200kHz oscillator

I would like to experiment a bit with magnetic parametric amplifiers.
For that purpose, I believe I need a 200kHz sine wave generator with
a watt worth of output power. My obvious choice was to construct
a current-fed resonant Royer oscillator. For a 1nF tank capacitor
it implies a 633uH inductor (2x5 turns on a 6400nH/turn^2 way to
big toroidal core), 2x ZTX851, a 100k base polarization resistor,
1 turn of feedback winding, 15 turns on the load winding. The
current-mode choke is 7 turns on a 5500nH/t^2 core. The
supply voltage is 8V.

And it works like a charm, the output waveform is *very* close
to a sine, 183kHz is also great for the first shot. But it works
only when not loaded. Even a 910 resistor shunting the load winding
is sufficient to kill the oscillations. Well, it still oscillates,
but the waveform does not resemble anything typical. A train of
weak wave packets with 2 maxima each, strange.

There is something terribly wrong with my circuit, people are
using it for muli-watt backlight applications without much problems.

What proven oscillator design would you suggest using instead?

	Best regards, Piotr

On Tue, 2 May 2017 19:56:46 +0200, Piotr Wyderski
<peter.pan@neverland.mil> wrote:

>I would like to experiment a bit with magnetic parametric amplifiers.
>For that purpose, I believe I need a 200kHz sine wave generator with
>a watt worth of output power. 

Most benctop function generators can just about do that.


-- 

John Larkin         Highland Technology, Inc
picosecond timing   precision measurement 

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

On Tuesday, May 2, 2017 at 1:56:55 PM UTC-4, Piotr Wyderski wrote:
> I would like to experiment a bit with magnetic parametric amplifiers.
> For that purpose, I believe I need a 200kHz sine wave generator with
> a watt worth of output power. My obvious choice was to construct
> a current-fed resonant Royer oscillator. For a 1nF tank capacitor
> it implies a 633uH inductor (2x5 turns on a 6400nH/turn^2 way to
> big toroidal core), 2x ZTX851, a 100k base polarization resistor,
> 1 turn of feedback winding, 15 turns on the load winding. The
> current-mode choke is 7 turns on a 5500nH/t^2 core. The
> supply voltage is 8V.
> 
> And it works like a charm, the output waveform is *very* close
> to a sine, 183kHz is also great for the first shot. But it works
> only when not loaded. Even a 910 resistor shunting the load winding
> is sufficient to kill the oscillations. Well, it still oscillates,
> but the waveform does not resemble anything typical. A train of
> weak wave packets with 2 maxima each, strange.
> 
> There is something terribly wrong with my circuit, people are
> using it for muli-watt backlight applications without much problems.
> 
> What proven oscillator design would you suggest using instead?
> 
> 	Best regards, Piotr

What kind of power?  Voltage and current.  
I was thinking a sig. gen. and power opamp. 

George H.  

On Tue, 02 May 2017 19:56:46 +0200, Piotr Wyderski wrote:

> I would like to experiment a bit with magnetic parametric amplifiers.
> For that purpose, I believe I need a 200kHz sine wave generator with a
> watt worth of output power. My obvious choice was to construct a
> current-fed resonant Royer oscillator. For a 1nF tank capacitor it
> implies a 633uH inductor (2x5 turns on a 6400nH/turn^2 way to big
> toroidal core), 2x ZTX851, a 100k base polarization resistor,
> 1 turn of feedback winding, 15 turns on the load winding. The
> current-mode choke is 7 turns on a 5500nH/t^2 core. The supply voltage
> is 8V.
> 
> And it works like a charm, the output waveform is *very* close to a
> sine, 183kHz is also great for the first shot. But it works only when
> not loaded. Even a 910 resistor shunting the load winding is sufficient
> to kill the oscillations. Well, it still oscillates, but the waveform
> does not resemble anything typical. A train of weak wave packets with 2
> maxima each, strange.
> 
> There is something terribly wrong with my circuit, people are using it
> for muli-watt backlight applications without much problems.
> 
> What proven oscillator design would you suggest using instead?
> 
> 	Best regards, Piotr

Generate at low level, buffer, filter, amplify the s**t out of it.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

I'm looking for work -- see my website!

