Nikola Telsa Coil

[chemelec]

Yep, the laws of physics are still holding here I had to demagnetise the nails before using this rig - there was a strong pull from both without any current flowing. Again the null is obtained when the electromagnets are equidistant from the compass. A few less turns but a greater flux than before but still the same result - the myth is still busted!

I won't be running this again with a battery in case anyone thinks that would make a difference. It probably would - but only because the current would fall as time went on. Maybe this is also why people think the Tesla winding makes a more powerful electromagnet - if they used it first they might indeed be able to pick up more paperclips

Wind the coils on Ferrite Cores!

[Tepco]

http://www.4shared.com/photo/JPB6JkwG/DSCF9424.html

Some (long) time ago, i have built two identical coils, one normal flat spiral and another in "Tesla" bifilar configuration, just for curiosity and for comparison tests. Coils made of 0.75mm PVC insulated litz wire, 13.5cm ID, 24.5cm OD, 26 or 2x13 turns on same former. I compared coils under identical conditions (except Rdump) on same machine. And "Tesla" configuration was always slower, by some say 30% compared to the normal one, and somewhat less sensitive too. As expected, larger self-capacitance. Unfortunately, "normal" coil is later graphite shielded and epoxyed, so i cannot post exact waveforms etc, direct comparison now is irrelevant unless i made another one. But, no magic happens whit bifilar configuration,regular flat spiral is tested better. Bifilar coils are good for other things, but not for PI monocoils. After all, Tesla is not became famous as metal detector designer.Idea: BUSTED, dont waste time on it.

[mikebg]

Tepco, explain please what is the advantage of your cascode TX stage!

[AdrianM]

I compared coils under identical conditions (except Rdump) on same machine. And "Tesla" configuration was always slower, by some say 30% compared to the normal one, and somewhat less sensitive too. As expected, larger self-capacitance. Unfortunately, "normal" coil is later graphite shielded and epoxyed, so i cannot post exact waveforms etc,

Yes, from my testing the winding topology appears to multiply the self capacitance by close to a factor of three (everything else being equal). However, we should not overlook other advantages as posted buy moodz The coil provides a differential signal for the induced eddy current and common mode for environmental EM which greatly increases signal to noise. It is also claimed that the capacitive coupling to ground effect is also eliminated (it also being common mode - but I don't quite see this).

Another feature is that the TX current can be switched into one half of the winding (having approx. 1/4 the total inductance of the coil) and auto-transformer action causes the field to collapse in both halves (giving twice the flyback voltage - only half of which appears across the switch). Then the full coil is available to convert the target field back into a differntial signal.

Eliminating the need for a shield makes some additional gains, and the losses due to extra capacitance can be minimised by careful choice of materials and construction. Indeed I think the capacitance can be brought down to the sort of values people are quite happy with in practical conventionally wound coils - just by using the kind of insulated wire rated for higher voltages.

This latter point I would imagine might be especially of interest to people playing around with cascode switching

mikebg, did you mean to post your question in this thread?

[Tepco]

Tepco, explain please what is the advantage of your cascode TX stage!

Advantage of this "cascode" circuit is basically to provide higher breakdown voltage switch. Usually, it is hard to find MOSFET whit over 800-900V rating, and typically those devices have high Rds on resistance and large output capacitance. In this way, high voltage bipolar is used, 1.5KV or more capable, and lower Cout. Cascode is just easiest way to drive it.

And advantage of this is minimizing, or eliminating breakdown condition (pulse width and coil dependent), thus minimizing nonlinear part of circuit behavior, and this will minimize ringdown time and enable faster sampling. (up to 30-40% faster, depending upon conditions) Try this trick and compare results, there will be improvement on almost any machine. Except penalty for higher current consumption, due to high base current requirement for upper NPN (almost comparable to coil current on some detectors).

And of course, be careful not to blow up something. This will produce 1.5-2KV microsecond peaks, so coil insulation, input and dump resistors (use multiple), protection diodes (1N5819 instead 4148 ) etc have to be considered.

[mikebg]

Alset coil

Alset coil - the opposite of Tesla coil

Unlike the Tesla coil having a decreased resonance frequency due to increased stray capacitance, the Alset search coil for PI (see below attached figure) has
increased resonance frequency. This is due not only by reduced stray capacitance, but by reduced self-inductance.

