Tesla's Magnifying Transmitter "Replications"

[Farmhand]

The magnifying transmitter has two coils on the output, one mounted above the other, having studied a few of his coils I think the trick is to have these coils wound opposite to each other like bucking coils.

Take a look at the coils in the model T buzz box COIL DOCTOR - Coil Documents and Diagrams These two output coils are wound opposite to each other The drawing on Naudins web site could be wrong The Tesla Magnifying Transmitter The L1 needs to be wound opposing each half and the coupling should be on both halves. Do this and you may get better results.

Just a suggestion.

Hmm there could be something to that, but I'm going by the actual patent of Tesla's.
In the drawing it can clearly be seen that all three coils are wound the same way in the normal winding direction.

I don't go by Naudins drawings.
I go by Nikola's.
ELECTRICAL ENERGY - Google Patents

I'm not sure it would work the same as a bucking coils, not that I've checked that out yet but I will of course. Sounds interesting.

Cheers

[Farmhand]

Hey dR, I am about to start putting together this new circuit for pulsing coils, it will double the voltage of the source or nearly double it into a big cap as a higher voltage supply for the coils to work from. It'll require at least one isolated "floating" driver I think not sure yet. Hopefully not. I'll need two PWM IC's. This circuit is still theoretical. I'll post some parts details as I go. It should be good for driving Tesla coils.

http://img812.imageshack.us/img812/7860/lightingcircuit001.jpg

Anyway while I was stripping an old project from the solderless boards to make room I thought I would show you a picture of the mosfet driver setup I was using and mainly the 2 Kohm resistor between the base and emitter on the PNP transister that drains the mosfet gate, the resister there drains the base of the two drive transistors. The driver works like this drawing below but the resistor is not drawn there.

ImageShack® - Online Photo and Video Hosting

And picture tells a thousand words.

ImageShack® - Online Photo and Video Hosting

I'm fairly sure they were working good like that, but I'll be testing them again very shortly anyway.

Here's a video of some results I got just fooling about.
Demo 2-1.wmv - YouTube

Cheers

[mbrownn]

Take a look at Teslas coils

The magnifying transmitter has two coils on the output, one mounted above the other, having studied a few of his coils I think the trick is to have these coils wound opposite to each other like bucking coils.

Take a look at the coils in the model T buzz box COIL DOCTOR - Coil Documents and Diagrams These two output coils are wound opposite to each other The drawing on Naudins web site could be wrong The Tesla Magnifying Transmitter The L1 needs to be wound opposing each half and the coupling should be on both halves. Do this and you may get better results.

Just a suggestion.

[Farmhand]

I forgot to mention, the analogy I use to visualise the fast switching and avoiding switch heating/failing is this -

I imagine a long large diameter pipe say 1 meter diameter with a section removed say 1 meter so that the pipe has two open ends facing each other over a gap then I imagine one opening having a sliding gate attached and holding water at great pressure so I have a pipe full of water with a sliding gate facing another empty pipe over a short gap the empty pipe leads downhill to a open pond, when I slide the gate open the water under pressure will shoot across the gap and down the empty pipe and fall into the pond. OK

So if was to open the slidegate (switch) slowly lots of water would spray all over the place (this is like heat) and also if I close it slowly the same happens in reverse (more heat) the heat is losses also reduces pressure (potential).

However if I open the slidegate very quickly like in nano seconds then close it again very quickly again in nano seconds what would happen is a cylinder of water would shoot across the gap and straight into the other pipe and down into the pond, with very little spray of water or (heat) / (losses)

And if the amount of time the gate is actually open is only small aswell the reduction in pressure would be less also. With enough time open there will be a certain pressure drop (voltage drop).

The effect we want is like a very rapid opening and closing of the slidegate to reduce the spray so to speak.

This effect can be seen with water and a hose in real terms. The effect is to "chop" the water or electrical current into pulses as neatly as is possible.

No spills.

Cheers

[Farmhand]

Hi dR, This circuit here is what I'm using now but with IRF740 mosfets.

N.B in this drawing the frequency adjustment is the 50K pot I did not label it well, sorry.


Uploaded with ImageShack.us

Working from a 12 or 24 volt source I haven't had any problems I did ruin a mosfet and IC chip trying to use a separate 160 volt supply but that was just being silly.

