Measuring crystal oscillators with scope ?

[Dannyx]

Good day folks. Today I had another (playful ?) argument with my stubborn co-worker about crystal oscillators - you know, those metal cans. Is it possible to measure these with a scope to see if they output a waveform ? What is the correct way to hook up the probes ?

It all started when he upheld the idea that a stereo CD/tape deck system doesn't display anything on its VFD because of a bad crystal oscillator. Now I personally haven't worked with these x-tals extensively because there was nothing much I could've done to them, plus I rarely encountered one failing and as such don't know too much about them.

He hooked up the GND probe to chassis GND and the positive probe of the scope to either pin of the x-tal and the line remained perfectly flat, leading him to conclude that it must be bad...I however disagree with both:
1) I don't believe that's the correct way of doing it, since the x-tal I've seen in most schematics is "isolated", meaning he should've measured BETWEEN the pins of the x-tal, as oppose to chassis GND
2) You can't use any regular old scope and probe to do the job - a high impedance probe is required from what I read (?)

Is any power required at all or can you just measure the cans alone in/out of circuit ?

Am I the idiot, or are we BOTH idiots ? I don't mind - it's all about learning.....even if requires getting called an idiot first

[eccerr0r]

Re: Measuring crystal oscillators with scope ?

Yes, if your oscilloscope has enough bandwidth, the crystal oscillator is powered, you should be able to see the waveform.

However if you have a plain crystal by itself -- not an oscillator -- then you need to have an amplifier that completes the oscillator circuit. If you're testing in-circuit, one end of the crystal is usually connected to an amplifier output that you can easily view through an oscilloscope provided the circuit is likewise powered.

If it's flatline, whether the amplifier is bad, crystal is bad, or some other condition not met is an exercise for the reader.

[Dannyx]

Re: Measuring crystal oscillators with scope ?

Ok, I was curious about the cans themselves if simply probing them should cause anything to show up on a scope. Got my facts a bit mixed up, since there's "oscillator" and "quartz/crystal". Oscillator is the circuit and "quartz" is the part...am I right ?

In some schematics, I've seen the crystal connected to two pins of a uC, so it seems to be "floating" relative to anything else...probing should be done with respect to GND in that case or between the two pins ? I distinctly remember a "jungle IC" for a Philips chassis CRT TV had this setup: two pins on the IC JUST for the quartz.

UPDATE: they changed the x-tal and it still doesn't work. When I say "they", I mean my colleague and his interns he was assigned to...corrupting young minds, that old bastard

[eccerr0r]

Re: Measuring crystal oscillators with scope ?

I forgot about another category, crystals used as filters, which are not part of oscillator circuits. These are a category of their own.

If both pins of the crystal is connected to a microcontroller or IC directly, likely that IC is doing the amplification and one of the two pins will likely to also have a low impedance signal that can be viewed through an oscilloscope, referenced to the IC's common point.

[Dannyx]

Re: Measuring crystal oscillators with scope ?

Just had a peek at what the boys are doing: it's a ceramic device with 3 legs they're using, just so we're on the same page. They're using a test circuit like the one shown here, but my colleague in his wisdom suggested that C1 and C2 are not required as they're built into the crystal element (no doubt fooled by the dotted line surrounding the crystal and caps)...curious to see what happens

Perhaps they ARE built in and I'm the idiot again

[stj]

Re: Measuring crystal oscillators with scope ?

ceramic resonators with 3 legs do contain the caps,
metal cans dont.

also, some microcontrollers have internal caps - but not many.

as for the scope, the capacitance of the probe is enough pF to fuck the output frequency a bit - but it should still oscillate (i hope)

[Dannyx]

Re: Measuring crystal oscillators with scope ?

Ok, so this one is a blue thing with 3 legs, so there should be caps in there. Initially, one of the kids failed to tie the middle pin to GND, but even after they did, it was still flat-lining...then again, I'm not sure the scope is operating properly OR the IC they used is fine.

Either way, I highly doubt the oscillator caused the display to fail. Smells like a bad uC to me somewhere...

[stj]

Re: Measuring crystal oscillators with scope ?

well i need to see the circuit!
and the frequency of the part.

[Dannyx]

Re: Measuring crystal oscillators with scope ?

well i need to see the circuit!
and the frequency of the part.

If you're referring to the test circuit they attempted to replicate, I linked to it in one of my previous posts. If you're talking about the circuit in the stereo, I haven't had a chance to have a go at it myself yet, but perhaps I'll at least pull the model so we can have a go with the service manual. If memory serves, it was something like a 10k crystal...maybe...take it with a grain of salt...

It's a Samsung from what I gather. Having a room full of junk is perfect for interns to mess with...even I have a go with them sometimes to learn stuff, like the scope thing right here

[stj]

Re: Measuring crystal oscillators with scope ?

10k ???
KILOhertz??

i doubt that - they may be mistaking it for a filter

[Dannyx]

Re: Measuring crystal oscillators with scope ?

10k ???
KILOhertz??

i doubt that - they may be mistaking it for a filter

Is that too much or too low ?
I probably got the value all wrong, but the part itself is certainly a crystal: blue with 3 prongs - I saw it myself

[budm]

Re: Measuring crystal oscillators with scope ?

May be 10.7 MHz FM IF filter?

[Dannyx]

Re: Measuring crystal oscillators with scope ?

I don't think so - it's on the board with the VFD, so I don't think it's for FM.

[R_J]

Re: Measuring crystal oscillators with scope ?

Based on the picture. It's a ceramic resonator and is likely bad.

Quartz crystals are more accurate and temperature stable than ceramic resonators. They are not the same but do a similar function and can basically be treated the same

[Dannyx]

Re: Measuring crystal oscillators with scope ?

This is what the part itself looks like, although it's blue and smaller. Not sure if the circuit I linked to is in operation as part of the stereo's display or it's just for testing purposes. Either way, from what I've learned so far is that these 3-legged devices (ceramic resonators) are NOT the same as THESE metal can things, are they ? The terminology gets a bit confusing there...

[budm]

Re: Measuring crystal oscillators with scope ?

That VFD display board probably has micro-controller (System Controller) so the Xtal is probably used by that IC.

[redwire]

Re: Measuring crystal oscillators with scope ?

Ceramic resonators are lower cost and tougher than quartz crystals, but not as accurate on frequency. The 3-pin parts have built-in loading capacitors.

A scope probe has capacitance and loads down a crystal or resonator oscillator. The MCU osc. has an input and output, and the input is more sensitive to loading. You can see a 100mVpp sine wave at a few MHz and know it works OK, on the the freq. function on a DMM to look for life.
Once the DMM or scope probe is connected, it adds say 20pF which can either make an oscillator stop, go lower in freq., or a dead osc. can start working because it's needing more load capacitance than the circuit has alone.

To say a VFD (MCU) is dead due to a bad crystal or resonator is a bit silly, it's a rare failure and not the greatest guess IMHO.

[llonen]

Re: Measuring crystal oscillators with scope ?

There are several different ways I would go about this, the simplest being a near field probe and pre-amplifier attached to a scope or spectrum analyser with the device under test powered. Directly connecting to the Xtal / resonator will likely result in the cessation of oscillation due to loading or capacitance effects.

[Dannyx]

Re: Measuring crystal oscillators with scope ?

To say a VFD (MCU) is dead due to a bad crystal or resonator is a bit silly, it's a rare failure and not the greatest guess IMHO.

Correct. I'm not even sure if the thing is working AT ALL. I have no idea what they were trying to do to it but I haven't heard it playing anything to begin with, so the display is probably the LAST thing these guys should've worried about...perfect example of getting priorities in the wrong order. Although I did hear them saying that board gets some 20+v on it somewhere...bad system control IC is still my guess without even going anywhere near the device and relying solely on what I heard around me

If someone has a cold and a stuffy nose, you don't check their nose first, do you ?

[Dannyx]

Re: Measuring crystal oscillators with scope ?

UPDATE: apparently it's a 4Mhz ceramic oscillator.