A bit of help with a crystal oscillator

A little searching turned up a few.

I found several guides on oscillators that refer to MOSFETs as being suitable "transconductance" devices.

Why would a MOSFET be a problem?

--

  Rick C. 

  - Get 1,000 miles of free Supercharging 
  - Tesla referral code - https://ts.la/richard11209

I did a google search. MOSFET Colpitts do exist. For example, see

"Loop Gain of the Common-Drain Colpitts Oscillator"

However, the problem with a Colpitts is you need two pins. One for the gate, the other for the source or drain whichever configuration is used. If you put the capacitors on the chip, you only need one pin for the external inductor that is also connected to ground. But I have always heard putting capacitors on an ic are very expensive due to the area required.

An alternative is to use a factory-trimmed RC oscillator if the frequency requirements are not too strict. Another alternative for precision work is to mount the oscillator externally and use one pin to drive the ic.

Design is full of tradeoffs.

z f r -

If

l g

I believe the inductor is not used when the crystal sets the frequency.

s

If an oscillator circuit (not any discussed here) is a bit more optimized f or a single digital pin, the input threshold crossing can be detected, the output then can be stimulated with an impulse of opposite polarity (equival ent of AC coupling) and on the opposite phase of the crossing the same thin g can happen with the opposite polarity impulse.

Oscillation can be initiated by simply driving the circuit with impulses of approximately correct timing. Rather than this being a stimulus in the AC domain, this is really to get the DC operating point established. If this were a dedicated output a series resistance would be provided for the outp ut internally on the chip.

This is very much analogous to driving a clock pendulum with an impulse at mid swing. The Shortt?Synchronome free pendulum clock did this wit h a gravity lever that was controlled electrically by the synchronized slav e pendulum. Other clocks do this using electromagnets, often controlled by a watch crystal oscillator.

--

  Rick C. 

  + Get 1,000 miles of free Supercharging 
  + Tesla referral code - https://ts.la/richard11209

Steve Wilson wrote:

Here is a Schmitt XO along with the PLT file. The oscillation is chaotic at first until you apply enough drive power to get to oscillate through the crystal.

The drive power may be excessive for most crystals, so you probably would revert back to a standard inverter. But the sym shows it can be done, it's just not practical.

Version 4 SHEET 1 1008 716 WIRE -32 144 -64 144 WIRE 224 144 -32 144 WIRE 640 144 336 144 WIRE 704 144 640 144 WIRE 752 144 704 144 WIRE -64 368 -64 144 WIRE 224 368 -64 368 WIRE 640 368 640 144 WIRE 640 368 304 368 WIRE -64 464 -64 368 WIRE 96 464 -64 464 WIRE 352 464 160 464 WIRE -64 544 -64 464 WIRE -32 544 -64 544 WIRE 80 544 48 544 WIRE 112 544 80 544 WIRE 240 544 192 544 WIRE 256 544 240 544 WIRE 352 544 352 464 WIRE 352 544 320 544 WIRE 400 544 352 544 WIRE 448 544 400 544 WIRE 512 544 448 544 WIRE 640 544 640 368 WIRE 640 544 592 544 WIRE -64 560 -64 544 WIRE 448 560 448 544 WIRE -64 656 -64 624 WIRE 448 656 448 624 FLAG -64 656 0 FLAG 448 656 0 FLAG 80 544 R5L1 FLAG 240 544 L1C1 FLAG -32 144 OscIn FLAG 704 144 Out FLAG 400 544 C1C2 SYMBOL cap 256 560 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 WINDOW 39 -28 32 VBottom 2 SYMATTR InstName C1 SYMATTR Value 0.1p SYMBOL cap 432 560 R0 SYMATTR InstName C2 SYMATTR Value 45p SYMBOL cap -80 560 R0 SYMATTR InstName C3 SYMATTR Value 45p SYMBOL res 608 528 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R2 SYMATTR Value 750 SYMBOL res 208 384 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 0 56 VBottom 2 SYMATTR InstName R3 SYMATTR Value 100k SYMBOL ind 96 560 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L1 SYMATTR Value 10m SYMBOL res 64 528 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R5 SYMATTR Value 25 SYMBOL cap 160 448 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C5 SYMATTR Value 4p SYMBOL 74hc14 272 96 R0 SYMATTR InstName U1 TEXT 40 -16 Left 2 !.tran 0 2m 0 1n TEXT 40 -48 Left 2 ;'Pierce Schmitt Crystal Oscillator TEXT 544 -16 Left 2 !.options plotwinsize=0 TEXT 752 -16 Left 2 !.ic V(OscIn) = 0 TEXT 544 8 Left 2 !.inc 74hc.lib TEXT 160 256 Left 2 ;May be overdriving the crystal RECTANGLE Normal 32 112 32 112 2