The 1n4148 diode and the PNP transistor make up a turn off sharpener which "drains" the gate charge to ground. When I use opto's with Bipolars I find I get the same problem if the some part of the signal train is not properly "drained".

As taught to me by some of the good folks around here, if the mosfet is turned fully "on" with an at least 10 volt pulse and turned fully off as quickly as possible either with a good mosfet driver IC or a made up turn off sharpener the mosfet will not get hot or fail. I have had problems using isolated fairly low voltages ruining mosfets because I did not reference the source correctly, it can get complicated.

I chose to forgo opto's because they don't work so well for me at higher frequencies.

With the circuit above I can drive the primary coil directly at frequencies up to 500 Khz or so and adjust the PW from 50% down to nothing. I can get up to almost 10 watts through the primary with 12 volts. Using the charging circuit I have now which is just a small "insertable core" inductor and a 2.5 nF capacitor. Changing the parameters of the charging circuit changes the possible magnitude of the current in the primary and therefore the power input.

What I did with Cody's help before i got my function generator was determin the approxomate resonant frequency of the secondary coil, "Terminal included", more capcitance in the terminal will lower the resonant frequency.

I used this calculator.
OLTC Calculator

Then I built a circuit that could pulse a coil around that frequency, which was the fastest circuit I built to work reliably, the one above.

Then I built the primary circuit to be resonant at a frequency that is just a tad lower that the secondary it should be as close as possible but not over, better under.

I used a RC calculator and an inductance meter to do that.

Then i hung the scope probe in the air about a foot away from the terminal and tuned to a resonant waveform, a sine wave.

The receiver needs to be also setup and connected to tune, preferably with some load, the coil "C" of the receiver the few turns one should have the same capacitance across it or in parallel with it as the transmitter does in order to make it have a resonant relationship with it's coil "A' and "B" combined, in my case it has the same 2.5 nF.

Basically you'll need a circuit that can drive the primary at the resonant frequency of the secondary, I didn't have much luck driving the primary at lower frequencies but got good results driving the primary at or just below the resonant frequency of the secondary.

You will almost certainly require a capacitor across the primary to set it's resonant frequency to either the same as the secondary or a quarter or half of it.

A magnifyer setup is a bit different in that it has "close" coupling beween coil "C" and coil "A" rather than "loose" coupling and a resonator coil "B" in series with Coil "A".

Tuned circuits require planning and tuning and to some degree adjustment as well to get everything right.

I hope some of that made some sense.

My coils are far from ideal they should all have a bigger diameter for a lower resonant frequency, but I just wound longer "B" coils in order to lower the resonant frequency to something i could reach. When you acheive resonance you'll notice it. I'm actually thinking about winding another set of rounder shorter coils, but I will put a lot of thought into them. And I won't unwind the one's I have until the new one's are online. I would like to build a set that are resonant under 200 Khz. Which will make the job a the solid state driving circuit much simpler.

To get resonance with a spark gap the primary still needs to have a cap of an appropriate value to make the primary ring in tune with the secondary unless of course the primary is naturally tuned to the secondary already.

Cheers

P.S. Thanks for jumping in to help 7imix much appreciated.

[7imix]

Another suggestion is to put a buffer or two inverters in a row before the opto to sharpen off the pulse. An op-amp might also be usable for the same purpose.

[7imix]

There are many other timer designs that are much better than the 555. For a simple, accurate, high speed one, look at the one farmhand often uses which is based off an inverter chip or NAND/NOR chip. I myself have been using an Arduino microcontroller recently to generate a clock.

There are also two circuits posted earlier in this thread, one by Farmhand and one by me, that use single IC PWM drivers to produce stable two phase clocks. One of the chips can do up to 400khz and another can do up to 1mhz.

If you have or get an arduino let me know and I'll paste the code I am currently using. It's super easy to use and extremely flexible as it is completely programmable.

[dR-Green]

Quick update...

I'm having some trouble getting this opto-isolator to work. It won't switch fully off at higher frequencies and a square wave output of the 555 becomes a triangle wave. So it looks like I'm limited to below 8kHz. Having got that part figured out, the transistor won't turn fully off so the transformer isn't transforming

I'm going to try a darlington pair to see if that helps, but if not then it looks like I'll just have to do away with the opto-isolator and buy a load of 555 chips ready for when they pop

[dR-Green]

Diagrams are not to scale [edit] The correct number of "bumps and grooves" are NOT shown, so don't count those from the image. Use only the measurements. Apparently my maths are 1-5mm out when adding everything together because I'm measuring the coil itself so I can't get the ruler exactly square up to edges and some wood is lost from sanding etc. I recommend cutting the base to size and make the rest fit around it.

