Frequency Splitting - Measurable at all? / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Frequency Splitting - Measurable at all?

1 2 next

Move Thread

LAN_403

Tesla_Bozo

Wed Jun 18 2008, 01:21PM

Registered Member #1527 Joined: Mon Jun 09 2008, 05:00PM
Location:
Posts: 4

Hi,

Is it possible to hook your primary side to an RF generator and observe frequency splitting through a probe connected to a sepctrum analyzer on the secondary? (i.e. observe two peaks to the left and to the right of the self-resonant frequency while sweeping frequencies on the RF generator). I've tried it myself, but I only get one peak at the self resonant frequency.

Any ideas?

GeordieBoy

Wed Jun 18 2008, 03:05PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Yes, but you don't connect the probe TO the secondary. You just need to hang it in the air about 6ft away so it samples the E-field without significantly detuning the resonator.

You can do it with:

1. A manually swept function generator and an oscilloscope.
2. An automated sweep generator and a spectrum analyser.
3. A combined generator / network analyser (VNA)

It goes without saying that the coil should be disconnected from the mains or any other HV source while performing this small-signal measurement with expensive equipment!

The above site explains the theory and gives equations for expected frequency modes based on coupling factor.

-Richie,

Tesla_Bozo

Wed Jun 18 2008, 03:35PM

Registered Member #1527 Joined: Mon Jun 09 2008, 05:00PM
Location:
Posts: 4

Many thanks Richie. That's one hell of a website you've got. Would you connect the RF generator in series with the LC tank (i.e. replace the spark gap with the generator), or would you connect it through a resistor to a shunted LC tank?

Also, I need to give an admission. I'm not building a 'normal' Tesla coil. My primary and secondary tanks are IDENTICAL, with only a few loops for the inductors on both sides. I don't know if that rules out the far-away-dangling probe measurment?

.. and sorry for asking too much.

Steve Ward

Wed Jun 18 2008, 10:44PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

You can directly probe these low impedance circuits, so dont expect to see much when dangling the probe in the air.

Coronafix

Wed Jun 18 2008, 10:45PM

Registered Member #160 Joined: Mon Feb 13 2006, 02:07AM
Location: Melbourne, Australia
Posts: 938

Seeing that frequency splitting is a product of coupling, or over-coupling as Hams would say,
is it more efficient to reduce coupling so that there is only one frequency?
Does this affect efficiency at all?
I know that DRSSTCs use a high coupling because of the lower E-field in the primary and achieve greater
spark length than similar size SGTCs, but the secondary is therefore "constrained" from self oscillating, and I
wonder if it wasn't for the efficiency of the silicon over the spark gap, would the lower coupling win.
Perhaps that's where Terry's SISG would come into it.

HV Enthusiast

Thu Jun 19 2008, 02:22AM

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

Its even easier to do with a current transformer.
Just hook the output of your wideband current transformer to a spectrum analyzer.

Steve Ward

Thu Jun 19 2008, 08:09PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I know that DRSSTCs use a high coupling because of the lower E-field in the primary and achieve greater
spark length than similar size SGTCs, but the secondary is therefore "constrained" from self oscillating, and I
wonder if it wasn't for the efficiency of the silicon over the spark gap, would the lower coupling win.

You are misinformed. 1) e-field in primary doesnt matter, its the B-field that matters. 2) DRSSTCs can provide just as much peak current (B-field) as a SGTC! Ive made DRSSTCs with lower coupling, it does work fine, but higher coupling is still more efficient, and spark gap coils would also favor higher coupling if insulation breakdown wasnt a problem.

Frequency splitting is *always* present with a 4th order coupled system, it can be less dramatic with lower coupling. Please realize that frequency splitting isnt a "problem" to be overcome, its just a fact of the physics of energy transfer between the systems!

GeordieBoy

Thu Jun 19 2008, 09:43PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Thanks for the comments about the website. That was written many many years ago so I hope there aren't too many errors!

If the two parallel resonant circuits are similar then their characteristic impedances are similar too. In this case you can connect a x10 scope probe directly across one of the tank circuits provided the added capacitance of the probe doesn't significantly detune that tank circuit with respect to the other one.

I would drive the primary tank circuit from the sweep generator with a 10k resistor in series with it's output. This should give little damping to the two parallel resonant circuits, so you should see two peaks at the output side when they are tightly coupled.

As someone else mentioned this over-coupling is often used in things like IF-transformers to get the desired bandwidth for radio receivers.

Hope this helps,

-Richie,

Coronafix

Fri Jun 20 2008, 01:20AM

Registered Member #160 Joined: Mon Feb 13 2006, 02:07AM
Location: Melbourne, Australia
Posts: 938

Steve Ward wrote ...

Please realize that frequency splitting isnt a "problem" to be overcome, its just a fact of the physics of energy transfer between the systems!

I wasn't seeing it as a problem, I was just thinking that because the tight coupling creates two peaks on either side of the resonant, that they are therefore harmonics and not the exact frequency of the system. I am probably wrong then.
I thought the E-field played a major part in TCs hence the reason we use higher voltages in the primary with a flat spiral?

Steve Ward

Fri Jun 20 2008, 04:29AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I was just thinking that because the tight coupling creates two peaks on either side of the resonant, that they are therefore harmonics and not the exact frequency of the system. I am probably wrong then.

The 2 frequencies are actually seen as "beating", and is an indication of energy moving between the primary and secondary systems. Ideally the primary current should notch (zero out from beating) just 1 time. When the primary current hits a zero on its envelope, the secondary current is maximum, so all the energy is in the secondary system, this is the ideal time to stop the energy from dumping back into the primary side. Lower coupling just makes the beat frequency lower, so you get more cycles to transfer the energy (well, ideally this is true). But, at the same time, you lose energy on every cycle, so you dont want to wait *too* long, otherwise you spend it all in the spark gap or conductors.

I thought the E-field played a major part in TCs hence the reason we use higher voltages in the primary with a flat spiral?

Ah, yes, this is the E-field of the secondary coil that is most problematic. The flat spiral helps to avoid severly distorting the voltage gradient, but the voltage gradient is from the secondary, not really the primary. You would have the same problem with grounded objects in the vicinity of the secondary coil, keeping them approximately large and flat gives the best voltage gradient for preventing sparks from forming.

Using higher primary voltages has to do with the efficiency of spark gaps. Spark gaps act more like a voltage drop device rather than resistive (well, its resistive, but that not the entire loss). So going with higher voltages not only allows larger tank impedance which can lower surge currents and thus lower the power loss in the spark gap.

1 2 next

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