[Transient Analysis] { Npanes: 4 Active Pane: 2 { traces: 1 {524290,0,"V(oscin)"} X: ('m',1,0,0.0001,0.001) Y[0]: (' ',1,0.9,0.3,4.2) Y[1]: ('m',1,1e+308,0.0002,-1e+308) Volts: (' ',0,0,1,0.9,0.3,4.2) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {34603012,0,"I(R5)"} X: ('m',1,0,0.0001,0.001) Y[0]: ('m',1,-0.003,0.0005,0.003)

Amps: ('m',0,0,1,-0.003,0.0005,0.003) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {524293,0,"V(c1c2)"} X: ('m',1,0,0.0001,0.001) Y[0]: (' ',1,0.6,0.3,4.2)

Volts: (' ',0,0,1,0.6,0.3,4.2) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {524291,0,"V(out)"} X: ('m',1,0,0.0001,0.001) Y[0]: (' ',1,-0.5,0.5,5.5)

Volts: (' ',0,0,1,-0.5,0.5,5.5) Log: 0 0 0 GridStyle: 1 } }

On Friday, 8 November 2019 14:21:20 UTC-8, Steve Wilson wrote: ...

...

I've seen one device that does use a single pin for a crystal connection in a 433MHz OOK receiver.

kw

Interesting. Thanks.

It says the oscillator is a Colpitts and the series caps are internal. But it doesn't say of the oscillator is a MOSFET or bipolar. I would suspect bipolar to maintain process compatibility with the rest of the circuitry.

The question is, does the oscillator start.

A normal oscillator is merely a noise amplifier with a frequency selective positive feedback. For each time the original white thermal noise around the feedback loop is amplified, the original white thermal noise is amplified only in the filter passband increasing the amplitude.

For each time around the loop, the spectral peak becomes narrower and narrower, much narrower than the filter Q would suggest, finally ending up as a single spectral line. The amplitude can't grow forester, either the amplifier gain is reduced or the signal is clipped.

While initially starting to amplify the white noise, it is essential the amplifier is in class A or AB. After a few iterations, when the amplitude has grown close to saturation, the amplifier can fall into class C and oscillate happily forever.

If the amplifier is initially deeply in class C, it doesn't amplify the thermal noise and the oscillator will not start. The Schmitt trigger makes starting even more unlikely.

Many faulty oscillator constructions starts simply by the operating voltage turn-on transient, putting the amplifier momentarily into class AB. Such oscillator may start when powered from battery and the DC voltage is applied by a switch, since the start transient is fast, but might not start when a mains powered device is plugged into mains, due to the slow charging of the storage capacitors.

A Schmitt RC oscillator doesn't work that way. It doesn't need noise to start.

Funny, they always work for me.

Again, a Schmitt RC oscillator always starts. It must oscillate because neither state is stable, and drives hard towards the opposite state. Linear analysis makes no sense.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

It depends. There are good and bad implementations of that. You can write one in software that measures the almost exact bit time, and then uses that to set either a software delay or programs the baudrate downcounters, that also works with non standard baudrates,

I was thinking[tm] OP is using 4800 Bd, that is very low these days, only context I use 4800 Bd is with GPS modules. Those have a crystal and the baudrate is very precise, Never had any problem with that with PICs running from internal oscillator, even outside in the cold, my drone's auto-pilot does just that

My GPS clock / radiation logger does that:

Also have a Raspberry Pi 2 with /dev/ttyAMA0 getting time and position from a GPS module at 4800 Bd, but the raspi clock is crystal based.

Most serial projects here run on 19200 or 115200 Bd. To make it easier for myself most send a 'hello' at power up so you can see if your baudrate matches or not.

The simplest form of a relaxation RC oscillator is implemented with a neon lamp, which has a higher turn-on voltage and a lower turn-off voltage, so there is hysteresis.

It is hard to imagine a crystal oscillators as an relaxation oscillator.

That's a really good explanation, I think.

snipped-for-privacy@downunder.com wrote

Steve Wilson wrote

Also Class A might be a problem if you are building a CPU which has very low power modes. You can't really throw away say 1mA.

For those who like to play with LTSpice, one could unroll the feedback loop by creating a few dozen modules containing an amplifier stage, an attenuator (to simulate feedback factor and filter losses) followed by a bandpass filter. Connect these modules in series.

At the input of this long chain, connect a resistor to ground,creating a thermal white noise density of -174 dBm/Hz. Observe the amplitude and spectrum after each stage.

If in a stage the attenuation is grater than amplifier gain, no output is produced from the chain. If the amplifiers are class C no output is produced from the chain.

This relaxation oscillator assumes that the capacitance to ground is sufficiently large and that the DC voltage to the inverter is applied fast enough so that the voltage across C1 is less than Vil and hence the output is "high" starting to charge the capacitor through R1.

The equivalent circuit of a crystal is a series resonance circuit consisting of Cs and L at Fs with a very small parallel capacitance Cp forming the parallel resonance Fp with the inductance L.