Well... what V, I and Z were you expecting?  Because it doesn't sound like 
much, assuming you chose the LC values intentionally.

1W at 50 ohms is 7V and 141mA, RMS.  A Royer doesn't really work at 
extremely low Q, so let's say the minimum Q should be 3, or a parallel 
resonant tank of 16.7 ohms.  Or 47nF and 13.3uH.

To keep it going under heavier loading, you probably want much less than 
that, so the Q stays above 3 or 5, say, at the highest loading.

An ungapped winding doesn't mean much, either.  You need air to store 
energy!

And you should probably use power MOSFETs, like a pair of IRF540.  No need 
to worry about base current, or teensy BJTs puffing off when you inevitably 
short it out. :-)

Tim

-- 
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Contract Design
Website: http://seventransistorlabs.com


"Piotr Wyderski" <peter.pan@neverland.mil> wrote in message 
news:oeah93$309$1@node1.news.atman.pl...
>I would like to experiment a bit with magnetic parametric amplifiers.
> For that purpose, I believe I need a 200kHz sine wave generator with
> a watt worth of output power. My obvious choice was to construct
> a current-fed resonant Royer oscillator. For a 1nF tank capacitor
> it implies a 633uH inductor (2x5 turns on a 6400nH/turn^2 way to
> big toroidal core), 2x ZTX851, a 100k base polarization resistor,
> 1 turn of feedback winding, 15 turns on the load winding. The
> current-mode choke is 7 turns on a 5500nH/t^2 core. The
> supply voltage is 8V.
>
> And it works like a charm, the output waveform is *very* close
> to a sine, 183kHz is also great for the first shot. But it works
> only when not loaded. Even a 910 resistor shunting the load winding
> is sufficient to kill the oscillations. Well, it still oscillates,
> but the waveform does not resemble anything typical. A train of
> weak wave packets with 2 maxima each, strange.
>
> There is something terribly wrong with my circuit, people are
> using it for muli-watt backlight applications without much problems.
>
> What proven oscillator design would you suggest using instead?
>
> Best regards, Piotr 

On Tue, 02 May 2017 11:29:20 -0700, the renowned John Larkin
<jjlarkin@highland_snip_technology.com> wrote:

>On Tue, 2 May 2017 19:56:46 +0200, Piotr Wyderski
><peter.pan@neverland.mil> wrote:
>
>>I would like to experiment a bit with magnetic parametric amplifiers.
>>For that purpose, I believe I need a 200kHz sine wave generator with
>>a watt worth of output power. 
>
>Most benctop function generators can just about do that.

Yup. 10V peak ~= 7V RMS into 50 ohms is about 1W. Which is why the
function generator output amplifiers need to be beefy. 

I had a look at one of the Agilent ones' output before sending it back
for repair (wasn't the output section, I suspect it lost its
calibration constants) .. parallel high frequency high performance
op-amps IIRC. 

--sp 


-- 
Best regards, 
Spehro Pefhany

On Wednesday, May 3, 2017 at 3:56:55 AM UTC+10, Piotr Wyderski wrote:
> I would like to experiment a bit with magnetic parametric amplifiers.
> For that purpose, I believe I need a 200kHz sine wave generator with
> a watt worth of output power. My obvious choice was to construct
> a current-fed resonant Royer oscillator.

If it's resonant and produces a sine wave it isn't a Royer oscillator, but a Baxandall oscillator.

http://sophia-elektronica.com/0344_001_Baxandal.pdf


> For a 1nF tank capacitor
> it implies a 633uH inductor (2x5 turns on a 6400nH/turn^2 way to
> big toroidal core), 2x ZTX851, a 100k base polarization resistor,
> 1 turn of feedback winding, 15 turns on the load winding. The
> current-mode choke is 7 turns on a 5500nH/t^2 core. The
> supply voltage is 8V.

