# A coil I built has divergent measurements respective to wire resistance and AWG size

I'm not exactly into building precision coils, but I decided to make one for a subwoofer voice coil.

I used what was supposed to be AWG 28 wire, it completely fit the diameter. I wound it tightly over a cylindrical metal piece and it took 15 turns until the length of the coil was 5 mm.

So I went on and built a 25 mm tall, 40 mm diameter voice coil with two layers (orthocyclic winding) and it has 78 turns on each layer (completely compatible with AWG 28). The legs are 10 cm each.

But when I measured the resistance with my multimeter it read 5.8 Ω. I was expecting 4.2 Ω.

Noting that I'm already accounting for the resistance of the multimeter leads and the electrical connection with the legs, I measured the resistance between two points in each leg and they both read 0.1 Ω (the whole coil measured 6 Ω, hence 5.8 Ω actual resistance).

I also tested my multimeter with resistors I have. A 100 Ω resistor measured 101.1 Ω, a 22 resistor measured 22.1 Ω and a 4.7 Ω resistor measured 5.1 Ω (all of that without subtracting the resistance of the leads this time, and all of them being 5% tolerance resistors).

It seems the multimeter is reasonably calibrated, values from 100 to 4.7 Ω show very little discrepancy.

What could be happening here?

I don't see any damage that could suggest fractures along the wire.

Could this actually be AWG 29 wire with a thick insulation?

Or a cheap wire with poor copper content?

Ps.: I wound it in an aluminium former which I then removed from it. Now there's only the copper wire with a tiny bit of resin that probably doesn't amount to 0.2 grams. I weighed it on a scale and it read 11 grams. With the error it could be 10 or 12, but an AWG 28 wire with 19.8 meters of length and a cross-section area of 0.08 mm² should weigh a bit over 14 grams. I tested my scale with water and a syringe and it seems the errors are always inside +- 1 gram (as expected.) This somewhy fits better with the weight AWG 29 wire would have with the same total length of wire.

• How long was your wire - you haven't demonstrated that you know how much wire you used. Apr 24 '20 at 8:32
• 78 turns per layer. How many layers? Apr 24 '20 at 8:38
• @Transistor Two layers (like most voice coils). Seems like I accidentally deleted the word "two" from the text. This gives 19.8 meters of wire. Apr 24 '20 at 8:39
• Noting that even if I added 1 meter more and rounded it to 21 meters this still doesn't give anything near 5.8 ohms. Apr 24 '20 at 8:41
• I've confirmed your calculations, allowed for increased diameter on first and second layer and tails and get 4.268 $\Omega$. Apr 24 '20 at 8:44

The above (modified) picture came from this wiki site on magnet wire. it suggests that the overall diameter is made up from about 8% insulation. The insulation is 0.015 mm overall. That site also states this: -

Breakdown voltage depends on the thickness of the covering, which can be of 3 types: Grade 1, Grade 2 and Grade 3. Higher grades have thicker insulation and thus higher breakdown voltages

So, do you know how thick that insulation is?

If all of this is enamel and this is AWG 29 wire, than the enamel layer is 0.016mm thick

OK, so assuming it is insulation grade 1, that explains half the thickness of varnish but, varnishes come in all sizes (as hinted above) and who is to say that the varnish on your wire isn't a little thicker than the "typical" grade 1 standard? I mean, where did the wire come from? Does it have any provenance?

• Well, it has a provenance. It was bought by me 15 years ago to replace a winding in a motor of mine. The labels, if there has even been any, disappeared from the reel, and I remember it was the only remaining reel of that AWG when I requested it, so I didn't have choices about insulation. And it wasn't a full reel, but the rest that remained in the reel, it was sold by the kilo. I remembered it as being AWG 28 and the measure I did seemed to confirm it was AWG 28, but now I remembered you're supposed to burn the wire before measuring to remove the enamel. Apr 25 '20 at 2:41
• @user2934303 if you are done with this question now you should accept the answer as per site etiquette. Dec 14 '20 at 11:40