Crystal oscillators / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Crystal oscillators

1 2 next

Move Thread

LAN_403

Marko

Fri Oct 12 2007, 12:40PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi all,

I just came into a need of a stable squarewave source in low Mhz.
Crystal oscillators are not something I'm familiar with and I would really want to know more about how they work.

I wanted to use a discrete quartz. First thing I lashed up was a crystal over a single 74hc14 gate - it worked, but stability was poor, sometimes it wouldn't oscillate before I touched it by a finger.
Used 4Mhz crystal, at most, 12Mhz worked.
I couldn't get the 32.768kHz watch crystal to work. Only briefly and noisy while I was actually touching the input - it would probably require careful biasing like 74LS04 gates did -see next..

I tried several different caps between crystal and ground (10-100pF) and some pullup resistors, none seemed to help much.

I was very unhappy with the output waveform from the IC too.
Output has to high impednce, rings a lot even with scope probe capacitance and duty cycle seems impossible to keep even.

My biggest problem is that I hardly understand how these things work at all.

I tried constructing an oscillator of 74LS04 gates which I hoped for to be faster, but I just couldn't get it to oscillate.

After half an hour fiddling with potentiometers and nearly exactly copying this circuit

I made it to work, and was stable, but waveform didn't look any better than HC14 version.

I know that it's bad to use logic circuits in linear mode like this, but I rather wouldn't use any special components here.

I tried to construct discrete pierce oscillators which output sinewave but hardly any worked unless it was a directly copied circuit with exact crystal.

What I want is to build an oscillator with discrete crystal with available components that will output at least a decent square wave of few Mhz into scope probe load.
I noticed micro controllers use comparators at their Xtal input - but I don't think I have any that fast.

The crystal appears to be usable both in series and parallel resonant modes. Most oscillators are parallel, though.

I would greatly appreciate all tips regarding this, and some designs I could use.

chip output; my scope could be too slow for this, but still I don't think I'm having duty cycle increased to 60% for that reason. I'm using two buffer gates here.

probe in this pic is not in it's place

ShawnLG

Fri Oct 12 2007, 01:10PM

Registered Member #286 Joined: Mon Mar 06 2006, 04:52AM
Location:
Posts: 399

I have read somewhere that a 74AC04 is recommended when using it as a crystal oscillator. Also put a 1M ohm resistor in parallel with the crystal to start the oscillation on turn on. Sometimes you would need to use a small pulldown capacitor around 22pf to 33pf on the input an output inverting pins to help stop higher frequency operation. One thing about this type oscillator is that it will not be 50-50 square wave. You will need to double your target frequency and divide it through a D-flipflop to get a 50-50 square wave.

Tom540

Fri Oct 12 2007, 03:09PM

Banned on 3/17/2009.
Registered Member #487 Joined: Sun Jul 09 2006, 01:22AM
Location:
Posts: 617

I've used crystal oscillators with 74HC14's before. Put the crystal in parallel with the inverter and 1M ohm resistor across the inverter just like Shawn said. I used 15Pf caps on mine but it depends on the crystal. Thats all you need. It works for the 32.768 crystals too which I used for IR detectors on robots. Since I'm bored at work I'm going to draw it for you.

Avi

Fri Oct 12 2007, 04:24PM

Registered Member #580 Joined: Mon Mar 12 2007, 03:17PM
Location: Melbourne, Australia
Posts: 410

If you connect the output to the input of another gate it will possibly have a better waveform and also will not 'load it down' when you try to use the signal.

Sulaiman

Fri Oct 12 2007, 07:03PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Marko, this

is quite good

Marko

Fri Oct 12 2007, 07:30PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Tom, that's *exactly* what I did adn I can't get the watch crystal to work. I used 12pf caps.

It oscillates only when I'm holding it with hand and waveform is noisy - scope can't trigger to it, and on some spurious frequencies 500khz and up.

I'l twiddle some more with it this evening.

Sulaiman: I actually have the article from very long ago.

As I figured out, the capacitors ''tune'' the crystal which behaves like a sort of inductor?
What caps do you guys use for 32.768kHz crystal?

Also, I do always use buffer gates, as I already stated. The shown waveform is actually inverted oscillator output.

Bjørn

Fri Oct 12 2007, 08:10PM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

Crystals are quite sensitive so if you don't get one to work properly there is a chance of it being damaged.

This page has good iformation about 32768 Hz oscillators:

Tom540

Fri Oct 12 2007, 08:20PM

Banned on 3/17/2009.
Registered Member #487 Joined: Sun Jul 09 2006, 01:22AM
Location:
Posts: 617

Hmm, it has been a long time since I've tried it but I have an old board sitting on top of my bench as we speak with this circuit In it. I'll fire it up tonight and double check the circuit. It uses a 40KHz crystal on it.

You do need the output going through another inverter. I got the 15pF caps value form the crystals datasheet.

By the way I remember someone saying that some crystals are meant to operate in harmonic mode. They might be squirrelly or something.

Amateur-Scientist

Fri Oct 12 2007, 11:18PM

Registered Member #1015 Joined: Fri Sept 21 2007, 06:43PM
Location:
Posts: 14

Have you looked at the SG-8002 from Epson? I've used these in the past and they work great. Digikey has them for about $5. You just tell them the frequency you want in the notes when you order.

Marko

Sun Oct 14 2007, 09:37PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

I always use at least one additional buffer gate; but still duty cycle is always uneven and waveform is ringy. Flip flop definitely helped there a lot!

BTW, I had no clue how accurate cap values need to be for those watch crystals. Thanks guys for now ^

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob