Magamp oscillator (2023 Update)

It's not an oscillator, either.

I don't see any gain. Wout = Win * ?

The 60 Hz pump trick adds energy to the resonant device, the pendulum, just as a transistor (or a varicap pump) adds energy to an LC. Adding energy keeps the oscillation from dying out.

A pumped resonator is an oscillator. As a bonus, the parametric oscillator is phase-locked to the pump.

You need to check your values. LTspice defaults to 1e-3 ohm ESR for inductors. XL of a 1uH inductor is 603.688 ohms at 96.08meg. This gives a Q of 603,688, which is a bit beyond current technology.

Similarly, the SRF of a 1uH inductor is around 200MHz. This gives a parallel capacitance of 0.63pf. Add 1.4pf for pad capacitance and you have around 2pf across the inductor.

Put 6 ohms in series with the inductor to drop the Q to 100 at 100MHz, which is probably unattainable with ordinary commercial inductors, and look at your waveforms. The F/2 disappears after a few cycles.

I'm not sure the model for the varactor is accurate. The datasheet shows a Q of 500 at 50MHz, but the model shows 5.41 ohms in series with the varactor diode, so the Q spec doesn't make sense.

Version 4 SHEET 1 880 680 WIRE -16 80 -48 80 WIRE 80 80 -16 80 WIRE 208 80 144 80 WIRE 320 80 208 80 WIRE 384 80 320 80 WIRE 400 80 384 80 WIRE -48 96 -48 80 WIRE 208 96 208 80 WIRE 320 96 320 80 WIRE 400 96 400 80 WIRE 208 176 208 160 WIRE 320 176 320 160 WIRE -48 192 -48 176 WIRE 400 208 400 176 WIRE 400 304 400 288 FLAG 400 304 0 FLAG -48 192 0 FLAG 384 80 tank FLAG -16 80 pump FLAG 208 176 0 FLAG 320 176 0 SYMBOL varactor 144 64 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName D1 SYMATTR Value MV2201 SYMBOL ind 384 80 R0 SYMATTR InstName L1

SYMATTR SpiceLine Rser=1u SYMBOL voltage -48 80 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value SINE(1 1 96.08meg) SYMBOL res 384 192 R0 SYMATTR InstName ESR SYMATTR Value 6 SYMBOL cap 192 96 R0 SYMATTR InstName PAD SYMATTR Value 1.4pf SYMBOL cap 304 96 R0 SYMATTR InstName SRF SYMATTR Value 0.63pf TEXT 160 8 Left 2 !.tran 1u TEXT 160 -24 Left 2 ;'F/2 Generator TEXT 416 192 Left 2 ;XL = 603.688 @ 96.08meg TEXT 416 288 Left 2 ;Q = 100.614

[Transient Analysis] { Npanes: 3 Active Pane: 1 { traces: 1 {524291,0,"V(tank)"} X: ('n',0,0,5e-008,5e-007) Y[0]: (' ',1,-4.2,0.6,3) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,1,-4.2,0.6,3) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {34603012,0,"I(D1)"} X: ('n',0,0,5e-008,5e-007) Y[0]: ('m',0,-0.008,0.002,0.01) Y[1]: ('_',0,1e+308,0,-1e+308) Amps: ('m',0,0,1,-0.008,0.002,0.01) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {268959746,0,"V(pump)"} X: ('n',0,0,5e-008,5e-007) Y[0]: (' ',1,0,0.2,2) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,1,0,0.2,2) Log: 0 0 0 GridStyle: 1 } }

Same problem as before. If the ASC file has a value of 1u (1e-6), Microsoft may change it to some other Unicode character depending on which OS version you are running. These will be unreadable by others who are not running the same version. Or maybe it won't be readable by anyone.

A suggested bypass is to omit symbols such as m, n, u, etc., and enter the value in exponential code such as 1e-3 = m, 1e-6 = u, 1e-9 = n, etc.

Microsfot cannot screw that up.

You obviously consider the pendulum an oscillator but I think you're alone. I see it as nothing more than a tank. There is no amplification or feedback. What's the transfer equation?

I don't see it as being phase locked, any more than an LC is phase locked (its dual), either.

Helical and coaxial resonators can easily hit Qs of 600.

Cheers

Phil Hobbs

As a minus, it must be phase-locked to the pump..... Why bother using a varicap? It's spice, after all.

Version 4 SHEET 1 880 680 WIRE 240 80 144 80 WIRE 288 80 240 80 WIRE 144 96 144 80 WIRE 288 128 288 80 WIRE 144 208 144 160 WIRE 144 208 96 208 WIRE 144 256 144 208 WIRE 288 272 288 208 WIRE 144 352 144 336 FLAG 288 272 0 FLAG 144 352 0 FLAG 240 80 tank FLAG 96 208 pump SYMBOL ind 272 112 R0 SYMATTR InstName L1

SYMBOL voltage 144 240 R0 WINDOW 123 0 0 Left 2 WINDOW 39 24 124 Left 2 SYMATTR InstName V1 SYMATTR Value SINE(1 1 96.08meg) SYMBOL diode 128 96 R0 SYMATTR InstName D2 SYMATTR Value 1SR154-400 TEXT 176 24 Left 2 !.tran 0 1u .9u

Hi Tim, Re: parametric amps/ oscillators, I think there is a nice discussion of this in AB. Pippard's "Physics of Vibration, Vol 1." (Wow searching on line I can get both volumes for a few dollars. I haven't read vol II. )

My copy is at home and it's been a while since I've read it. So perhaps I'm mis-remembering.

George H.

I'm at work and not at home where the text is, but

Yes, John has mentioned that before.

Doesn't do any good. You can change the series resistance in the ASC file I posted to give a Q of 600. Increase the time to 5us. The oscillations peter out in about 2us.

The F/2 signal is merely an artifact of the startup transient. The circuit does not generate a continuous F/2 signal.

I am doing further studies on the MV2201 varactor ESR. The datasheet says the series resistance (Rs) is 5.41 ohms. A simple series resonant circuit in LTspice shows the ESR is abot 1.01 ohms at 0.25V bias. So there appears to be a discrepancy between the datasheet and the SPICE model.

Same problem. Add some realistic ESR and stray capacity and the F/2 signal disappears.

In a university campus, in a galaxy far, far away......

RL

You don't think that something that oscillates is an oscillator?

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

If you add an idler resonator, you can make a paramp with gain anywhere you like (below the pump frequency). There's an additional degree of freedom then, so there's no necessary phase relationship between the output and the pump, except that idler phase + output phase = pump phase.

The pendulum is really an injection-locked system rather than a phase-locked one, if one wants to be pedantic about it.

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

Not true in general. You can buy frequency halvers, no problem. Making the varactor a smaller part of the tank capacitance fixes it.

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

I invite you to read the paper I linked, about how to optimize a circuit that you appear to be claiming is impossible.

Because that one is behind a paywall, I've posted a several-page excerpt from the ARRL VHF Handbook, First Ed. (1965) about how paramps work and how to build them. 'taint rocket surgery.

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

Parametric amplifiers are well-known. The paper you reference talks about a down converter that requires a signal input, a pump signal, and the output. That requires two input signals to generate the lower output frequency, similar to a mixer.

They do not have a single varactor and inductor with a single input frequency as shown in Jeroen's circuit.

I don't understand your statement that I claim the circuit is impossible. Of course you can run anything in LTspice. You don't have to include realistic stray capacitances and impedances. That doesn't mean it will work on the bench.

The circuit I posted shows the F/2 signal dies out in a few cycles when you include realistic ESR and stray capacitances.

Please post a circuit showing how to optimize a single varactor and inductor as in Jeroen's circuit to produce a sub-multiple of a single input signal.

I am unable to find commercial versions of a frequency halver. Can you supply some links?

In Legg's circuit, as soon as you add realistic ESR and stray capacitance across the inductor, the F/2 signal disappears