Much has been said about TC efficiency over the years, and a certain amount of controversy still exists. This is an area that I've been investigating for some years now and some of my findings have been surprising and thought-provoking. I'll present some ideas in a question and answer format.
1) Wire thickness, secondary proportion issues, etc.
Ques. I've heard that even for a small TC project, it's important to use thick wire on the primary and secondary, to seal the PVC secondary form before winding, to use robust connections, etc. Is this really important?
Ans. You are correct to question those statements. In reality none of those things will have a noticeable effect on the spark length of a small Tesla coil. In fact, contrary to popular belief, the use of thinner wire and more turns for both the primary and the secondary will result in longer sparks.
Ques. Thinner wire... won't that increase the wire losses?
Ans. Yes, it will, but the thinner wire lets more turns be used on the primary which will increase the primary surge impedance thereby reducing the gap losses. The benefits of the reduced gap losses outweighs the negative effects of the thin wire, so the overall sparks will be longer, brighter and better.
Ques. And this really makes a noticeable difference?
Ans. In a comparison I did, the sparks were 10% longer for the same power input using thinner wire.
Ques. I looked at a lot of websites, and no one talks about this method.
Ans. The only other person that I know of who tried this is Richie Burnett of the UK, he also gained 10% spark length when he used more turns of a thinner wire on his secondary.
Ques. OK, but why does a higher primary surge impedance reduce the gap losses?
Ans. Gap losses are proportional to the primary tank current, a higher impedance results in lower currents and lower losses.
Ques. You say thin wire is good. How thin should it be? Should a huge coil use real thin wire?
Ans. The wire should be proportional to the size of the coil. I like to use about 1600 turns on the secondary, and about 24 to 34 turns on the primary. Realize though that it is not the thin wire or the many turns that is giving the benefit, rather it is the resulting many turns in the primary and higher primary inductance and Zsurge that gives the benefit.
Ques. Do the proportions of the secondary matter?
Ans. It seems that wider secondaries are better, because they give a greater inductance which also helps to reduce gap losses, by requiring the use of more primary turns as explained above. A proportion of 1:4 or 1:5 is probably quite good.
However, excellent results can be obtained from narrower secondaries, esp if one wants to build a compact coil and is willing to sacrifice some efficiency.
Ques: Well, suppose I just build a huge secondary so I can get a high inductance but still use fewer turns of thicker wire?
Ans: This might work, but often it is not desired to have short wimpy sparks coming off of a huge TC. Generally the size of the coil is matched to the size of the sparks to keep a sense of proportions. Also, if the secondary is too large, it will shade too much of the toroid's area, and decrease its effectiveness.
Ques: Just how much loss is there in the spark gap?
Ans: There's some controversy here, but it may be 15% to 30% or more depending on the coil design.
Ques: Is it possible that the lower frequency that results from using more turns or a wider secondary also helps the spark length?
Ans: Yes, it is possible that the lower frequency also helps by reducing losses or helping the quenching. If the lower frequency is an important factor, then larger more powerful coils may benefit less than smaller ones from the use of thiner wire with many turns, because the frequency of large coils tends to be low anyway, and may be low *enough*. I still think the larger coils should benefit from the reduced gap losses when using more turns of thinner wire though.