If you make the choke too inductive in a classical Baxandall oscillator it can "squeg". This only happens if you use bipolar transistors as switches, and doesn't happen if you use MOSFET switches (though start-up voltages can go rather high with a high feed inductor). 

With a supply voltage of 8V you can't just connect the gate of one MOSFET to the drain of the other - you get more than 20V between gate and source, and it can be even more during start-up.

The simplest MOSFET driven circuits can be difficult to get to start up neatly - using a pair of bipolar transistors to drive the MOSFETs offers a tidy solution - see the oscillator part of the circuit at the bottom of the page.

http://www.sophia-electronica.com/Baxandall_parallel-resonant_Class-D_oscillator1.htm
 
> And it works like a charm, the output waveform is *very* close
> to a sine, 183kHz is also great for the first shot. But it works
> only when not loaded. Even a 910 resistor shunting the load winding
> is sufficient to kill the oscillations. Well, it still oscillates,
> but the waveform does not resemble anything typical. A train of
> weak wave packets with 2 maxima each, strange.

If you load a Baxandall enough, the harmonic content of the output goes up a lot, and you can kill the oscillation completely. The late Tony Williams had found that a Q higher then ten didn't give you a better sine wave, and a Q less than five gave you something less than good.
 
> There is something terribly wrong with my circuit, people are
> using it for muli-watt backlight applications without much problems.

100k of base polarisation doesn't sound like much.

Spice it and look at the actual currents.
 
> What proven oscillator design would you suggest using instead?

Baxandall should work fine, but you do have to think about what is going on.

-- 
Bill Sloman, Sydney

On a sunny day (Tue, 2 May 2017 19:56:46 +0200) it happened Piotr Wyderski
<peter.pan@neverland.mil> wrote in <oeah93$309$1@node1.news.atman.pl>:

>I would like to experiment a bit with magnetic parametric amplifiers.
>For that purpose, I believe I need a 200kHz sine wave generator with
>a watt worth of output power. My obvious choice was to construct
>a current-fed resonant Royer oscillator. For a 1nF tank capacitor
>it implies a 633uH inductor (2x5 turns on a 6400nH/turn^2 way to
>big toroidal core), 2x ZTX851, a 100k base polarization resistor,
>1 turn of feedback winding, 15 turns on the load winding. The
>current-mode choke is 7 turns on a 5500nH/t^2 core. The
>supply voltage is 8V.
>
>And it works like a charm, the output waveform is *very* close
>to a sine, 183kHz is also great for the first shot. But it works
>only when not loaded. Even a 910 resistor shunting the load winding
>is sufficient to kill the oscillations. Well, it still oscillates,
>but the waveform does not resemble anything typical. A train of
>weak wave packets with 2 maxima each, strange.
>
>There is something terribly wrong with my circuit, people are
>using it for muli-watt backlight applications without much problems.
>
>What proven oscillator design would you suggest using instead?
>
>        Best regards, Piotr

Do it the China way:
 http://panteltje.com/pub/inductive_coupling_real_power_300Vpp_IMG_6092.JPG

:-)

Jan Panteltje wrote...
>
> Do it the China way:
> http://panteltje.com/pub/inductive_coupling_real_power_300Vpp_IMG_6092.JPG

 More details, please.


-- 
 Thanks,
    - Win

Den torsdag den 4. maj 2017 kl. 23.36.18 UTC+2 skrev Winfield Hill:
> Jan Panteltje wrote...
> >
> > Do it the China way:
> > http://panteltje.com/pub/inductive_coupling_real_power_300Vpp_IMG_6092.JPG
> 
>  More details, please.

the pcb looks like one of those cheap chinese induction heater driver, 
based roughly on this:

https://markobakula.files.wordpress.com/2012/10/royer_ih.png