How do they wind coils with ridiculous gages like 56AWG?

[Alex Eisenhut]

I'm trying to figure out one step in the manufacturing of floppy drive R/W heads. At least the older ones I am looking at.

I know no one makes those, but someone is still making ridiculously small and fine coils for teeny-tiny earbuds, etc

I mean just in general. The wire is so fine just yanking it a bit too much it breaks. Soldering it is another mystery, solder dissolves copper. I managed with much bigger headphone wire but even that was tough.

How do you wind something this fine let's say on a 1mm diameter coil? Or remove the enamel on the ends?

What terms should I google for?

[amyk]

Look for "fine wire winding" machinery.

https://azonicproductsinc.com/products/wire-tensioning-devices/model-3535-ultrafine-tensioner-awg-45-to-56-wire-52533641

https://audemars.com/micro-coils-manufacturing/

[Alex Eisenhut]

Thanks. I know this guy managed 60 years ago
https://pasadenahistory.org/collections/micromotor/
so it's not impossible...  But there isn't much detail about how he did it.

[Ian.M]

As long as the winder is designed with the known tensile strength of the wire in mind, I don't see any problems.  Stripping the wire can be done chemically, and soldering it just needs a solder alloy with a composition that has a near-saturation copper concentration in it so it has very little tendency to dissolve the wire.

[Circlotron]

A related issue - I used to work in a SMPS lab where we had reels of various diameter winding wire. To make sure of what we chose when winding up a prototype we would use a pair of calipers to measure the wire diameter and then subtract 0.1mm to allow for the thickness of the insulation coating. This worked great for wire of 0.25mm upward, but what about say 56AWG like the OP mentions? The insulation thickness would account for a much greater percentage of the overall diameter, and therefore going to a thinner wire would have less and less usefulness. Are there different insulation thicknesses on very fine wires?

[Alex Eisenhut]

As long as the winder is designed with the known tensile strength of the wire in mind, I don't see any problems.  Stripping the wire can be done chemically, and soldering it just needs a solder alloy with a composition that has a near-saturation copper concentration in it so it has very little tendency to dissolve the wire.

For sure, if someone showed me the tools I can learn. I'm just fixated right now on how these floppy heads were made. Because in my ... let's say enthusiastic ignorance, I feel with enough knowledge I can take the defective coils out of the head (it's just goop in there, it's not potted) and ... somehow, magically get a new coil in there. No I don't know how to get the coil off from a closed loop of ferrite... one step at a time.

A related issue - I used to work in a SMPS lab where we had reels of various diameter winding wire. To make sure of what we chose when winding up a prototype we would use a pair of calipers to measure the wire diameter and then subtract 0.1mm to allow for the thickness of the insulation coating. This worked great for wire of 0.25mm upward, but what about say 56AWG like the OP mentions? The insulation thickness would account for a much greater percentage of the overall diameter, and therefore going to a thinner wire would have less and less usefulness. Are there different insulation thicknesses on very fine wires?

I found a video of a guy who bought a spool of AWG56 and I think the micrometer squashes the wire and you can't get a reliable reading. In any case the manufacturer of the fine wire can't even measure it, the dimensional chart says dimensions are "theoretical". 

https://youtu.be/pn8ufOdbJ5M

[thermistor-guy]

I remember meeting the CEO of a UK company that made 2-to-4 wire transformers (called "hybrids").
These were used in telephone line interfaces. The company had problems using fine enough wire.

I advised him to check out Japanese manufacturers of low-power 48 V telecom relays. At the time,
not every relay OEM offered these, because the wire needed was too thin, and the normal manufacturing
process would result in too low a yield. Judging by their pricing at the time, Japanese OEMs like Takamisawa
had mastered the necessary art, and we used these by the boatload.

[iMo]

FYI - 2 weeks back I was repairing an 1MegaOhm wire-wound resistor, the wire invisible with a naked eye - see below the shot with a piece of the wire wound on a matchstick (the control line is 0.076mm wide).
I spent 3 hours with it and the resistor works again (but it is by 8kOhm lower)

[MisterHeadache]

Constant control on wire tension is a must.  My experience is limited to no smaller than 42 gauge, which we wound onto sensors and coil bobbins for automotive sensors using Marsilli industrial machines.  The terminations were wound into a structure called a 'skein', which essentially is the wire wrapped around itself several times to make a blob thick enough for subsequent soldering or resistance welding.  But even so, the transition from the skein to the single strand was always a weak point and prone to breaking.

