5MHz TTL clock oscillator giving 20Hz. What's wrong?

[robertbaruch]

Hey all,

I'm using this TTL oscillator, specifically the MXO45-3C-5M0000.

It seems simple enough. So, I get out my benchtop power supply, set to 5VDC, and a breadboard.

• Pin 1 is on the lower left, looking down at the package. That's the enable pin, so hook it to +VCC.
• Pin 14 on the upper left is power, hook to +VCC.
• Pin 7 on the lower right is GND, hook to GND.
• Pin 8 is the output, connect to oscilloscope.

Power up with 5VDC, get... 20Hz out. WAT.

Replace, try another one. Get... 150Hz. WAT!

Okay. Clearly there's something I don't know about this thing. Does anyone have an idea what I'm doing wrong?

[void_error]

Did you use a ~100n ceramic decoupling cap between +VCC and GND? Lots of weird things can happen on a breadboard if you don't use one.

[robertbaruch]

*sigh* Never mind. Apparently I turned the time setting on the Rigol that I'm using down to 1 ms/div, and the input capacitance at that point was enough to drop the oscillator into some other mode. When I moved the time up above 20 us/div, I was finally able to see high-frequency oscillation, and at 50ns/div I can clearly see bang-on 5MHz.

[Andy Watson]

*sigh* Never mind. Apparently I turned the time setting on the Rigol that I'm using down to 1 ms/div, and the input capacitance at that point was enough to drop the oscillator into some other mode.

I suspect that what you were seeing was aliasing. Try putting the scope in a different capture mode.

[uncle_bob]

*sigh* Never mind. Apparently I turned the time setting on the Rigol that I'm using down to 1 ms/div, and the input capacitance at that point was enough to drop the oscillator into some other mode. When I moved the time up above 20 us/div, I was finally able to see high-frequency oscillation, and at 50ns/div I can clearly see bang-on 5MHz.

Hi

That's not an input cap issue. As mentioned in another post, digital scopes are not ideal broadband frequency counters. They only take just so much data and only do so just so fast. When you slow the sweep speed down, the sample rate drops as well. This gets you into trouble if you try to look at a very fast signal. It will down convert it into a perfectly good looking signal ... at the wrong frequency.

Bob

[robertbaruch]

Ah, now that makes sense. Thanks!

[tggzzz]

Apparently I turned the time setting on the Rigol that I'm using down to 1 ms/div, and the input capacitance at that point was enough to drop the oscillator into some other mode.

The input capacitance is invariant, unless you have a very strange and faulty scope!

Always learn to use your instruments by playing around with a known signal. Make sure you can relate what you see to what's inside the instrument. All instruments lie in one way or another, so you need to learn each one's foibles.