asking. Copies usually available from Arc Euro and also Amazon.

Essentially you need the magic gears, but I suspect from the date you have the early version of the quick cutting gearbox, the difference is a 2:1 reduction in the later box which means for you that you need a 17 and a 16-1/2 tooth gear on the mandrel to cut accurate metric threads, not the 34 and 33 tooth gears. If you can work out how to post a picture of the right hand side of the gearbox we can determine which gear box you have, and hence which gears are needed.

[Edit:  the 30/12 cluster gear in the gear train you have described in your initial post confirms that we're dealing with the older style gearbox.]

The 16-1/2 tooth gear sounds mighty odd, but don't worry all will be explained. You might like to do a bit of back ground reading on the subject here:

https://www.model-engineer.co.uk/forums/postings.asp?th=125630

QCGB1 should take you there

Also:

https://www.model-engineer.co.uk/forums/postings.asp?th=131985

QCGB2 links to it.

To post a picture you have to upload a JPG to a site album first, then reference it in your post. So , log in, go to Albums on the header and start a new album giving it a suitable name. Now up load your picture in jpg format, it helps if the file size is sensible but I think I've read that the forum software sizes the file anyway if needs be. Now you can write a forum message, and include your picture by using the little camera icon in the message header. This opens a screen where you browse to the picture you want to include in the album you just started, and then pastes it into your message for the rest of us to see. Hope that makes sense. and if I've missed anything out please could those with more knowledge correct my summary.

Hope that helps, this is a subject with plenty of reference stuff on the forum, so I'm sure we can sort out the problem you have encountered.

Best rgds Simon

Edited By Simon Williams 3 on 24/11/2019 09:42:53

Simple factual answer to the question as stated - "could the e stop be wired into the single phase supply" is "Yes!"

But the question deserves some more thought than that, and the pro's and con's of putting an E Stop in the supply need a bit more consideration.

Firstly, my apologies if I've missed the relevant post, but I don't think I've seen what sort of machine this is. This arises because I think the proper answer to your question needs a rudimentary risk analysis to inform the design decisions. If the machine is something nasty and dangerous, then the risk of injury is offset (controlled) by being able to stop the machine quickly, and switching off the mains supply won't necessarily give you the optimum control. If your arm is in the mangle then having an active machine braking function activated by the emergency stop would probably be a good idea.

If the hazard is of the machine going berserk because of the inverter losing control then shutting the mains off is your only option. For the milling machines and lathes I have in my shed I think this is a very unlikely that a modern good quality inverter is going to go doolally, so I have the E Stop function as a stop signal to bring the inverter to zero frequency and with DC braking if appropriate. The inverter stays energised so it can do what it needs to do to effect a rapid stop pf the offending machine.

So the short and simple answer to your question is "Yes" - you can connect some sort of E Stop functionality to remove the mains supply from the machine motor, but my preference is to use the extra control facilities of the inverter to give me a quick stop function and for this I leave the supply on the inverter but tell the inverter to stop the machine. There's a big red latching button on the machine within reach, and if I bang it the machine comes to a stop whichever way it is configured.

Horses, as they say, for courses.

Hope this helps Simon

Another vote for 34 degrees or something close from me.

I've read that the included angle is (slightly?) dependent on the radius of the vee pulley groove - the tighter the bend the less the included angle. Whether it's significant and to what extent the default 34 degrees figure needs to be modified is very likely something Mr Fenner's drawing office keeps in a safe place.

I've made Z and A section pulleys of a variety of diameters using 34 degrees included angle and been happy with the results.

HTH Simon

Hi Roger

I've got a Mk3 in the shed, and it's been sat on a shelf for about 30 years. I pulled it down to admire it not long ago and realised that the needle was sticky. Pulling the back off shows very fine iron filings aggregated to the magnets in the coil gap.

On one of the radio/electronics forums I was advised not to take the magnets off the movement, as cycling them through the magnetic remanence path of removing them and replacing them will change their strength. These are red eclipse pot magnets before the days of rare earth.

It may be an old wife's tale, but if so it's well entrenched in Avo folklore that removing the magnets is a one way trip.

If you know different or have a cunning work-around do tell!

Best rgds Simon

Good evening again, looking for some advice please.

I am planning a house move, and included in my goods and chattels is a Tom Senior Light Vertical.

The question arises of how to pick it up safely. I don't like the thought of putting a sling around the cast iron bar at the top of the machine where the head is attached - all of the weight has to be carried by those two little castings clamping the head. Or am I being over-cautious? Having the casting crack would be the end of Mr Senior.

I moved it on a pallet the last time, but of course it's top heavy.

Any thoughts?

Why?

Is this because the oil is viscous when it is cold, or perhaps the engine is a bit tight yet and will slacken off when the running in process is complete.

Cure the problem, not the symptoms.

I'm uncomfortable with the idea of dumping oil flow back to sump during start-up, that's when you most need the oil in the bearings where it belongs. High pressure on start up suggests the oil isn't getting where it needs to be - or at least not in the quantities it should. I'm assuming the oil pump is a positive displacement device so the rise in pressure indicates a restriction in the flow around the engine. If it is actually caused by the oil being a bit thick and treacly when it is cold then there are other oils to choose, at least for the running in phase.

I'd fit a higher scale pressure gauge as a temporary measure and see how this problem develops as the engine hours accumulate.

HTH Simon

Try googling "belimo actuator" for a wide range of damper actuators - essentially a motor driven crank that turns trough 90 degrees. Other variants are available but looking at ebay at the moment there seem to be quite a few on offer around the £25 to 30 sort of mark.

HTH Simon

I know I keep banging on about this, but the idea that there might be enough resistance in the earth path to allow any item within the installation to reach a voltage sufficient to experience a tingle needs de-bunking, as it flies in the face of everything the Reg's have to say about earthing.

Sure, the earth path will have some resistance, but the installation must be such that the potential reached by any exposed conductive parts (metalwork) is always below the safe touch voltage, even under fault conditions (whilst blowing a fuse). Safe touch voltage is usually deemed to be 25 volts RMS, that won't give a tingle though the actual requirement is that anyone touching an exposed metal surface must not be given sufficient shock to cause an involuntary movement.

This places some pretty tight constraints over what earth path resistance can be accepted, and rightly so.

Cows being very susceptible to electric shock there are specific additional requirements for livestock pens, farms etc.