small stone "colour switch"

[markphaser]

The boss Boss PH-2 mode switch what does this mode switch do? is it feedback paths? i don't have the schematic so i don't know where the paths of this mode switch does

The small stone "colour switch" what does this do in the LFO circuit? it seems to change the LFO circuit but what does the colour switch do to the LFO?

The Boss PH-3 has a rise and fall time for the LFO how did they do this? putting a diode on the output of the LFO to just get the rise?

Whats unidirectional phasing?

[markphaser]

Does the "colour switch" change the DC offset in the small stone LFO?

[George Giblet]

> Does the "colour switch" change the DC offset in the small stone LFO?

From my notes:  It makes the sweep wider. ie the notches start at lower frequencies and move up to high frequencies compared to the normal mode.  As a side effect the sweep rate is a little slower.

[markphaser]

Thanks for the information

how does it change or make the sweep wider?

How would i change a MXR phase 90 LFO to get the sweep wider like the small stone "colour switch"?

The colour switch must change the sweep range? but how does it do this?

[bioroids]

It does not make the sweep wider, as far as I know. It only introduces feedback to make the peaks "peaker" and incidentaly, also changes the lfo speed range (slower?)

Miguel

[puretube]

it`s a 2pdt switch: 1p increases LFO output amplitude,
the other switches on audio FB.

[markphaser]

introduces feedback to make the peaks "peaker"

It increases the feedback peaks in the LFO? how does it do this please?

How does the "colour Switch" increase the amplitude? is it extra gain in the LFO circuit?

[stm]

it`s a 2pdt switch: 1p increases LFO output amplitude,
the other switches on audio FB.

So it sort of "flangerizes" the circuit:icon_question:

[puretube]

it`s a 2pdt switch: 1p increases LFO output amplitude...

this part of the switch takes care of the LFO voltages

[puretube]

...
the other switches on audio FB.

this part switches feedback of the phase-shifted signal in and out of the audio-circuit

[Mark Hammer]

The MXR Phase 100 has a 4-position switch that gets the 4 combinations of more/less resonance and wider/narrower sweep.  The Small Stone "color" switch reduces this to 2 combinations. In one position it produces a narrower sweep range with less regeneration/feedback/resonance (pick your favourite word for it).  In the other position it produces a wider sweep (generally sweeper a bit higher, rather than starting lower as was suggested), and introduces a bit more feedback/regeneration.

If you look at the Piedrita swchematic at Tonepad (a Small Stone clone that is identical but substitutes different OTA chips), you will see that after the last phase shift stage based around IC4, the phase-shifted signal follows two paths: one via a 27k resistor to the output mixing stage, and the other via a 3k3 resistor, 27k resistor, 0.1uf cap and either a 270k resistor or a straight wire path back to the input just after the 100k input resistor.

At the same time, the other set of contacts either shunts the 22k and 15k resistors, or places a 1k8 resistor in parallel with the 1k resistor.  The circuit built around IC5 is the LFO.

If a person wanted to, it would be an easy and smart thing to split the two functions of the Color switch into two toggle SPDT switches: one for sweep width, and the other for regeneration.

Alternatively, you can substitute a 500k pot for the 270k feedback resistor.  The Color switch either presents zero ohms or 270k along the feedback path.  As the resistance goes down, the amount of feedback/regeneration is increased.  Replacing the 270k fixed resistor with a 500k variable resistor will let you go from practically zero feedback to the maximum value for the Small Stone.  If a little more feedback or playing with the edge of oscillation appeals to you, try reducing the 27k resistor before the 0.1uf cap in series with a 22k value or if you're daring an 18k part.  You could also consider replacing the 4k7 resistor in the 3k3/4k7 network (which serves as a preset attenuator) with a 6k8 unit.  Do not replace the 27k series resistor AND the 4k7 resistor to ground.

As for making the width variable, I'm afraid this is beyond my knowledge of OTA based oscillators.

[George Giblet]

> t`s a 2pdt switch: 1p increases LFO output amplitude,

.... which makes the sweep wider.


Your original question related to the LFO.  As stm has detailed the other part of the color switch adds peaks to the audio through feedback.

Many phasers user feedback to add peaks (check out Ibanez units for example).  I'm sure the Boss unit would do the same, it's a common trick.

