Crystal oscillator question

obitus1990

Well-known member
Joined
May 25, 2020
Messages
2,546
Reaction score
2,114
Location
Metairie, Louisiana
Should both legs of a crystal exhibit the same frequency when checked with a scope? I have a partially working board, in which one of the oscillators on the board (there are 3 total on it), when checked with the scope, outputs 10MHz on one leg, yet the other leg just goes high to 5V and stays there. I am assuming that both legs should be the same, as the other two crystals have the same frequency on both of their legs.

Thanks!
 
Crystals, not oscillators, are high impeadance devices and the circuitry is sensitive to capacitive loading. Because of this it's common to "kill" the oscillation when you probe on the side of the crystal that connects to the input of the buffer. On the side that connects to the output of the buffer MOST of the time you can probe without effecting the circuit but even on that side there's the potential to effect the circuit. The best place to probe is on the buffer or IC that's NOT connected directly to the crystal. If you have a good clock at that point then the crystal circuit is working fine.

Oscillators are crystals with the buffer INSIDE the can, these require a power pin to power the internal buffer. Since these are buffered internally you can probe on the output without effecting the actual crystal.

Hope that helps.
 
Prior to posting, I swapped out the crystal for a new one of the same 10MHz frequency. With the new crystal, both legs produce a 10MHz wave on my scope, versus one leg just going high like the original did. Did I use a different type of device (an oscillator versus a crystal) here?
 
Prior to posting, I swapped out the crystal for a new one of the same 10MHz frequency. With the new crystal, both legs produce a 10MHz wave on my scope, versus one leg just going high like the original did. Did I use a different type of device (an oscillator versus a crystal) here?
Like I mentioned the circuit is "sensitive" to capacitive loading, meaning depending on many variables like the capacitance of your probe, the "Q" (quality factor) of the specific crystal, S parameters of the buffer device, value of coupling capacitors/resistors etc. you MAY cause the oscillation to stop but you also MAY NOT so it's kind of "hit and miss".

Crystals and oscillators aren't interchangeable since the circuitry around them is different, so I don't think you changed to an oscillator.

If you want to verify that's what is happening you can use one probe on the leg that you are seeing the signal on then with a second probe connect to the "non working side" you should see the signal on the "working" side go away since you are stopping the oscillation from happening.

The thing to understand is if you are seeing a valid signal on EITHER side of the crystal then the circuit is working and you can move on to downstream circuitry.
 
Last edited:
Back
Top Bottom