The question is, does the parallel capacitance Cp with some internal feedback resistor R1 parallel create sufficiently fast relaxation oscillations to excite the crystal parallel resonance Fp to produce a voltage swing larger than the hysteresis and hence maintain oscillation.

Watch crystals use an inverter running off a tiny battery. They last for years.

Peter wrote:

The Schmitt has no stable state so it has no choice but to oscillate.

Getting it to oscillate at the crystal frequency may take a lot of drive power. This may exceed the crystal's capability. Here's an example. You need Helmut's 74hc.lib to run it.

Version 4 SHEET 1 1008 716 WIRE -32 144 -64 144 WIRE 224 144 -32 144 WIRE 640 144 336 144 WIRE 704 144 640 144 WIRE 752 144 704 144 WIRE -64 368 -64 144 WIRE 224 368 -64 368 WIRE 640 368 640 144 WIRE 640 368 304 368 WIRE -64 464 -64 368 WIRE 96 464 -64 464 WIRE 352 464 160 464 WIRE -64 544 -64 464 WIRE -32 544 -64 544 WIRE 80 544 48 544 WIRE 112 544 80 544 WIRE 240 544 192 544 WIRE 256 544 240 544 WIRE 352 544 352 464 WIRE 352 544 320 544 WIRE 400 544 352 544 WIRE 448 544 400 544 WIRE 512 544 448 544 WIRE 640 544 640 368 WIRE 640 544 592 544 WIRE -64 560 -64 544 WIRE 448 560 448 544 WIRE -64 656 -64 624 WIRE 448 656 448 624 FLAG -64 656 0 FLAG 448 656 0 FLAG 80 544 R5L1 FLAG 240 544 L1C1 FLAG -32 144 OscIn FLAG 704 144 Out FLAG 400 544 C1C2 SYMBOL cap 256 560 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 WINDOW 39 -28 32 VBottom 2 SYMATTR InstName C1 SYMATTR Value 0.1p SYMBOL cap 432 560 R0 SYMATTR InstName C2 SYMATTR Value 45p SYMBOL cap -80 560 R0 SYMATTR InstName C3 SYMATTR Value 45p SYMBOL res 608 528 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R2 SYMATTR Value 750 SYMBOL res 208 384 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 0 56 VBottom 2 SYMATTR InstName R3 SYMATTR Value 100k SYMBOL ind 96 560 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L1 SYMATTR Value 10m SYMBOL res 64 528 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R5 SYMATTR Value 25 SYMBOL cap 160 448 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C5 SYMATTR Value 4p SYMBOL 74hc14 272 96 R0 SYMATTR InstName U1 TEXT 40 -16 Left 2 !.tran 0 2m 0 1n TEXT 40 -48 Left 2 ;'Pierce Schmitt Crystal Oscillator TEXT 544 -16 Left 2 !.options plotwinsize=0 TEXT 752 -16 Left 2 !.ic V(OscIn) = 0 TEXT 544 8 Left 2 !.inc 74hc.lib TEXT 160 256 Left 2 ;May be overdriving the crystal RECTANGLE Normal 32 112 32 112 2

[Transient Analysis] { Npanes: 4 Active Pane: 2 { traces: 1 {524290,0,"V(oscin)"} X: ('m',1,0,0.0001,0.001) Y[0]: (' ',1,0.9,0.3,4.2) Y[1]: ('m',1,1e+308,0.0002,-1e+308) Volts: (' ',0,0,1,0.9,0.3,4.2) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {34603012,0,"I(R5)"} X: ('m',1,0,0.0001,0.001) Y[0]: ('m',1,-0.003,0.0005,0.003)

Amps: ('m',0,0,1,-0.003,0.0005,0.003) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {524293,0,"V(c1c2)"} X: ('m',1,0,0.0001,0.001) Y[0]: (' ',1,0.6,0.3,4.2)

Volts: (' ',0,0,1,0.6,0.3,4.2) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {524291,0,"V(out)"} X: ('m',1,0,0.0001,0.001) Y[0]: (' ',1,-0.5,0.5,5.5)

Volts: (' ',0,0,1,-0.5,0.5,5.5) Log: 0 0 0 GridStyle: 1 } }

The traditional crystal oscillators use the grid-to-cathode diode of a tube or gate-to-source diode of a JFET as a means of generating an amplitude-dependent bias on the control electrode. A MOSFET needs an external diode to make it work.

--

-TV

It will oscillate with any value of capacitance or any value of resistance (as long as gate DC current doesn't make too much offset in the resistor) and will oscillate no matter how long it takes for the supply to come up. It has no stable state.

I've done CMOS schmitt oscillators with no explicit capacitor.

Schmitt inverters also oscillate with no R or C: just connect input to output. Add a long trace, a PCB delay line, to better define the frequency.

Right. That's the issue.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

LT Spice doesn't handle nonlinearity in a frequency sim, and parts are noiseless in a time domain sim. Time-domain oscillators sometimes need a kick of some sort.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics