Class C Power Oscillator

Hi bitrex, (I didn't download your spice file.) I've tried to simulate oscillators a few times in LTspice with various degrees of success. One trick is to kick start them... a big voltage or current pulse to get things going.

What's the power, freq, and load impdeance of your oscillator?

George H.

I'm not at my computer ATM, but in order to start reliably, the oscillator needs to be in Class A at zero signal, and have enough gain to be unstable. Otherwise a slow turn-on or brief brownout can stop it from oscillating.

Cheers

Phil Hobbs

Better to use a _small_ kickstart--just enough to get past roundoff and the various SPICE voltage and current tolerances. That leaves the active devic e in the small-signal regime, so you can verify that it starts reliably.

Cheers

Phil Hobbs

he various SPICE voltage and current tolerances. That leaves the active dev ice in the small-signal regime, so you can verify that it starts reliably.

right... forget about the simulation..

how does a pure class C osc get started in real life even?

Mark

Two ways. Bad ones, i.e. those whose quiescent bias is Class C, require a kick at turn-on. Good ones, which have unstable Class A quiescent bias, start up from noise.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

This is a reword Kickstart (eeww), or arrange the thing so that it's biased to be a class- A oscillator on startup, and therefor starts oscillating on its own.

--

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

On 9/24/2015 7:31 AM, bitrex wrote:

(snip)

I can rearrange the parts and get it to oscillate easily as a non-linear (class ?)arrangement. Here ya go...

Version 4 SHEET 1 880 680 WIRE -80 32 -144 32 WIRE 64 32 -80 32 WIRE 208 32 208 -16 WIRE 208 32 64 32 WIRE 608 64 304 64 WIRE -144 112 -144 32 WIRE 208 112 208 32 WIRE 304 112 304 64 WIRE 608 112 608 64 WIRE -272 128 -272 96 WIRE -80 128 -80 32 WIRE 64 128 64 32 WIRE 64 144 64 128 WIRE -272 240 -272 208 WIRE 304 240 304 192 WIRE 608 240 608 192 WIRE 608 240 304 240 WIRE 64 256 64 192 WIRE 208 256 208 192 WIRE 208 256 64 256 WIRE -144 272 -144 176 WIRE -80 272 -80 208 WIRE -80 272 -144 272 WIRE 464 272 -80 272 WIRE 464 288 464 272 WIRE 464 288 320 288 WIRE 608 336 608 240 WIRE 208 368 208 256 WIRE 400 368 320 368 WIRE 144 416 -160 416 WIRE 208 560 208 464 WIRE -160 640 -160 416 WIRE 400 640 400 368 WIRE 400 640 -160 640 FLAG 208 -16 Vcc FLAG 208 560 0 FLAG 608 336 0 FLAG -272 240 0 FLAG -272 96 Vcc SYMBOL ind2 192 96 R0 WINDOW 3 34 74 Left 2 SYMATTR Value 47uH SYMATTR InstName L1 SYMBOL ind2 288 96 R0 SYMATTR InstName L2 SYMATTR Value 100uH SYMBOL npn 144 368 R0 SYMATTR InstName Q1 SYMATTR Value 2SCR533P SYMBOL ind2 336 384 R180 WINDOW 0 36 80 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName L3 SYMATTR Value 4.7uH SYMBOL cap 48 128 R0 SYMATTR InstName C1 SYMATTR Value 10n SYMBOL res 592 96 R0 SYMATTR InstName R1 SYMATTR Value 10k SYMBOL res -96 112 R0 SYMATTR InstName R2 SYMATTR Value 47k SYMBOL voltage -272 112 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value 12 SYMBOL cap -160 112 R0 SYMATTR InstName C2 SYMATTR Value 1n TEXT 296 8 Left 2 !K1 L1 L2 L3 1 TEXT -392 296 Left 2 !.tran 0 2m 1m startup uic

Turn it into an amplifier and jigger it around until you have lots of gain at 0 degrees phase shift.

Here's a start. It's for your non-working amplifier and it shows gain, which puzzles the heck out of me -- but figuring it out is your job.

Version 4 SHEET 1 880 680 WIRE 64 32 -80 32 WIRE 208 32 208 -16 WIRE 208 32 64 32 WIRE 608 64 304 64 WIRE 208 112 208 32 WIRE 304 112 304 64 WIRE 608 112 608 64 WIRE -272 128 -272 96 WIRE -80 128 -80 32 WIRE 64 128 64 32 WIRE 64 144 64 128 WIRE -272 240 -272 208 WIRE 304 240 304 192 WIRE 608 240 608 192 WIRE 608 240 304 240 WIRE 64 256 64 192 WIRE 208 256 208 192 WIRE 208 256 64 256 WIRE 464 288 320 288 WIRE -80 320 -80 208 WIRE -16 320 -80 320 WIRE 96 320 48 320 WIRE 464 336 464 288 WIRE 608 336 608 240 WIRE 208 368 208 256 WIRE 400 368 320 368 WIRE 400 400 400 368 WIRE -288 416 -352 416 WIRE -192 416 -224 416 WIRE -80 416 -80 320 WIRE -80 416 -112 416 WIRE -32 416 -80 416 WIRE 96 416 96 320 WIRE 96 416 48 416 WIRE 144 416 96 416 WIRE -352 448 -352 416 WIRE 400 496 400 480 WIRE -352 560 -352 528 WIRE 208 560 208 464 FLAG 208 -16 Vcc FLAG 208 560 0 FLAG 464 336 0 FLAG 608 336 0 FLAG -272 240 0 FLAG -272 96 Vcc FLAG 400 496 0 FLAG -352 560 0 SYMBOL ind2 192 96 R0 WINDOW 3 34 74 Left 2

SYMATTR InstName L1 SYMBOL ind2 288 96 R0 SYMATTR InstName L2

SYMBOL npn 144 368 R0 SYMATTR InstName Q1 SYMATTR Value 2SCR533P SYMBOL ind2 336 384 R180 WINDOW 0 36 80 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName L3

SYMBOL cap 48 128 R0 SYMATTR InstName C1 SYMATTR Value 10n SYMBOL res 592 96 R0 SYMATTR InstName R1 SYMATTR Value 10k SYMBOL res -96 112 R0 SYMATTR InstName R2 SYMATTR Value 47k SYMBOL res 64 400 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R3 SYMATTR Value 470 SYMBOL cap 48 304 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C2 SYMATTR Value 100n SYMBOL voltage -272 112 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value 12 SYMBOL res 384 384 R0 SYMATTR InstName R4 SYMATTR Value 50 SYMBOL res -96 400 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R5 SYMATTR Value 50 SYMBOL voltage -352 432 R0 WINDOW 123 24 124 Left 2 WINDOW 39 0 0 Left 2 SYMATTR Value2 AC 1 SYMATTR InstName V2 SYMATTR Value 0 SYMBOL cap -224 400 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C3 SYMATTR Value 100n TEXT 296 8 Left 2 !K1 L1 L2 L3 1 TEXT -392 296 Left 2 !.tran 0.1 startup uic TEXT -392 328 Left 2 !;ac dec 40 10k 10meg

--

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

On Thu, 24 Sep 2015 08:31:08 -0400, bitrex wrote:

L3 shorts out the base bias. Put in a blocking cap and it sings.

The "treat it as an amplifier" analysis says that it should work with R1 at 100 ohms, but in reality if you put R1 much less than 10k then Bad Things happen -- I don't know why, chasing that down can be an exercise for you, the reader, to figure out.

You may want to try a plain ol' Colpitts topology and see if you can find a combination that works.

Version 4 SHEET 1 880 680 WIRE 64 32 -80 32 WIRE 208 32 208 -16 WIRE 208 32 64 32 WIRE 608 64 304 64 WIRE 656 64 608 64 WIRE 208 112 208 32 WIRE 304 112 304 64 WIRE 608 112 608 64 WIRE -272 128 -272 96 WIRE -80 128 -80 32 WIRE 64 128 64 32 WIRE -272 240 -272 208 WIRE 304 240 304 192 WIRE 608 240 608 192 WIRE 608 240 304 240 WIRE 64 256 64 192 WIRE 208 256 208 192 WIRE 208 256 64 256 WIRE 464 288 320 288 WIRE -80 320 -80 208 WIRE -16 320 -80 320 WIRE 96 320 48 320 WIRE 464 336 464 288 WIRE 608 336 608 240 WIRE 208 368 208 256 WIRE 400 368 320 368 WIRE -80 416 -80 320 WIRE -32 416 -80 416 WIRE 96 416 96 320 WIRE 96 416 48 416 WIRE 144 416 96 416 WIRE -80 480 -80 416 WIRE 208 560 208 464 WIRE -80 640 -80 544 WIRE 400 640 400 368 WIRE 400 640 -80 640 FLAG 208 -16 Vcc FLAG 208 560 0 FLAG 464 336 0 FLAG 608 336 0 FLAG -272 240 0 FLAG -272 96 Vcc FLAG 656 64 Vout SYMBOL ind2 192 96 R0 WINDOW 3 34 74 Left 2 SYMATTR Value 47?H SYMATTR InstName L1 SYMBOL ind2 288 96 R0 SYMATTR InstName L2 SYMATTR Value 100?H SYMBOL npn 144 368 R0 SYMATTR InstName Q1 SYMATTR Value 2SCR533P SYMBOL ind2 336 384 R180 WINDOW 0 36 80 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName L3 SYMATTR Value 4.7?H SYMBOL cap 48 128 R0 SYMATTR InstName C1 SYMATTR Value 10n SYMBOL res 592 96 R0 SYMATTR InstName R1 SYMATTR Value 10k SYMBOL res -96 112 R0 SYMATTR InstName R2 SYMATTR Value 47k SYMBOL res 64 400 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R3 SYMATTR Value 470 SYMBOL cap 48 304 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C2 SYMATTR Value 100n SYMBOL voltage -272 112 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value PWL(0 0 5m 12) SYMBOL cap -96 480 R0 SYMATTR InstName C3 SYMATTR Value 100n TEXT 296 8 Left 2 !K1 L1 L2 L3 1 TEXT -392 296 Left 2 !.tran 10m startup uic

--

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

Try this, it even has regulation (bottom left):

(Dang, don't have core data written down on that one, you'll have to guess.)

Because of the high voltage load, the capacitance is fairly dominant and it runs more-or-less full wave class C/D/E, rather than square flyback like a conventional blocking oscillator supply. In your case, that'll be the transducer providing most of the loading (capacitive for high frequencies, and complex around resonance).

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

I'm trying to design an ultrasonic class C power oscillator to use with a project I'm working on. I'm vaguely basing it off a circuit I recall seeing in an early 1960s-era book called "Industrial Transistor Circuits". It seems I'm not remembering it correctly, and I don't really know much about the design of this type of oscillator as it's having some problems oscillating in simulation, even doing the "start supply voltages at 0 V" thing.

I'd like to use a small air core, maybe 1.5 inches in diameter by 5 inches in length, something like that.

Here's the .asc:

Whooooooops! That was a dumb mistake. Must have been no caffeine in the morning coffee...

I found "Oscillator Design and Computer Simulation" by Randal Rhea to be very helpful. I never designed power oscillators with it, however.

Given the size of components these days, a master oscillator followed by a power amp may not be much bigger and can almost certainly be made more efficient.

--

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

If you can live with something slightly more complicated, with a ferrite core in a wound transformer, and a separate inductor (which you can probably buy off the shelf), the class-D Baxandall oscillator can be very designable.

The original paper from 1959 is written around bipolar transistors

In some applications, MOSFET switches work better, as I've discussed

--
Bill Sloman, Sydney