NNTP-Posting-Date: Wed, 06 Nov 2002 23:40:40 -0600
From: "Dave VanHorn"
Subject: Re: Problems with cristal as clock for atmel AT90S8535
Date: Thu, 7 Nov 2002 00:38:55 -0500
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers
X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly
> serie? are they not paralel
> | |
> -X- (cristal)
> | |
> = = (capacitors)
> | |
> ----------- ground
They are in series for the purpose of calculating the load capacitance.
If the xtal spec says 22pF, (typical) then you need twice that, minus the
actual load C presented by the chip and wiring, say 5pF for a guess, so
maybe 39pF on each leg.
> > > -----| 22pF
> > > X2 |-------|-------][---|
> > > | cristal 8Mhz |
> > > X1 |-------|-------][---|
> > > -----| 22pF |
> > > GND
> > Does the ground lead go directly, and only, to one of the 8535 ground
> yes ,the same ground that is connected to the ground of the atmel (pin 11)
Good dog! :) It does make a difference.
> which cristal should i use?
One you have a spec sheet on.
> an oscilator with 4 pins (5 V,ground, output ) ?
> and connect that one to X1 ?
That's an oscillator, not a crystal.
> > Probe the xtal out pin only.
> with a other cristal oscilator?
> but then a DIL version:
> like this
> i use now someting like this:
Oscillators and crystals aren't the same thing.
Oscillators have crystals (or something like that) in them.
Oscillators take 5V power, and ground, and produce an acceptable clock
They have at least three active pins.
Crystals are passive devices, and can't make an output signal without some
more circuitry, which is included on the chip.
> > > so i assume the atmel has an amplifier build in for the cristal.
> > > I thougt the two small capacitors were there to make teh cristal
> > > out of it self and it did not need any extra signal to make a signal.
You COULD design a crystal to run with no caps, but it would be very flaky,
because one chip might have a little more or less stray capacitance than the
rest. The external large caps make the chip's variability proportionally
smaller. If a 5pF cap on the chip varies by 1pF, that's 20%. If 39+5pF
varies by 1pF, that's only about 2%.
> so i do not need to have any voltage input for the kristal?
don't think of it that way. It's not an active circuit, it's simply a slab
of rock that has the interesting property of having exactly a 180 degree
phase shift at X MHz, when in the right circuit.
Have a look here