Balance Springs II

[Sam StonesParticipant @samstones42903]

Help !!!

[Sam StonesParticipant @samstones42903]

Gentlemen,

Once more, I need some help.

 

Further to my posting – Balance Springs – back in June this year, a significant turning point has been reached in completing the escapement of my skeleton clock, ie. winding the helical spring. Its dimension are :-

 

Wire diameter = 0.008″ (0.20mm)

Nominal spring diameter = 3/8″ (9.5mm).

Number of turns = 9 (ish)

Pitch = 0.04″ (1.0mm)

Spring material – Plain steel guitar string

Tests have show that a mandrel diameter of about 3.9mm will produce the necessary nominal spring diameter.

On many occasions, I have succeeded in winding springs onto a mandrel (for other purposes) and with larger diameters of wire. The shank end of a twist drill proved to be ideal, with the flutes being somewhere to shove the end of the wire. It was also easy on my ML7, with back-gear, forward and reverse, and a Norton gearbox to (instantly) select the desired feed pitch.

My first effort today, resulted in a complete tangle.

The wire (`wanting’ to both uncoil, and twist sideways), flew around itself, ending in a scrambled mess. I had actually envisaged something like this happening, but felt somewhat hamstrung as to how to stop it. So I simply cut the `feed’ wire after winding enough wire onto the mandrel.

Clearly, the problem is “How does one control the uncoiling so that the coils remain uniform?”

I wondered if it would be possible to slip a correctly-sized sleeve over the coiled wire, (before making the cut) and then releasing the spring, one coil at a time. That will be my next test, unless your comments offer a more compatible solution. Please note that although I have the use of a lathe, I otherwise have a very limited workshop and materials.

I found the following web page, but there was nothing mentioned about controlling the uncoiling wire.

http://www.oldengine.org/members/holland/images/MakingSprings/Thumbnails.html

Your help would be most appreciated.

Regards,

Sam

Edited By Sam Stones on 23/10/2010 05:46:12

[Ian S CParticipant @iansc]

I had a look on google, put in How to make hair springs for clocks, and found;

http://www.spring-makers-resources.net/blued-steel-hairsprings-for-watches-and-clocks.html       Might be worth looking at, I imagine you are making a flat helix, not a cylindrical spring. Ian S C  I’v tryed clicking on the above site and could’nt get in , you might have to go the long way round.

Edited By Ian S C on 23/10/2010 08:00:12

[john swift 1Participant @johnswift1]

Hi Sam

 

why don’t you slowly reverse the spindle by hand to  unwind the spring

to take the tension off before you cut it

 

could you try a threaded mandrel ?

 

 

I assume the wire is straight before you make the spring

so you don’t have a combination of the bend you are putting into the wire and any existing one

 

 

    John

 

 

Edited By john swift 1 on 23/10/2010 11:15:32

[maurice bennieParticipant @mauricebennie99556]

Hi SAM   PLEASE do not cut a spring under tension on a mandrel ever.

My workmate did ,and the wire spun round and sewed itself down his finger.

I had to cut him loose from the lathe. It was not a pleasant sight.

best wishes MAURICE. 

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

The traditional method of making very small springs with limited equipment is to pass the feed wire through a hole in a piece of flat soft wood and arrange for the flat part of the wood to bear down on the coils that you have already made . Until such time as you deliberately remove the wood the coil remains tight and under total control . You cut off surplus feed wire with the wood still in place and then slowly lift it away . As an added bonus there is an element of screwcutting going on in the wood and after a few turns  a neat set of grooves form which help to keep the winding pitch constant . Please be very careful when winding springs under power – it is a terribly dangerous process .

[Sam StonesParticipant @samstones42903]

Gentlemen,

I really appreciate all your ideas and advice, especially about safety. As something of a coward who hates the sight of blood, particularly my own, I’m exceptionally cautious. I’m taking Warfarin too, which adds even more to my caution. I also turn off the power to the lathe so that there is no possibility of starting it by mistake. There’s no back gear either, and speed changing is too much trouble. For me at least.

Ian,

I also tried the website and failed.

John,

Because the wire is so thin, I suspect that the coils will try to twist sideways once they are set free.

Maurice,

That’s certainly good advice for everyone. I’m taking this task relatively slowly and warily, even though the wire is quite thin.

That said, it appears that I wasn’t all that clear with my description and, so that I don’t confuse others like I often confuse myself, here are a couple more comments.

The guitar-string wire is quite fine, measuring only 0.008″ (0.2mm) diameter, or about twice the thickness of a rather thick human hair. The string is sold, loosely coiled in a relatively large envelope, and is almost perfectly straight when released. The balance spring is helical (in wikipedia, under Balance spring, they call it a chronometer helix). It measures about 3/8″ in diameter, has about 9 turns of wire with a pitch of about 0.04″. This has to produce a spring length of 3/8″. On the clock, it is positioned vertically above the balance wheel, and connects this wheel with the clock frame.