The 33mm diameter ring represents where the 13.3cm length piece of bendy MDF will come to when it's put up to the outer 30cm edge of the base. The 140mm ring represents where the 7.8cm pieces of bendy MDF end over the same diameter circle. This is for knowing where to glue the pieces on to the base so everything lines up, and the wires will form a nice circle.

[dR-Green]

dr-Green, thanks a lot for the specs. I wonder where I can find some of that bendy stuff.

I wouldn't hook a device with a spark gap directly up to an oscilloscope... you're liable to fry it.

I think you should be able to get the bendy MDF from a local DIY store. Another name for it is Neatform. Check in the MDF/wood section. It comes in the same size panels as the other sheets of MDF in case it doesn't look like you think it does. Mine was 1220 x 607 x 6mm

Innovative Panel Products by Neat Concepts

NEATCONCEPTS Introduction

If you can't get that then you can just cut notches into strips of wood manually. I went for the bendy MDF because it seemed like the job was already done, but it also has its disadvantages in that you have to be very careful working with it because it's not very strong. Sanding it for example puts pressure on it in ways it's not designed to handle, so it will break easily. Or rather you'll be left with a bare strip of MDF, but no notches and bumps because they pop right off. So I suppose it depends which you want to spend your time on, measuring and cutting all the notches, or the delicate work.

My personal experience from it involved cutting and sanding 80 support blocks, and also 32 delicate bits of this bendy MDF... So I was pleased when all that was finished to say the least

In that respect, I'll do a test to see if wood near the coil affects the performance, and if it doesn't then it will save you the time of rising it off the main base. These won't be my last spiral coils anyway so if there are better or more efficient ways to do it then I intend to find it

I have the scope connected to the transistor collector. Safely away from the HV side I haven't paid too much attention to what's going on there yet, I just noticed there's not a constant frequency. But I also know that with wireless experiments the scope will respond to things without being connected to anything, so I'll have to look closer.

Oh, and I almost forgot, last night I connected a CFL to the coil output with 1 wire. 2 wires seems to short it out. But with the 1 wire going in to the CFL, there's an even higher voltage coming out the other wire and it will ark about 3 times the distance of my spark gap.

[7imix]

dr-Green, thanks a lot for the specs. I wonder where I can find some of that bendy stuff.

I wouldn't hook a device with a spark gap directly up to an oscilloscope... you're liable to fry it.

[dR-Green]

It might get complex to explain, I'll have to draw diagrams to make it clear and there's a modification/improvement I've thought of since which would make what is a square base uneven, but I'll explain what I can.

Here are the specs of the Bendy MDF. The grooves are approx 2mm wide with 4mm between them. That is, the wire sits in a 2mm groove, and there is 4mm between each turn. Each section that the wire sits in is cut 1cm wide. I'll make a diagram to show all the pieces with measurements etc.

The base is 30cm x 30cm. However, here is the modification I would suggest: 30cm x 30.6cm.

The reason for this is to accommodate one extra "groove and bump" of Bendy MDF on one of the supports, to allow the start of the primary winding to be on the outer dimension.

The way I currently have it wound is that because it's an even circle, or 30cm x 30cm, the start of the winding has to come from somewhere, so it's taking up winding space, when it should be coming from the outside INTO where the winding should be.

Am I making sense? For that, one of the supports needs to be on the outer dimension and directing it into the "next turn" if you will, where that turn SHOULD be starting if there were more turns on the outside.

In other words, I hadn't counted on how to start the winding before I started building the base. Now I can see that the primary isn't an even circle because there's nowhere else to begin the winding. So, I'd add the extra 0.6cm on the base

Anyway, the primary is 17 SWG, secondary 24 SWG. 29 grams each, give or take as close as I could approximate how much wire would be leading in and out of the actual winding. The actual windings are probably within 0.5g of each other. Primary is approx 1.5 meters in length, and the secondary approx 13.5 meters.

The way I did that was to put 2 turns of 17 SG primary on the base, wind it back up again and weigh it, then weight the same amount of 24 SWG wire. (You may remember first I used 26 SWG but it was too long to fit on the base, so I had to use 24 SWG. There is space left over, if it was available in the local shop I'd probably see if 25 SWG could fit on there).