Alset coil can be regarded as a flat spiral coil, which are missing several turns in the middle of spiral. In fact, it is made of two flat spiral windings (coils) connected in series. Their connection forms center tap to which we can connect two damping resistors. For critical damping they have greater resistance than a resistor damping the whole coil.

The coefficient of coupling "k" between two windings is reduced due to the increased distance and different diameters. This leads to reduced self-inductance of search coil.

[AdrianM]

Alset as in Tesla spelt backwards? Very funny
But is it really a goal to reduce self-inductance in a PI coil? I would have thought it would reduce the sensitivity. There is no inherent common mode rejection with this winding either, and this is the primary advantage of the Tesla coil from my perspective. Like I said above, I think it's quite easy to nail the capacitance problem (while at the same time creating a coil that will take a 3kV flyback!)

[Tepco]

Alset coil - the opposite of Tesla coil

Unlike the Tesla coil having a decreased resonance frequency due to increased stray capacitance, the Alset search coil for PI (see below attached figure) has
increased resonance frequency. This is due not only by reduced stray capacitance, but by reduced self-inductance.

Alset coil can be regarded as a flat spiral coil, which are missing several turns in the middle of spiral. In fact, it is made of two flat spiral windings (coils) connected in series. Their connection forms center tap to which we can connect two damping resistors. For critical damping they have greater resistance than a resistor damping the whole coil.

The coefficient of coupling "k" between two windings is reduced due to the increased distance and different diameters. This leads to reduced self-inductance of search coil.

AlseT coil, what a name! But this is exactly the point. If someone stretch this idea further, say 4 or 8 sections whit its own dumping resistor...and end up whit resistor on every winding, this will create artificial transmission line of certain impedance. Distributed L,R and C. And i suppose this will be fastest. Passive dumping? Remotely associate me to Ellen Ott (US6456079 and others) although no mention of transmission lines there, just passive dumping.

Really, if anyone here tried something similar, and what results can be obtained?

(I did, to some degree, using highly conductive coatings, even samples of conductive rubber etc, but due to lack of proper materials, testing and time i cannot claim anything. What is needed for test is printed circuit flat coil, some math and about 30 SMD resistors, all of different value.)

Anyway, nice idea, and even better name!

[chemelec]

On the Flat Wound Tesla Coils I make, I am Not able to find ANY SELF RESONANCE, at Any Frequency.

[AdrianM]

On the Flat Wound Tesla Coils I make, I am Not able to find ANY SELF RESONANCE, at Any Frequency.

Yes, I saw you posting that before and I thought "how very odd"... I've made several now and they all resonate like heck! I mean, if one is within four feet of a working PI there's almost a volt across the O/C Tesla coil. I actually have to remember to put them out of the room while I'm testing other coils on a driver - otherwise the distant ringing even shows up on the RX waveform

I've also found the resonant frequency very easily feeding the coil through a 1M resistor and using a 1pF wire twist to couple the scope. Resonance is usually down a bit lower than a conventional mono coil - the most useless one I made was 242kHz, the best 500kHz but boy oh boy do you ever get power coupling between two identical coils when you drive one at that frequency (whatever it is). This, I think, is what attracted Tesla to this invention.

IIRC there is only one resonance, nothing crops up at odd or even harmonics (although I only remember looking half-heartedly). If that resonance doesn't leap out and hit you in the eye I would suspect something was wrong with your winding connections?

As a spin-off, I'm making use of a pair of my "failed" PI 12" coils to wirelessly trickle charge my "Sunday best" car when stored in the garage. The power RX coil goes under the plastic front bumper about 10" above the TX coil on the ground. Perfect alignment isn't necessary but I'm putting a charging LED on the dash to tell me when to pull-up (useful in its own rights). Still tweaking, but looking good so far.

[AdrianM]

Here's a quick photo demonstration of resonant power coupling - just an LED across the coil with a pretty big air-gap. The coil on the floor is way off resonance as well (450kHz but being driven with 250kHz - the resonance of the RX coil) matched coils would probably destroy the LED at this range! (those are 12" diameter coils to give you an idea of scale).