Ramp generator help

momentary toggle switch SPST or DPST

You want a MOM-OFF-MOM switch, I think. Look in DigiKey.

Do you want it to rise to VCC as if through a low-pass, or do you want it to follow A*e^(t/a) until it hits VCC, then level off? The former is easy, the latter less so.

Bitrex-

From an analog computer course I had back in the 60s, I think you are looking for an integration circuit. Your input is a square wave, your output a linear ramp up to the point where the output reaches its limit.

You should be able to do it with an op amp with series input resistor and capacitor feedback. If you want an output current ramp rather than voltage, you may need a second op amp with series input resistor and your load as feedback.

Fred

Which exponential? 1 - exp(-t) cap discharge, or exp(t) with saturation?

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

One unipolar case, in an smps, developed a linear ramp (constant current into capacitor) that was converted to second capacitor current, giving increasing dv/dt during the ramp interval, but a reset pulse was available to discharge all caps between periods. Counting on a discharge the starting ref would mean very low duty, as these discharging strays degrade fidelity. The starting ref on each node also wasn't zero, but could be quickly acheived by the reset hardware, due to similar compatible unipolar offsets.

A bipolar application would be more than twice as complicated, discrete component-wise.

RL

It sounds like you want to make a resistor divider (string of resistors) feeding an RC delay. The momentary switch (button or toggle types are available) would be wired to shunt one resistor of the divider string, when it's in 'closed' position.

If you're feeding current into the PNP transistor's emitter, that's a low impedance load (and the RC delay part ideally feeds a HIGH impedance load). You might want to consider component values carefully. I'm envisioning an LM13700 current programmed by a 0-5V signal, with the PNP doing the level translation.

So I'd like: engage momentary toggle, exp(t) with saturation, release toggle, 1 - exp(-t) cap discharge.

HuH?? Ramps up _exponentially_? Do you really mean to say the curve is like a^x ? Good trick..driving a iron/ferrite core toward saturation seems the only decent way to do approximate that.

Yes; correct question WRT poor description.

assume 0

The exponential rise can be accomplished by linear upramp drive of Vbe of a transistor, and looking at the collector current. You'll want to limit the base current, though, and there will be thermal effects.

The exponential-time driven gain can be a useful function if you want to determine (for instance) an AC signal level; just ramp amplifier gain by a few dB per millisecond, and check your stopwatch when the first saturation occurs. That's a dB-to-digital scheme. Not sure why the controlled fall response would be required, though.