I'll have to make a quick dash here because there's a friend coming over, but thanks for the info and I'll get back with more details and diagrams soon. I suppose a better quality camera would also help so I'll get on that too

[Farmhand]

Hi dR, I have a space problem too, all this stuff takes up a fair bit of room and it can't be left just anywhere for people to poke at.

The thing with the SG sounds good, maybe the Tesla coil made the ground more negative than the battery negative, that will charge the battery, to charge a battery you can add charge to the top "positive" or take charge from the bottom (negative) and it will charge. Usually we reference the negative to "ground" and add to the top to charge but we can reference the positive and take from the bottom to charge also, "Tesla Switch" negative load kinda way. In this case the ground would be the load taking potential from the negative and pulling it down.

Be prepared for wierd things to happen that are difficult to explain. I hesitate to say some of the things I see and hesitate even more to offer an explanation. Good work, I am seriously considering building a spiral coil or two like your's. I can sweep the coils with the function generator to find the resonant frequency. Could you take the time to make some measurements of the specs of the spiral coils and post them.

Primary, ouside and inside diameter, wire size, turn count and inter turn spacing. Same for secondary.

Cheers

[dR-Green]

A quick side note... I just activated the circuit, and forgot I had my SG earthed on the same line as the Tesla coil. The SG sped up and having just checked the voltage before getting started, the battery had risen from 14.17v to 14.70v. So I thought that's not good, and disconnected the SG from earth. [edit] Which is a 36Ah battery and takes a much longer time for the 2 circuit 200mA output SG to get it to that level, be it a fuzzy charge or whatever. I'm only looking at the difference in voltage here.

A check a couple of minutes later read 14.60v which confirmed my suspicion and I didn't imagine the numbers I thought I saw earlier. I swapped over to the spiral coil and left it going for a minute or two. For no apparent reason I checked the battery voltage again, this time 14.68v, with no earth connection between the devices. To be sure, I left the voltmeter on and activated the spiral coil. The battery's voltage does indeed start to rise and went to 14.70v in a couple of seconds before I came to report it here.

Haven't repeated this test with the helical coil yet. The SG in the pictures is to the right of the helical coils on my desk to get an idea of the positioning. I'm using the spiral coil on a seat because there's no space on the desk, so it's even further away from the SG. The "termination" so far is just a neon bulb held on by crocodile clip.

[dR-Green]

If you put an optocoupler or an optoisolator between the 555 output and whatever you are driving with the 555 you will avoid frying them.

Do you have an oscilloscope?

Physical relays are good for extreme power switching, but they waste a lot of energy and they limit the frequency of switching. I suggest using a rotary commutator if you want to try physical switches instead of transistors. Connect the shaft of a regular DC motor to the shaft of one from which you have removed the windings. Drive the regular DC motor normally to get the desired spin rate, and use the commutator from the second DC motor which has had the windings removed to make and break the circuit containing the primary.

That's a good idea with the motor, thanks for the suggestion. I'd like to stay away from moving physical parts if I can though, but it's good to have the options and I hadn't thought of using the commutators, nice idea. I've got plenty of opto couplers here so I'll give that a try I do have a scope. It's getting a bit crazy when I'm using a spark gap though, the frequency is jumping all over the place. Thanks for your suggestions

Either way, initial tests of the flat spiral coil (only one used so far) are more impressive than the helical coil. I have no way to measure accurately, but based on the fluorescent bulb I'm guessing I'm getting about double the output from the spiral coils. From the pictures you can see the difference in amount of wire/copper used for each coil, the flat spiral coil looking quite bare compared to the helical coil, but the spiral coil gives a better output already with exactly the same input, so that's interesting. Input current draw also goes down when I load the "output"... Although the coil setup it not configured properly for this purpose yet, and by "output" here I mean the centre of the coil, whereas Eric Dollard in his videos is using the outside start of the coil as the output, the bit that's supposed to go to ground. But this was just a quick test of the coil and already better results as I say, so very interesting I'll take some pics to compare the lighting of the fluorescent and post them here, I might have to wait until it gets a bit darker though because there's not that much to see from the helical coil at this point and I'm using a webcam to take the pics so it's not too good at capturing the things I want it to capture.