Ramp Type Oscillator with Bottom Release

I think he means he wants the bottoms of the sawtooth to be "pointy" and without dead-time where the current source current shoots thru the discharge switch while the oscillator bottoms out and starts to recover. Which means you have to snap your discharge transistor off a little prior to the cap fully discharging.

A simple 555 + current source ramp-er "solves" this problem by only letting the ramp capacitor swing between 1/3rd and 2/3rds Vcc at the comparator inputs. A naive implementation using a comparator that could swing nearly to ground (whatever it is) would have to sense when the ramp was a couple mV above ground. Which is hard due to offset; if the offset is particularly bad your ramp might smack into the rail prior to the comparator actually firing

For a single-supply comparator with a P-structure input the acceptable common-mode voltage goes below ground, like up to -0.3 volts for an LM358. So the offset can too, it doesn't "know" your output can't actually invert.

I understand that, but who cares what the 'bottom' looks like if your other inputs, not able to accurately go that low, can't use it anyhow?

That's what I was hinting at--offsetting the whole waveform's voltage up to where it's within analog inputs' voltage-range. I've done that for a PWM circuit that had to adjust close to zero duty-cycle. (I made an oscillator that swung from 1/3rd to 2/3rds Vcc, followed by a comparator with a threshold adjustable from 1/3rd to 2/3rds of Vcc.)

If we knew what Jim was actually trying to do (e.g. make a ramp for a PWM, vs. ramp for an A/D, vs. ....) there would likely be all sorts of alternate approaches.

Cheers, James Arthur

Sometimes when people use a sawtooth turns out a triangle wave would also do - just wondering if this application might be one of those? Of course only worthwhile if triangle wave is easier to make in this process.

piglet

1.2V Supply :-( Ramp is for timing events.

I need maximum swing. Upper side is limited by mirror headroom. I was trying to squeeze low side to as near zero as I can manage. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

             I'm looking for work... see my website. 

Thinking outside the box...producing elegant & economic solutions.

What's the approximate Vt of the MOSFETs on a chip like this? At that voltage and supply current seems like there's so little gm to work with that the ramp generated by a simple CCS is going to start to look super-sloppy long before it ever gets close to the positive rail...

Can you capacitively couple to that charge cap and, while it is being discharged, use that as a "rate-of-change" signal? When the rate of change drops off -> it's close enough to zero. That ought to work with a tiny on-chip capacitor because the ramp-down is fast.

--
Regards, Joerg 

http://www.analogconsultants.com/

This is the result of my own best-effort at the problem just using the

2N7002-type devices included with LTSpice:

The bottoms look nice but it's already getting pretty sloppy at around

4.3 volts as it runs into the mirror headroom, also circuit is at 5 volts and consumes ~3 mA

You certainly have enough dv/dt in that discharge ramp to provide a full ga te on voltage for a termination circuit that turns off quickly at the botto m, no picky thresholds required, although a little Schmitt action would be helpful. There's an old Tektronix patent for a feedforward technique used t o substantially sharpen those turnaround corners in triangle waveform gener ators.

[snip]

I'm seeking the CMOS equivalent of this bipolar phenomenon...

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

             I'm looking for work... see my website. 

Thinking outside the box...producing elegant & economic solutions.

Since there is an oxide layer I can't see anything static (DC) like that. You might be able to grab the dynamic gate current via Miller which goes away once Vds has arrived at the bottom and no longer changes. Looks like your negative ramp is fast enough to do that.

--
Regards, Joerg 

http://www.analogconsultants.com/

Late '60's ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

             I'm looking for work... see my website. 

Thinking outside the box...producing elegant & economic solutions.

Node (A) on this classic CMOS oscillator swings about [-.5 x Vdd .. 1.5 Vdd]. You probably don't need protection resistor R1 @ 1.2V supply, and you may have to replace Rt with something fancier if linearity requires, but perhaps this basic notion will inspire...

.------[Rt]------. | | |\ Ct | |\ | .---| >o---||---+--[Rl]---| >0---+---> | |/ (A) |/ | | | '--------------------------------'

Cheers, James Arthur

Well, that changes things some.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Had a dream last night. I think a (CMOS) cascoded current mirror will "collapse" and give me the signal that I seek.

As for oscillators, this is far superior...

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

             I'm looking for work... see my website. 

Thinking outside the box...producing elegant & economic solutions.

That solution is too pedestrian! :-)

[...]

--
Regards, Joerg 

http://www.analogconsultants.com/

The best solutions are always head-slappers... why didn't I think of that before ;-) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

             I'm looking for work... see my website. 

Thinking outside the box...producing elegant & economic solutions.

Sounds like a winner.

That loses the advantage of 2 x Vdd swing. You said you wanted "maximum" swing.

The extra swing also makes the original circuit more independent of supply and threshold drift, but yes, it does ping the input diodes with a few uA. I figured you wouldn't need gate-protection diodes though, not at Vdd=1.2V.

Cheers, James Arthur

I said I wanted a _ramp_, neither oscillator satisfies that. But my oscillator version is ultra-stable, proven in several chip designs (without the 2:1 attenuation... no ESD diodes internally, so flying above and below rails is Kosher). ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

             I'm looking for work... see my website. 

Thinking outside the box...producing elegant & economic solutions.