4.5 MHz coil / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

4.5 MHz coil

1 2 3 next

Move Thread

LAN_403

WaveRider

Mon Feb 13 2006, 12:51PM

Registered Member #29 Joined: Fri Feb 03 2006, 09:00AM
Location: Hasselt, Belgium
Posts: 500

Hi all..
This is a continuation of the 10 MHz coil thread from the old board. I thought I would give an update about the modifications that I made over the weekend.

Basically, spurious radiation was a BIG problem. Disruption of various digital devices was an indication of this. So, over the weekend, I build a cage to put my coil in..

Notice that all joints are soldered:

This has the double effect of keeping the resonator Q high as well as preventing unwanted radiation. In fact, the arc strikes at much lower currents than when the coil was without a shield. The drive current peaks at below 4 amperes for 30 VDC from the bench power supply.

When the coil was "open", the current draw was nearly 7 amperes. My guess is that most of this "missing " power disappeared into the ether as radiation as a result of coupling into the power-supply leads. With this "cage", the radiation problem has disappeared and the transistors run much cooler and the power supply seems to be well decoupled.

Probing with a wire loop connected to my CRO indicates strong fields near the enclosure, but 1 or 2 meters away, they are barely detectable. The open top does not permit much radiation and allows access to the coil..

Now I can play all I want without angry radio operators beating on my door!

Cheers!

Marko

Mon Feb 13 2006, 01:02PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

If you add just small topload and remove faraday cage capacitie coupling to neon bulbs around would be great..
im wondering from how far could it light up neon lamps..

How did you construct the driver, do you have a schematic??

WaveRider

Mon Feb 13 2006, 01:14PM

Registered Member #29 Joined: Fri Feb 03 2006, 09:00AM
Location: Hasselt, Belgium
Posts: 500

The faraday cage is as much for safety as for legal use. Without the shielding, tens of watts of RF power were likely coupled into the power suppy leads, which acted as an antenna. This caused havoc with digital clocks and embedded u-controllers in household items (this says alot about the EMC of said devices

)

The Faraday cage enhances the desirable Tesla-coil properties (i.e. arc power) because rediation is greatly reduced. And it lights fluorescent tubes very nicely!

A rough schematic is in the old article. When I get around to it, I will draw a proper schematic...

Dr. Shark

Mon Feb 13 2006, 09:21PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

I did follow the old thread and find it very intersting, and its great to see that you continue working at it. This might be the furure for small, simple SS coils (maybe SSSSTC?). As you may have noticed I am also very interested in this stuff, so when I have time I will definitely try this out!

About your faraday cage: I am very surprised that this actually reduces radiation losses, as the wire will most likely absorb the radiation, but not prevent it from being radiated in the first place. If it does give an improvement in streamer length I would be very surprised, but then it might act as an additional capacitance and load down the coil somewhat, which could mean that primary and secodary were out of tune before.

Marko

Mon Feb 13 2006, 09:56PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Tesla coil of any kind is based upon its resonance and earth is just a plate of capacitor in that LC circuit.
In normal conditions earth has relatively higher resistance than iron, so that kind of ''plate'' is more lossy than cage (cage doesnt apsorb energy, it helps LC circuit to keep it)

Ofcourse if it arcs to cage its like we have dielectric breech in LC and this is considered as loss...

If it is removed obivously coil would be far harder to tune and more lossy, as moving people, objects, neon bulbs and resistances aroung affects its Q factor and reconant frequency...

WaveRider

Tue Feb 14 2006, 10:03AM

Registered Member #29 Joined: Fri Feb 03 2006, 09:00AM
Location: Hasselt, Belgium
Posts: 500

The wire of the cage will absorb some of the power, but most will just be reflected. In fact, the coil behaves like a typical enclosed helical resonator. With no arcing (at low power) , I get an amazingly clear 4.5 MHz tone (sampling away from the primary coils...where harmonics are present). Moreover, by reducing inadvertent coupling into external objects, I get a bigger spark for less DC supply current.

It is interesting to note that the primary has 2 turns and the secondary has 86 turns. Breakout occurs when the DC supply is around 17VDC. A 43:1 turns ratio gives about 730 volts (assuming perfect coupling). If we assume a resonator Q of about 10 with the arc, we expect about 7300V at the top of the coil...right order of magnitude, I guess, given that air ionisation is easier at high frequencies.

