1
\$\begingroup\$

I have a crystal oscillator (full can) running at 14.31818 MHz. When I put that on my scope, it shows about 14.3~ MHz. My scope is a Rigol DS1102E 100 MHz so I understand that you need about 10x the frequency of what you're measuring so I would need about 140+ MHz scope to accurately see the 14.31818 MHz frequency. So I get that. Anyway, oscillator seems to be working.

Next, I wanted to "square out" the wave a little by running it through a SN74AC04 HEX inverter. So when I tie the output of my oscillator to A0 I get nothing but a flat line on O0.

So thinking maybe the '04 was bad, I swapped it out. Same thing.

Next, I decided to back down the frequency. So I plugged in my SIGLENT SDG805 waveform generator into A0 (I unplugged the oscillator).

I then tried different frequencies. Starting at 1 KHz all the way up to 5 MHz (the maximum I can generate with the SIGLENT). The scope shows a very nice square wave for the most part. 5 MHz is a little off but the scope shows pretty much the same thing as if I had a 5 MHz crystal.

Anyway, why does the 14 MHz oscillator show a flat line? The specs of the '04 say (from what I understand) that it should be able to handle that frequency.

I'm running everything at 3.3v.

The specs say tPLH and tPHL at 3.3v should be about 10ns and 9.5ns respectively. 10ns should allow 100 transitions per micro-second right?

So wouldn't I get around 6-7 transitions at 14 MHz?

Anyway, what am I doing wrong?

Thanks!

\$\endgroup\$
7
  • \$\begingroup\$ Do you have access to any 74xx parts from faster families such as LVC, AUP, etc? \$\endgroup\$ – Krunal Desai Dec 28 '15 at 20:49
  • \$\begingroup\$ Unfortunately, no. I have 74HC and 74LS versions. 3.3v is a must as I'd rather not use level converters or resistors for the number of data lines in my design and I believe AV is 3.3v tolerant. Thanks \$\endgroup\$ – cbmeeks Dec 28 '15 at 20:52
  • \$\begingroup\$ I think the AC part isn't fast enough but I'm having trouble thinking of a clear explanation of it. The LVC parts are 3.3V and are quite speedy -- there's also ALVC, and AHC: ti.com/lit/sg/sdyu001aa/sdyu001aa.pdf \$\endgroup\$ – Krunal Desai Dec 28 '15 at 20:55
  • \$\begingroup\$ 74AC is pretty fast, certainly fast enough at 5V but you need to double-check its timings at 3V. I'd be checking the DC conditions at the input, in case the osc had an AC coupled output or something daft. \$\endgroup\$ – Brian Drummond Dec 28 '15 at 21:49
  • 1
    \$\begingroup\$ At 3.3 V, a 74AC04 should be able to handle typical frequencies of about 40 MHz. \$\endgroup\$ – WhatRoughBeast Dec 28 '15 at 23:13
1
\$\begingroup\$

How are you plugging this all together? On a proto plug board?

If so your problems could be related to this. Proto boards can cause a lot of problems when the frequency starts to get up there over a few megahertz. The problems relate to coupling between adjacent buss nodes and the larger than normal circuit capacitance that the buss tracks being to the party.

\$\endgroup\$
0
\$\begingroup\$

That you want to "square out" your oscillator suggests that it is not providing a logic-level signal, but rather something like a sine wave. Such a sine wave is, of course, referenced to zero, so the highest amplitude portion of the waveform simply never gets high enough to drive the inverter low. Try this sort of circuit

schematic

simulate this circuit – Schematic created using CircuitLab Make sure to use a ceramic cap. And be careful to make your connections, especially ground, as short and thick as humanly possible.

\$\endgroup\$
1
  • \$\begingroup\$ These cans typically output logic type signals. The apparent problem with the direct output is most likely probe bandwidth. \$\endgroup\$ – Chris Stratton Dec 28 '15 at 21:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.