LM555: Want functional LM555/TLC555 spice model, or help with transition from square wave output to sinusoidal output.

Will,

1) I could not find any LM555 or NE555 timer models.
2) I often use TLC55 model instead. It works well in TI-Tina for me.
3) No model other than one that models every component in LM555 separately would be helpful to model this type of oddity.

Wow, that is really strange behaviour. LM555 has so much hysteresis that I see no way it could produce a sine wave.
Do you have any device pin waveforms showing this behaviour?

Thanks for your quick response, Ron. I kind of assumed the LM555 wouldn't have a model (or you probably would have it on your website's page for that component ;  ). As far as using the TLC555 model, I also assumed that it would work but TI-Tina must have a default or specially defined VT parameter that LTSpice does not have. Usually, I have no problems importing a TI model into LTSpice, as LTSpice will run any standard PSpice subcircuit model, but I guess this sim is different enough or something.

I'm going to ask IT to install TI-Tina on my computer, and hopefully they'll get around to it before the universe ends.

We're troubleshooting the connections the timer makes. Sorry about the quality, and also sorry about this technically not being the direct output of the 555 timer (this is coupled through the field director's transformer). Approximate frequency of the 555 timer is ~3.3kHz, and the sine frequency seen is ~20kHz (but 20kHz is the frequency of motor commutation, and is most like EMI. You can see it in the first capture anyways, as those enormous spikes).

I'll give you some application information in case it helps you think of what could be causing this:

The 555 timer feeds several demodulators and a high-power motor's field director. The sinusoid seen is consistent with the motor's turning, and since it's such a high power motor some EMI coupling is expected. The 555 timer's output starts out as a roughly square wave with a little distortion (above) with sharp spikes at the frequency the sinusoid will occur at. The 555 timer actually works, right up until some point (we haven't figured out the trigger that causes this and are still investigating), but eventually this near-perfect sinusoid at the motor frequency replaces the signal. And once the signal is replaced, the square wave behavior doesn't return. Is it possible to latch up the LM555 so that it's output goes high impedance? That effect would match what we are seeing here.

Did some digging. The parameter causing the simulation to hang up in LTSpice is VT: VT is the thermal voltage of a diode, calculated as VT = k * T / q, where k is Boltzman's constant, T is the diode temperature in °K, and q is charge on an electron. VT is approximately 25mV @ room temperature.

Simulation is showing correct operation with noise injected onto GND, albeit the output is heavily distorted by the GND noise (as expected). It never reverts to a perfect or nearly-perfect sine, suggesting the issue is either noise coupling into something else OR the 555 timer is being disconnected.

I reiterate: is it possible to cause this 555 timer to latch up as high impedance?

555 timer is not latching up and stopping. We finally had a breakout cable made for the ship that let us probe the pin the 555 timer comes out on (instead of after the resolver's transformer). The 555 timer transitions frequency from 3.3kHz to 20-25kHz. When it does that, the 555's output is a little more distorted at the higher frequency (less than perfect square wave), coupled with the fact the resolver attenuates higher frequencies the measured value became a pure/semi-pure sine wave of the fundamental frequency.

Ron, you've already been so helpful, but can you tell me any way the 555 timer's frequency can be moved, other than by the 2 resistors and the capacitor that sets it up? I checked the parts, and they are military qualified resistors, 1% tolerance; and military qualified 10% tolerance BX dielectric capacitors -> all components can drift but these should both be very stable and incapable of jumping to 6x their values in one transition. And your forums seem to indicate that sometimes you can get the 555 timer to drift with load...?

Will,

The jpg images appear to be a broken in some, way. I can't open them.
A capacitor on the control pin helps with noise on VCC. The comparator (trigger and threshold) offset could be affected by EMI on those pins. If the EMI was large enough the effect could be significant. Looking the the timing cap ramp will show at what voltage the comparators trip.