[coppercone2]

for soldering wire of small gauge you need a solder alloy that has the wire material in it so there is less tendency to dissolve.

I.e. copper bearing solder. It will protect plating too. Same idea as silver bearing solder for tektronix metalized silver ceramics.

As to what % you need, I have no idea.

My procedure would be this:
1) clean wire (not sure how, maybe molten KOH, wear goggles + safety shield + long gloves + apron .. nasty work)
2) strain relief the wire with a silicone or q-dope drop after its positioned correctly for the solder joint
3) allow for full cure
4) perform solder joint with correct metal bearing solder and flux
4) don't touch nothing

You don't want to be bending or positioning the wire or anything during soldering IMO. It needs to be laid correctly then just submerged. And I would say you should make sure not to over heat it, so the rosin does not burn too much, so you can flow clean it, and possibly use a brush (not a tissue) to agitate it a little.

[DavidAlfa]

Using a Sn-Cu alloy should help a bit, as the tin has already "relieved" some of its affinity for copper.

[CatalinaWOW]

Once you have equipped and trained yourself to work at these gauges welding might be easier than soldering.  Thing of it like wire bonding on ICs

[Alex Eisenhut]

I've contacted the Swiss company. I'm pretty sure there's not much that can be done to revive these dead old heads, but if this company can wind me a few test coils for fun, who knows?

Thanks all!

[Ian.M]

See https://www.eevblog.com/forum/projects/soldering-ultrafine-wire/
Apparently Sn60/Pb/Cu2 solder will reduce the rate that 54AWG copper wire immersed in molten solder dissolves at, sufficiently that it will last over a minute, so you can expect negligible wire thickness loss at the surface of a normally made solder joint.

[HighVoltage]

The smallest wire I have wound with a Swiss made winding machine was 0.008 mm.
It is more a question of good quality tentioners.

If all is set correctly this is not a challenge anymore.

[Bud]

Now, how do they make such thin wires 

[Alex Eisenhut]

Now, how do they make such thin wires 

Very carefully.

[coppercone2]

you need swiss people to make a smaller version

[MathWizard]

No wonder I've found it hard to solder headphone wires. Not only is there something like cotton wound in for strength. And the wire is enamelled and burns onto the wire so it's hard to wet. And now I see the wire also could be dissolving a bit.

[Ian.M]

https://www.eevblog.com/forum/beginners/tips-for-soldering-in-ear-earphone-wires/

[Edwin G. Pettis]

Copper being a relatively 'soft' element is limited to how small a wire can be drawn out to, 56 AWG is approaching the physical limit of drawing it out.  However, the 1 Megohm resistor mentioned was not made with cooper wire, at the very least it was Manganin and more likely an Evanohm alloy.  Evanohm has been drawn down to as small as 0,000,4" (0,010,16mm) for many years, I have wound resistors with 0,000,7" (0,017,78mm) wire and it is not easy.  The wire will break at the slightest jolt to it and it is a lot tougher than copper or copper alloy.  Yes the wire is difficult to 'see', the lighting must be aimed exactly in order to 'see' the wire, preferably with a darker background for contrast.  There has been several manufacturers in the US over the years that made this fine wire, no need for the Swiss' fine capabilities though.

Depending on the size of the bobbin, the 1 Meg resistor was wound with wire <0,000,9" and some of the smaller ones used 0,000,4" wire.  Note that the wire sizes quoted here are the actual diameter of the resistance wire and not the finished size with enamel coating on it.

I do not know all of the finer details of making very fine wire but diamond dies are used with progressively smaller holes in them.  As a nod to the difficulty of drawing out these fine wires, the last quote I got on 0,000,7" Evanohm wire was $18,000/pound and that was a few years back!

[Alex Eisenhut]

It looks like the wire in the floppy head is nowhere near 56AWG, looks more like 39AWG AFAICT. But winding a coil is the least of the problems...
Thanks all, I guess I'll try to find local coil winders if they exist. Not that I want anything right now but in case I make a breakthrough in how to swap coils in a head.
I know it's likely not possible but whatever.