The new mandrel (on which I’ll be winding the wire), will be about 3.9mm diameter, and cross drilled with a 1mm hole near the lathe chuck end. The end of the wire will be held in this hole with a tapered pin. From the tool post I (currently) have a rod and a pulley sticking out towards me. In that position, the wire can be weighted and can hang down in front of the lathe and bench. I haven’t found any signs of this pulley affecting the straightness of the wire. I’ll be winding the chuck backwards to achieve a RH helix.

After some overnight thoughts about what is happening to the wire after winding it onto a mandrel, I think Michael’s comments fit very closely to my current set-up.

Considering the small size of the wire and the winding pitch, along with the atypical nature of wood grain, my choice of material to `trap’ the coils will be some sort of plastic. Because I’m very familiar with polypropylene, I’ll try that material first, probably cutting a suitably shaped section from an old plastic crate. This material has a slightly higher softening point and a higher coefficient of friction when compared with polyethylene, which may be too soft. I shall also wind the wire slowly so as to avoid melting the plastic due to friction. Teflon (PTFE) is more attractive in terms of temperature resistance, but I would expect that there’ll be insufficient friction to control the wire. In any case, I’ve given away my stock of raw materials.

I intend to `close’ wind the wire (0.008″ pitch), and will either pull the coils apart by hand later, or perhaps even rely upon the weight of the balance wheel to do that. This should also have a subtle benefit in reducing the (balance wheel) load on the bottom thrust plate.

I’ll incline the flat face relative to the lathe’s main axis so that as I back off the top slide, the right hand coil is the first to be released. Further backing of the cross slide, should allow subsequent coils to be released in sequence. That’s perhaps an ideal situation, but the proof of the pudding . . . etc.

Best regards to all,

Sam

Edited By Sam Stones on 24/10/2010 04:22:09

[John HaineParticipant @johnhaine32865]

I suspect the problem may be that guitar string wire is hardened and isn’t taking up the helix, so when you release it it just goes straight again.  Why not soften it by heating to red, wrapped round the mandrel, allowing to cool slowly, then re-temper?

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

I forgot to mention a little bit of engineering science :

 

When winding any sort of spring you are not only winding the wire around the mandrel but you are also rotating it about its own axis . Conceptually the wire has two twists in it – one to curve it around in a circle and a second smaller one to twist it sideways . The quality and performance of precision springs is greatly affected by how this second twist component is managed . If the feed wire is free to rotate then there is no problem and a stable spring will result . If , however , the feed wire is prevented from rotating then a quite large torsional stress gets trapped in the wire and when you release the completed spring from the mandrel this stress is released and in addition to the normal spring back that occurs the wire of the spring will rotate about its own axis all the way along and you end up with a ‘ coiled coil ‘ and an unstable or unuseable spring .

 

 

 

[Roger WoollettParticipant @rogerwoollett53105]

Could it be you do not have enough tension on the wire while you are winding the spring. To get a permanent bend it is neccessary to exceed the elastic limit.

I have wound a number of small coil springs using guitar strings without serious problems.The fact that the spring ends up nearly twice the size of the mandrel also seems to suggest this possibility.

[Gordon WParticipant @gordonw]

I made a tool for winding small springs from guitar wire, just a piece of bar with two brass washers and a bolt, adjust tension by trial and error, as noted you must have a fair load on the wire to make it stay bent. I use a hand wheeel on the mandrel for winding.

[Ian S CParticipant @iansc]

I got the wrong spring, but this may help all the same:Quote from the above site.

Carbon balance springs by:Anonymous.

What I learned at watch school is that two (for better quality springs) or three (for lesser quality springs) carbon steel wires of the correct dimentions are sandwiched together and wound in a spiral in a round metal box of the correct size and heated in a furnace to harden, then temper.

The springs are then removed from the box and separated. The space between the spirals is created by the width of the one or two thicknesses of the other wire(s) wound between them.

Thesprings are then polished,(to do this they are pulled gently into a cone shape), and then blued.

 

I imagine that 2 or 3 wires wound on a cylinder of the right size, then heat treated before removal would be the way to go.  Ian S C

[Sam StonesParticipant @samstones42903]

Gentlemen,

Thank you for your ideas and suggestions. I hadn’t expected this to stretch as far as it has.

That’s springs for you!

I thought it pertinent to comment upon each postings in turn. This will help to straighten my own thinking, and presumably will be of use to others who may be quietly following this thread or who pick it up later.

In his text about the clock, designer John Stevens explains the difficulties he experienced with this spring, but then he did begin from a different, perhaps more traditional point than I.

 

John,

I think you are right with regards to the hardness of the wire. It is very springy.

