Crystal Oscillator Problem

The circuit layout near a crystal needs to be tight and tidy, the IC it's connected to needs good decoupling on it's power supply too.
Read the IC data sheet, there may well be a paragraph on the oscillator layout or Google application notes for your IC, there will prob' be good info' there.
Also, how do you know it's a 9Mhz sine wave? Connecting a 'scope can stop correct operation of the osc' especially if it's already marginal (or your scope may be aliasing!)

The capacitance of the circuit board tracks count towards the capacitnce needed by the crystal, so you may need to adjust the values (usually go to a smaller C)
Del

First of all thanks for the reply! I don't think its an aliasing problem, I am using a scope with a bandwidth of 200 MHz. But the datasheet does not include any specifications on the crystal, since I read all the datasheet. I am currently using a breadboard for all the connections since I am still in the designing part, do you think the breadboard can alter the frequency? (since the breadboard acts as an antenna at high frequencies!)

I know its a 9Mhz since I measured its frequency on the scope. Also when I read the clock that got through the IC was also 9 MHz :( I can't figure it out

I bet the breadboard is adding a load of 'C'.
try bending the legs of the IC upwards out of the breadboard and soldering the crystal straight across the legs. Make sure you have a 0.1uF across the power pins of the IC too.
Del

will it help if I use a veroboard (stripboard) instead of using the breadboard? since now I am only testing and simulating and also the IC is set on a breakout-board and is not connected to the breadboard.

Bread boarding HF is difficult. Typically you can use blank copper clad PC board as a ground plane and then apply or glue strips or pieces of the same PCB to solder components to. We did this type of bread boarding with a lot of RF stuff.

We simulate everything in LTSpice (its free from Linear Technology website) and then go directly to a PC house for prototype boards. Express PCB charges $100 for a 4-layer production style board 2.5" by 3.8". You get three boards for that price. That includes silk screen and solder mask. They have a free PC layout and schematic capture software that is a breeze to use.

Also, if you want a good solid 27.0000 mHz signal, use an oscillator IC instead of a crystal to feed the input of your chip. You will get a nice TTL square wave with plenty of fanout.

Just try what I said in #3!

Your crystal is probably an overtone type and too much added capacitance or other motional parms make it oscillate on its fundamental frequency. Did you follow the original design including crystal type or you're improvising? Not all crystals are created equal. S.M.

i got the crystal from a store and did not include any datasheet so i dont know the exact load capacitance. i added 2, 30pF capacitors in the normal topology connected to ground and it seems to have a "non noisy" wave but still i am reading 9.07MHz i also think its something related to the fundamental or ovrtone frequency since 9 is a multple of 27 sincemy xtal is 27MHz

Well, I have experience chasing my tail. You may be having a similar experience. You need to recheck the way you are measuring and what you are measuring with to be sure you are seeing what you really think you are seeing.

Sometimes a rose is just a rose, too. You may simply have the wrong part that either got dropped into the wrong bin at the store or was mismarked by the manufacture.

I'd start by half or third of the capacitance, 30pf is kinda high for 27 Mhz clock. Also crystal's and IC's datasheet can give you some clues. But there's always the chance your IC clock gen cirquit is static damaged. In most cases one of iC pins connected to crystal can be used as an external clock input for testing, just mind the input levels. S.M.

As (almost) always I've read all members replies, including OP's clarifications, long after posting.(LOL). And don't take me wrong, what I said stands, but only in normal conditions. Breadboarding 27 Mhz (!) is no normal condition. Hell, most HF and up crystal osc stages, can't handle the 15pf+ added from a 10X or some 300pf from a 1X probe tip as was mentioned. Added to the high capacitance, God knows how long your cap leads are, and how close to digital ground the crystal is refered. A photo could give us more info. S.M.

_{http://www.northcountryradio.com/PDFs/070302007.pdf http://www.4timing.com/termcrystal.htm http://www.imec.be/esscirc/ESSCIRC2002/PDFs/C02.01.pdf http://www.westfloridacomponents.com/mm5/graphics/F10/CSA358MG300DETF.pdf http://ebookbrowse.com/eecs242-lect22-phasenoise-pdf-d54720490 http://global.kyocera.com/prdct/electro/pdf/add_pdf/hand_e.pdf http://www.jauch.de/ablage/med_00000619_1193753698_Quartz%20Crystal%20Theory%202007.pdf http://www.crystek.com/documents/appnotes/VCXOarticle.pdf}

^RANDOMNESS

I had a look around and found a chip with XIN, XTAL and there are no capacitors needed. Search and find a datasheet on the chip you are using and see what it recommends for the Xtal.

From memory way back usually the input capacitor is about half the output capacitor.

If you CRO does not have 10:1 probes then when you connect it to the device it will slow it down usually.