I have a Schmitt trigger inverter oscillator setup like below. I don't remember how it first looked on the scope, but I'm showing pos and neg spikes. At one point in changing values I did wire it up incorrectly and power it for a minute or two.

Are the spikes normal?...Or did I damage the chip? I'm using Philips 74HC14, by the way. And I don't have a another one to try. The spikes are more common than the screenshot shows. Neg spikes are about 2V; pos spikes are about 1.5V.

 

Those spikes are not uncommon due to stray wiring circuit capacitance and inductance.

 

Can I remove them? Diode clamps didn't seem to work.

 

Your circuit has no supply voltage, maybe it is out of spec?
Your schematic shows an antique TTL IC but your text here says a modern 74HC IC.
The output is supposed to be squarewaves, not spikes so replace the IC.

 

Power supply is a regulated 5.0V. Image is something I grabbed from the interwebs to avoid having to draw my own. My circuit is the same...I'm just using a Philips 74HC14.

EDIT: And there's 100nF across the 74HC14 power supply pins.

 

Oh, I may have miscommunicated. Those "spikes" in the oscilloscope image are actually low duty cycle pulses. By "spikes", I mean the excursions above and below the power supply. I probably should have used a different time base for the screenshot.

 

Better image. Spikes on second pulse.

 

Are you using a 10:1 probe with the probe ground connected to circuit ground?
If not, that's how you need to measure the waveform.

 

Yes.

 

This is on a breadboard, by the way. Will moving it it to a ground plane PCB get rid of the spikes? I don't want to solder this up yet, as I'm still messing with component values.

Better image:

 

Can I remove them? Diode clamps didn't seem to work.

If you are using a real fast Schottky diode (in other words with short leads) and the spikes are still there across the diode, its a 99% chance that what you are seeing is a grounding problem. Using one of those plastic plug-in breadboards is inviting spikes like that. This sort of thing is better done on a copper clad board with solder.

 

I'll solder it up and order some more chips then. Thanks. Seems like prototyping directly on copper clad could be a challenge, though.

 

I find it interesting that what looks like an over / undershoot is not on every edge,

 

Right, it isn't on every edge. If I let the oscilloscope run it shows up often enough. But for the screenshots I have to take several grabs to get one with spikes. Which is why I'm confused about it being breadboard capacitance. I do use quality breadboards, but I'm still going to solder it up to see.

 

You could try measuring with a ground probe tip:

I have used it in the past to reduce observed ringing on fast edges.
You can make your own with some copper wire.

 

Breadboards cause most problems people have with circuits:
1) the connections are intermittent.
2) Each wire and each row of contacts is an antenna that picks up interference and produces positive feedback oscillation.

Maybe your scope does not have a shielded cable.
Maybe the scope probe is not frequency compensated correctly.
Maybe your scope is digital that misses showing some of the very narrow pulses.

 

Right, it isn't on every edge.

That may be an oscilloscope problem.
A AG noted, digital 'scopes can miss short spikes at lower sweep speeds due to the finite number of samples they take.
Increase the sweep speed so you are only looking at one edge and see if that helps.