A nagging question remains: Why is the resonant frequency 4.5MHz for a 25 m long winding? 1/4 wave at 4.5MHz is 16.67m. I expected the resonance to fall somewhat below 3MHz... Need to give this some thought!

Marko

Tue Feb 14 2006, 01:56PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Cage reflects only EM field but not electic, just acts like low resistance way-to ground, with smaller losses and higher Q of resonator.
If that electric (magnetic is not so bad actually) goes wild out of cage anytging grounded becomes part of 'capacitor plate', wic can result in ultimate deats of electrical equipment even if they are far away from arcs.

Also EM field if released would cause havoc in radio equipment but nothing more, but this probably wouldnt be noticed if there is no cage.

Maybe its best to understand ground (or cage) like other plate of capacitor.

Resonant frequency is determined by LC circuit, not coil itself and wire length.
Angular frequency is root of LC^-1, inductance is that of coil and capacitance is interturn, depends on surface and wire thickness, and capacitance to earth (thats why it works better and more stable within cage).
Just length of wire means nothing here, frequency would be lower if you just used wire with thinner insulation, more of it or added topload for more capacitance to ground.

WaveRider

Tue Feb 14 2006, 02:44PM

Registered Member #29 Joined: Fri Feb 03 2006, 09:00AM
Location: Hasselt, Belgium
Posts: 500

Hmm.. The interference with relatively distant devices was strong when there was no shield. This is indicative of radiation/power-line coupling.

Cage reflects only EM field but not electic, just acts like low resistance way-to ground, with smaller losses and higher Q of resonator.

The shield blocks E-fields extremely well (even tho' E and H fields are inseperable here)! The concept of "ground" in the traditional sense has little meaning at high frequency. The water pipes and electrical system in my apartment building make better antennas at 4.5 MHz than being useful as a fictious "ground".

Resonant frequency is determined by LC circuit, not coil itself and wire length.

Just length of wire means nothing here, frequency would be lower if you just used wire with thinner insulation, more of it or added topload for more capacitance to ground.

Horse-hockey! The length of the wire is everything! Just how would you define C? L? I have verified that the current at the base of the coil is high and is nearly zero near the top. How do you explain this with a lumped LC model? There is no "topload" in the traditional Tesla Coil sense. The resonance is indeed a "self-resonance" of the secondary coil. This is a distributed effect.....inter-turn mutual inductance and capacitance as well as distributed capacitive effects with the cage. Capacitive loading on the top of the coil will indeed lower resonant frequency, however. End-loading of helical and coaxial resonators is indeed a way to tune them. Helical resonators can be thought of as a helical slot transmission line of about 1/4 wavelength (hence my question about the resonant frequency in my previous post).

I would suggest a good introductory text on EM theory to understand the basic mechanics. Also, have a look at references on helical resonators, if you have access to a uni library or IEEE Xplore.

HV Enthusiast

Tue Feb 14 2006, 03:19PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

wrote ...

Just length of wire means nothing here, frequency would be lower if you just used wire with thinner insulation, more of it or added topload for more capacitance to ground.

Waverider,

I believe you misunderstood what Firkragg is saying. He is basically saying the same thing as you, that length of a wire alone does not determine fres, and that parasitics (mutual turn-to-turn inductance and capacitance, stray capacitance to environment, etc...) present within the secondary must be accounted for when determing fres.

What he is saying is that if you took a 1000 turn secondary with x amount of turn-to-turn space (due to insulation), and used the same wire with smaller insulation, x-y amount of turn-to-turn space, that indeed the fres will change for that secondary, even though length of the wire has remained constant.

WaveRider

Tue Feb 14 2006, 03:47PM

Registered Member #29 Joined: Fri Feb 03 2006, 09:00AM
Location: Hasselt, Belgium
Posts: 500

If I sounded a bit harsh, I apologise

.

It is true that there are more variables than just the wire length. However, given a fixed wire/insulation thickness and winding radius, fres is clearly strongly influenced by the wire length. Also, the inter-turn spacing and shield spacing (or lack of shield) is indeed important and is one of the things I wish to understand better... What I understand so far is that traditional "lumped" models are not very good at describing what is happening here...

Cheers!

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob