Quadrature Oscillator

Nice I saved that.

Ahh, That's what I thought. I'll need 4MHz synced with the 1 MHz for the flip-flop thing. And then I'll want some phase "tweaking" knob down stream too.

Nice, I do like Charles W's website.. lotsa great stuff!

I want to thank you for the ideas and leg work. At the moment I'm going to put the quad oscillator/ quad signal generation to bed. (Maybe to be used sometime in the future.) I've got a Rigol DG1022 which is a ~$350 solution to the problem. It can give me both I and Q.. and then I just need to make a switched gain stage (+1/-1) and a low pass filter on the output and I'm there. (The switched gain stage will be tricky.. but hey in the short term maybe just an AD630.) One thing I like about the external generator idea is that it will be easier keeping the reference oscillator from leaking into the signal chain. That can be a real pain.. all sorts of sneak paths at the 100nV-1uV level. (I've still got battle scars, from my first lockin.:^)

Oh a final 'scope shot. For the 5MHz opamp quad oscillator the 16pf scope probes were distorting the signals some. I tacked in 1pF coupling caps and got this,

Still not a perfect circle. But OK.

George H.

But how would you do that in real life? There aren't any really big varactors left, and anyway, they'd cause sine distortion (and probably parametric phase shift as well). It could probably be done, but not so easily.

With good components, the adjustment would only have to be a percent or two, which won't make much difference to the measurement. I'm not wedded to the idea of servoing the frequency, I just thought it was fun. A dpot in front of the integrator's input resistor would work too, but you'd have to watch the capacitances.

I don't recall your being the OP. I use Johnson counters fairly often. My first PLL, back in 1981, used one to generate a quadrature signal for the lock detector since the main PD servoed on the null.

So they're great. But George was asking for sine waves, unless I'm very much mistaken.

Which is much easier to do well than square->sine.

Not hard at all. Capacitive coupling into a low-offset comparator, or even a crappy comparator with a simple positive/negative feedback network to force the duty cycle to 50%.

I don't know about that. The ADA4817 is a unity-gain stable op amp with a GBW of 1.4 GHz.

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

At 1 MHz, it's not hard; find/make a transformer with a center-tapped secondary, and ground one leg on Q, the other on /Q. Take your output from the center tap.

Ahha! The Baxandall Class-D demodulator.

If you want to go back almost that far, the 1966 Faulkner and Harding long-tailed pair current-steering demodulator is probably more like it.

--
Bill Sloman, Sydney

e

Jim Thompson would use a gyrator. I'd use an AD734 configured as a current source and driven by the quadrature waveform to produce the current that wo uld be drawn by a positive or negative capacitor (or a negative or positive inductor) depending on the sign of the - adjustable but quasi-constant - v oltage applied to the other input of the multiplier.

Varactor's aren't the only way of getting an adjustable reactive current. N ice and simple to wire in, but total swine as far as far as linearity goes.

--
Bill Sloman, Sydney

I was sort of teasing John S., who seemed to be temporarily stuck in LTspice mode. It's far easier to tune an integrator by adjusting the resistor. Failing that, I'd probably use a fixed cap wrapped round a variable-gain amplifier, but that's a bit niggly when you're building an integrator, because neither end of the cap is grounded.

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

A friend recently used series/parallel combination of 480 varactors to tune a hand-held sampling head (spanning 1-5MHz). So that's one option, I guess :).

Yowza. A bunch of Y5V caps in series would do about as well, I expect.

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

the

Leif Asbrink uses 24V zener diodes in a 14Mhz VCO:

Of course, you need a CV analyzer to characterize the zeners, which is why I decided to build one.

The sine distortion can be minimized by using two diodes back to back and adding the control voltage to the junction.

as

fun.

The pot is probably a better idea. It should not need continuous adjustment, so it could be a simple tweak pot on the pcb. Only needs to be tweaked as the component age and the circuit drifts.

I never said I was. Once you figure out what George is trying to do, it is obvious sine waves are not needed. In fact, they are an unnecessary complication.

very

He didn't realize he could do the job with square waves. Once he realized he would have to convert them into square waves, he stated:

Quote:

And yeah I think it will all get turned into square waves to turn switches or something on and off. So your D-flip flop circuit may do the trick.

Maybe not. See the approach at the end.

a

Too complicated, adds jitter, difficult to guarantee quadrature, difficult to calibrate, drifts, etc.

It is much simpler to start with a 4X clock and use a Johnson counter.

would

with

One degree of phase shift at 10 MHz is 100e-9/360 = 270 picoseconds. Good Luck.

Where do you get a low jitter, low harmonic distortion, phase and amplitude lockable sine wave oscillator at 10 MHz?

Sine wave oscillators are notoriously difficult at RF frequencies. They have lots of jitter and harmonic distortion, and have difficulty controlling the amplitude and locking to a desired phase. They also tend to drift as the components age and the temperature changes.

This approach is very low jitter and has essentially the same jitter as the source clock. According to LTspice, the second harmonic distortion is about -80dBc, which is instrumentation-class performance. It can easily be locked to the desired phase, and the amplitude is easily controlled. It is inexpensive and takes very little space on the pcb.

The CV analyzer may have to deal with signal amplitudes in the microvolt or nanovolt range. This requires a high gain ampifier in the detection chain, which will have considerable propagation delay. This means the quadrature square wave signals will be out of phase with the sine wave.

A second Johnson counter is needed to generate the quadrature detection signals. It could be driven from the same 4X clock through a delay or adjustable phase circuit.

Here are the LTspice files:

SHEET 1 1820 692 WIRE 1136 -128 608 -128 WIRE 608 -48 608 -128 WIRE 640 -48 608 -48 WIRE 864 -48 800 -48 WIRE 944 -48 864 -48 WIRE 1472 -48 1104 -48 WIRE 1488 -48 1472 -48 WIRE 640 0 608 0 WIRE 832 0 816 0 WIRE 944 0 912 0 WIRE 1136 0 1136 -128 WIRE 1136 0 1120 0 WIRE 1472 0 1136 0 WIRE 1488 0 1472 0 WIRE 608 96 608 0 WIRE 672 96 608 96 WIRE 912 96 912 0 WIRE 912 96 672 96 WIRE 608 112 608 96 WIRE 864 128 864 -48 WIRE 1200 128 864 128 WIRE 1472 128 1200 128 WIRE 1488 128 1472 128 WIRE 832 160 832 0 WIRE 1472 160 832 160 WIRE 1488 160 1472 160 WIRE 960 192 944 192 WIRE 1008 192 960 192 WIRE 1088 192 1008 192 WIRE 608 208 608 192 WIRE 1008 208 1008 192 WIRE 1088 208 1088 192 WIRE 1488 240 1232 240 WIRE 1008 304 1008 272 WIRE 1040 304 1008 304 WIRE 1088 304 1088 288 WIRE 1088 304 1040 304 WIRE 1232 304 1232 240 WIRE 1232 304 1088 304 WIRE 1264 304 1232 304 WIRE 944 320 944 192 WIRE 1088 320 1088 304 WIRE 832 368 832 160 WIRE 880 368 832 368 WIRE 1200 368 1200 128 WIRE 1200 368 1152 368 WIRE 608 400 608 384 WIRE 736 400 736 384 WIRE 944 432 944 416 WIRE 1024 432 944 432 WIRE 1088 432 1088 416 WIRE 1088 432 1024 432 WIRE 1024 448 1024 432 WIRE 1152 496 1088 496 WIRE 1264 496 1152 496 WIRE 1152 544 1152 496 WIRE 1024 560 1024 544 WIRE 1024 656 1024 640 WIRE 1072 656 1024 656 WIRE 1152 656 1152 624 WIRE 1152 656 1072 656 FLAG 608 208 0 FLAG 672 96 Clk FLAG 1472 128 A2Q FLAG 1472 0 A3!Q FLAG 1472 -48 A3Q FLAG 1472 160 A2!Q FLAG 608 400 0 FLAG 736 400 0 FLAG 608 304 Vcc FLAG 960 192 Vcc FLAG 736 304 VEE FLAG 1072 656 VEE FLAG 1040 304 Sine SYMBOL digital\\dflop 720 -96 R0 WINDOW 3 8 168 Invisible 2 SYMATTR Value TD=2n TRise=2n VHigh=0.95 SYMATTR InstName A2 SYMBOL digital\\dflop 1024 -96 R0 WINDOW 3 8 168 Invisible 2 SYMATTR Value TD=2n TRise=2n VHigh=0.95 SYMATTR InstName A3 SYMBOL voltage 608 96 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value SINE(0.5 0.5 4e6) SYMBOL cap 992 208 R0 WINDOW 3 23 53 Left 2 SYMATTR Value 304pf SYMATTR InstName C1 SYMBOL ind 1072 192 R0 SYMATTR InstName L1

SYMATTR SpiceLine Rser=24 SYMBOL npn 880 320 R0 SYMATTR InstName Q1 SYMATTR Value 2N2369 SYMBOL npn 1152 320 M0 SYMATTR InstName Q2 SYMATTR Value 2N2369 SYMBOL npn 1088 448 M0 SYMATTR InstName Q3 SYMATTR Value 2N2369 SYMBOL voltage 608 288 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V2 SYMATTR Value 5V SYMBOL voltage 736 400 M180 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V3 SYMATTR Value 5V SYMBOL res 1008 544 R0 SYMATTR InstName R1 SYMATTR Value 15k SYMBOL voltage 1152 528 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V4 SYMATTR Value 3 TEXT 616 -184 Left 2 !.tran 0 50u 45u 200p TEXT 616 -216 Left 2 ;'I-Q Quadrature Generator With Sine Output TEXT 984 -184 Left 2 !.options plotwinsize=0 TEXT 984 -160 Left 2 !.options nomarch TEXT 1144 480 Left 2 ;AGC Voltage TEXT 1296 344 Left 2 ;PLL and Amplitude TEXT 1336 376 Left 2 ;Control TEXT 1272 224 Left 2 ;Sine Wave To Buffer Amp TEXT 1304 48 Left 2 ;Quadrature Signals to TEXT 1304 72 Left 2 ;I-Q Detect TEXT 1184 -184 Left 2 !.options numdgt=15 RECTANGLE Normal 1488 528 1264 272 2

Here is the PLT file:

[Transient Analysis] { Npanes: 4 { traces: 1 {524293,0,"V(sine)"}

Y[0]: (' ',1,3.8,0.2,6.2) Y[1]: ('m',1,1e+308,0.0004,-1e+308) Volts: (' ',0,0,1,3.8,0.2,6.2) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {268959748,0,"V(a3q)"}

Y[0]: (' ',1,-0.1,0.1,1) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,1,-0.1,0.1,1) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {268959747,0,"V(a2q)"}

Y[0]: (' ',1,-0.1,0.1,1) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,1,-0.1,0.1,1) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {524290,0,"V(clk)"}

Y[0]: (' ',1,0,0.1,1) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,1,0,0.1,1) Log: 0 0 0 GridStyle: 1 } }