make LTSpice simulate switching spikes?

[cellularmitosis]

Hey guys,

I whipped up an ICL7660 circuit to generate a negative rail, but noticed it created significant switching spikes.

I thought I'd mock it up in LTSpice so that I could try simulating the effects of ferrite beads, etc, but using an LTC1044 (Linear's ICL7660), I don't see anything that looks like a switching spike.

Is there anyway to get these spikes to show up in LTSpice?

[TerminalJack505]

Are you sure what you saw wasn't just picked up by the scope's ground wire?  It can act like a nice little antenna and make it look like you have switching spikes of 10mVpp to 20mVpp. 

Use the low inductance ground attachment (usually a spring-like thing) and see if the signal looks less noisy on the scope.

Other than that you can have a similar issue where long wires on a breadboard can pick up similar noise. 

Also, breadboards have the issue of having a capacitive coupling type of effect for adjacent pins.

[T3sl4co1l]

Try looking at I(V1), and also setting V1 for, say, 10 ohms ESR, 20nH ESL: very crudely representative of a small electrolytic.

Or better yet, draw the components -- I hate that LTSpice hides all those properties from the schematic view, so you don't know at a glance what ESR/ESL you're playing with.

You aren't going to get much from as soft and slow a chip as that, and depending on whether they modeled shoot-through (if any) or not.

If you like, you can model your own CMOS inverter with a 2N7002 and BSS84.  On a supply of about 4V or more, you'll see plenty of shoot-through, to the tune of ~20 ohms shunted across the supply for whatever the rise/fall time is.  I suggest driving them from a PULSE source with, say, 10-50ns rise/fall time.

It's difficult to model a realistic, representative power supply -- every wire segment or trace acts like a little inductance, and all the bypass caps interact (think vibrations on a string with beads spaced out irregularly).  For best experience, do it in the real world.

Tim