I know I read this somewhere

[Papa_lazerous]

It must be over a year since I made myself a Phase 90 and at the time I remember reading about it and what made it work....  I could sworn there was a technology of artticle over at Geofex that went right into the Phas90 but there isnt.

I am most interested in understanding the LFO side of things more as I have a good understanding of the rest.  Anyone willing to shed some light on some material for me to read?  all my googling on LFO's using opamps arent coming up with the same thing thats on the Phase 90 so its not really helping.

Cheers

[George Giblet]

Try,

opamp schmitt-trigger oscillator

http://ocw.mit.edu/NR/rdonlyres/2A11B939-5C27-4DBF-A358-40DBBAF37835/0/24_op_amps3.pdf

The two common types of opamp based schmitt-trigger oscillators.  The above is one, the other is the type that uses an integrator and is found in things like the Boss BF2 and CE2.

[Paul Perry (Frostwave)]

All I can say is, thanks for asking! because, i hadn't looked closely at it before - and now I DO look at it, it is much more tricky than I thought it would be.

Glancing at the circuit at http://www.geofex.com/FX_images/p180plus.gif
I could be completely wrong - I'm not an EE - but, my guess is that the 10uF is the one that charges & discharges. The .01 cap across the op amp is just to stabilise the op amp.

When the system starts, the 10u is discharges & the output of the op amp is +9 (from amplifying the + voltage at the junction of the 150K & 470K).
The 10u cap starts charging up (via the 500K speed control) & when the voltage on the 10u is greater than the voltage on the + input pin on the op amp, the output switches over to 0v (or, as low as it can go) and the cycle starts over.

[R.G.]

I think I have typed it in here a time or two.

Most LFOs are either the phase shift oscillator (not phaser!) type or the integrator-Schmitt trigger type which produces a triangle wave and a square wave. The P90 has the integrator-Schmitt type.

An integrator produces the summation of the input voltage times time. That's confusing to say, but easy to think about. If you put a voltage into an integrator, the output is just voltage times the amount of time the voltage is there. For instance, an integrator with a gain of unity, fed a voltage of a steady one volt starting ...now!... will have an output of one volt at one second, two volts at two seconds, three at three seconds, and so on until it hits its power supply limit, or until we change the input voltage. An integrator produces a ramp from a steady voltage with a slope proportional to the voltage.

If we feed an integrator a voltage of +1 and let it ramp for two seconds, then change the voltage to -1, the output ramps up to 2V, then ramps downward because the polarity of its input has changed, telling it to ramp down. If we let it go for four seconds, it gets to -2V. If we then change back to +1V at the input, it can ramp up for four seconds to +2; if we're really clever about feeding it the right voltage and flipping its polarity at the right time, we get a smooth ramp up and down - a triangle wave.

It's a true pain to do this manually just to get a triangle wave. So we use a trigger circuit. The trigger circuit watches the output of the integrator and feeds it +1V until the output gets to the trigger level, which we'll set at 2V for consistency. When the output of the integrator hits 2V, the trigger causes the input voltage to invert to -1V. That forces the integrator to ramp down. At the same time... here's the Schmitt trick... the trigger changes its own trigger threshold so that now it inverts when the integrator output gets down to -2V. Schmitt triggers can be made from one opamp or the equivalent and do this little trick, watching a voltage and flipping states when it gets to or outside some middle area.

That's the basics in a nutshell. An integrator ramps up or down depending on the polarity of the voltage it's fed. A Schmitt trigger watches that voltage and makes it change ramp direction when it gets to an upper or lower limit. In the P90, the integrator is replaced by a capacitor, which is a crude imitation of an integrator. The opamp stage in the LFO is the Schmitt trigger. It's arranged to feed its own output voltage, either full high or full low, into the capacitor through a resistor. The resistor is the speed pot, and it limits how much current gets into the cap, and therefore how fast the cap ramps up or down. Small resistance equals fast charging, bitg resistance means slow charging.

The Schmitt trigger thresholds are set by the resistors and bias voltages, and do not change. The integrator "gain" doesn't change. All that changes is how fast the speed pot lets the cap charge up and down. So you get an almost-triangle wave. If the cap was replaced by a proper opamp integrator, you'd have a more or less accurate triangle wave.

There are some fine points I glossed over in trying to explain the overview. (That's for the purists in the audience.)  But it is how it works. Tell me what still doesn't make sense and we can dig in there.

[Papa_lazerous]

Thanks RG, I think I got my head round what you are saying.  I am goign to google some now I know what its called and also read the above link and if I am still not sure I will probably ask somemore

[The French connection]

Hi! Are you refering to this:
http://www.geofex.com/Article_Folders/phasers/phase.html

Dan

[Papa_lazerous]

Hi! Are you refering to this:
http://www.geofex.com/Article_Folders/phasers/phase.html

Dan

Thanks but no I wasnt referring to that article, Its one I have read and understood before it was more about the oscillator type than how the phasing worked Thanks for checking that I might have missed it

On a plus note my copy of "the art of electronics" turned up from Amazon 2 days ago  188 pages read and digested and interestingly just started reading about opamps as integrators.....

I recommend it as a good read! I know there is a hookey pdf floating around 10 mins with that made me realise it was time to delete it and buy the book

[DougH]

I have a library copy of AoE in my office. I need to buy a copy for home, though. It really is a good investment. It's easy to read and understand and is a gold mine of information.

[Paul Perry (Frostwave)]

The only problem with the Art of Electronics is that the index could be better.
On the bright side, that means you have to read it right through at least once...