However, I have tried to avoid going through any heat treatment processes, even though annealing (stress relieving) may be required after winding. The original notes for this clock advocate that the (rectangular section 0.005″ x 0.02″) wire be wrapped around a brass or copper mandrel which has been carefully grooved to the requisite pitch. John Stevens then goes on to describe how to carry out fairly standard heat treatment while the wire is still wrapped around the mandrel. For me, this approach has to take a back seat, owing to my own very limited workshop circumstances.

I should point out that although this was my first attempt with this wire, it does take on a definite curl after winding, and I was able to determine the OD from the albeit poor results. There’s more about this aspect below.

Michael,

The engineering science is more great advice. Thank you for being so explicit.

Although I’m aware of the torsion component in the wire of helical springs when they are in service, I knew nothing of this (twisting) aspect during winding. It also explains why the weight I was using to apply tension to the wire was spinning slowly (after I had stopped winding). I thought that I’d perhaps caught it with my knee.

In addition to my previous interpretation of your notes, I envisage moving the pulley much further away from the mandrel. This should provide more freedom for the wire to rotate. For the same reason, I shall also use a weight with a minimal moment of inertia (eg. a small-diameter rod hung vertically), so that it will exert a lower level of inertia, offering more freedom for the wire to untwist on its own axis.

Roger,

Thanks for your posting. From my albeit poor results, I got a very curly, but screwed up bit of wire. I believe therefore that there was enough tension on the wire, and that the elastic limit was exceeded.

In bending, it is generally accepted that a beam is subjected to tensile stress on the convex side of the neutral axis. Sorry for the theory. I’m therefore more convinced that this part of the problem of uncoiling relates both to those in Michael’s notes about the wire being allowed to twist and, that the wire has already been heat treated. It is certainly very springy.

In Machinery’s Handbook, there’s a table of mandrel (arbor) sizes quoted for `music’ wire. The smallest diameter wire is the same as mine (0.008″), and the mandrel size for a 1/4″ OD spring is quoted at 0.129″. Unfortunately, the table does not extend to 3/8″ OD, the size I require. It is however, indicative of the amount of relaxation taking place. An extrapolation of the figures and my own mandrel size (guestimate) inserted, confirms that I am on the right track.

Gordon,

Thanks for your comments. Would your two brass washers be working in the same as the method used for controlling the thread tension on a sewing machine?

 

Ian,

I really appreciate your comments about winding springs. The method you describe appears to relate more to flat spirals. From my limited results and what Michael has mentioned, I could imagine getting myself and the springs into a real tangle if I wound two or more together.

Once I have succeeded in making a spring (whatever the pitch), I will rely upon the two ends of the spring being pulled apart by the clock’s own anchor points. These are just visible in my picture of the balance wheel.

I’ll now hang off doing anything else until you good people have had a chance to examine the above notes.

Regards,

Sam

[Gordon WParticipant @gordonw]

Sam -yes, I guess the washers work the same as a sewing machine. Sorry I was a bit rushed. What I have is a bar in the toolpost, with a 1/4″ hole at the end, a 6mm bolt with 2 thick brass washers clamps the spring wire, clamping load found by practice. end of wire caught under a chuck jaw/mandrel.Wind a few turns , adjust clamping load until the spring starts to form. For small diameter springs the tension can be high enough to bend the unsupported mandrel! Release the tension on the clamp washers slowly before cutting the wire !! First few turns will be scrap. Washers will need adjusting as a groove wears in.

[Ian S CParticipant @iansc]

The copper mandrel sounds like the way to go. I think the heat treatment could be done with a spirit lamp and a bit of tube for a blow pipe. Ian S C

[Sub MandrelParticipant @submandrel]

Sam,

 

As a long-standing guitarist, Johnis right. Guitarists regularly wind 0.008 top e-strings round 1/4″ diameter mandrels (tuning pegs) at high tension. After being left in position for a several years (if lucky enough not to break one doing a Pete Townshend impersonation) they uncoil to about 1/2-3/4″ diameter loose coils on unwinding.

 

I can’t see guitar springs behaving to make a usable spring without being  annealed first and then re-tempered.

Neil

[Sam StonesParticipant @samstones42903]

Hi Neil,

 

When the diameters of the 1/4″ tuning pegs and the uncoiled e-string are extrapolated, they fit neatly into the Machinery’s Handbook table.

 

I just hope I don’t have to apply too much heat to the springs, because there’s already a nice polish on the wire. Having to re-polish the wire does not appeal to me.

 

Anyway, thanks for your comments.

 

Sam

 

 

[Gordon WParticipant @gordonw]

My tame guitarist gives me old springs when he can remember, which isn’t often. I have just made a tension spring, about 3mm i/d x 20mm long, pull -off for the 2nd choke on a carb. Using the tension method ,no heat needed, and a neat spring. I’m no expert, and found this method by trial and error.

[Author]

Posts