One thing about the Small Stone is the feedback is done in a slightly different way: the clean signal path is "modified" by the freedback.

[markphaser]

"Small Stone is the feedback is done in a slightly different way: the clean signal path is "modified" by the feedback"

Doesn't all feedback "modify the clean or dry signal? how does the small stone "modify" the clean signal different?

[markphaser]

 sweep width variable, I'm afraid this is beyond my knowledge of OTA based oscillators

Whats the different between a OTA Based oscillator VS a regular LFO oscillator?
the OTA based oscillator output Current?
the regular LFO oscillator output voltage?

How do we change the sweep width in most common LFO oscillators? it seems the swing voltage rails would make the LFO sweep width different because of the swing range?

[George Giblet]

>Doesn't all feedback "modify the clean or dry signal? how does the small stone "modify" the clean signal different?

The idealized model of a phaser is that one path is clean and second path is modified using an all-pass filter.  The all-pass filter characteristics are modulated by the LFO, the clean path is unaffected.  When feedback is used the feedback is applied to the all-pass filter, the clean path is again unaffected.   The two paths are then mixed at the end.   A phaser doesn't *have* to follow this structure there are actually other ways of achieving the same final result.

>  Whats the different between a OTA Based oscillator VS a regular LFO oscillator?

The main difference/advantage is the OTA based LFO can easily produce a parabolic shaped LFO output - as opposed to the linear/triangle outputs produced by opamps etc.  A parabolic LFO output sounds more natural and smoother.   For the record you can get linear/triangle from an OTA.  And with non-linear devices like diodes you can get parabolic outputs from an opamp (see the Mutron biphaser).  The multiplying nature of the OTA provides a natural and predictable means of getting a parabolic sweep.

> the OTA based oscillator output Current?
> the regular LFO oscillator output voltage?

That's part of the internal behaviour.  While most OTAs are based on current outputs somewhere deeper in the circuit, usually the current output is converted to a voltage output with buffer.  So really you end-up with voltage outputs.  That's not the important difference.

>  How do we change the sweep width in most common LFO oscillators? it seems the swing voltage rails would make the LFO sweep width different because of the swing range?

Yes the voltage rails will change it but the voltage rails are limited, it's not the place to do it.  A triangle output LFO swings at less that the supply range.  The voltages that correspond to the peaks of the triangle output are set with a threshold in the circuit - more often than not a Schmitt-Trigger circuit is used.  To increase the LFO swing you widen the two threshold points of the Schmitt-Trigger which makes the LFO output go higher/lower.

The side effect of *just*changing the thresholds is that the LFO speed will drop - that's because the rate the triangle ramp is increasing is fixed and now you have told it to go higher and it takes longer to to that.   So, to compensate to have to modify the timing part of the circuit to speed it up, for example by decreasing the timing capacitor.

You probably need to study the two opamp Schmitt-Trigger/Integrator LFO circuit.  There's plenty of these on the Web and you should be able to find some theory on it.

[markphaser]

Thanks George Giblet alot for the information

LFO types:
                  #1.) two opamp Schmitt-Trigger/Integrator LFO circuit- mostly outputs a linear triangle waveform
                  #2.) Phase shift oscillator LFO(univibe)- mostly outputs a sine waveform
                           the phase shift network in the negative feedback in the oscillator section sets the range and width?

                  #3.) OTA oscillator LFO- outputs a parabolic shaped LFO output

Why do they use a parabolic shaped LFO for OTA's?
what does the parabolic shape differ from a sine wave phase shift oscillator?

The parabolic shape waveform would move/shift the notches/peaks up and down different than a sine wave (univibe) phase shift oscillator LFO?

Phase shifters move the notches/peaks up and down in frequency or ampitude?

a flanger moves the notches/peaks in Time not up and down but left to right "shifting" the notches/peaks

 

[markphaser]

parabolic waveforms look like:
http://crca.ucsd.edu/~msp/techniques/latest/book-html/node176.html

[George Giblet]

> Why do they use a parabolic shaped LFO for OTA's?

As I mentioned before the parabolic sweep sounds more natural.

Take this idealized case to see why:  When you have a linear sweep each say 0.1V of sweep corresponds to fixed change in notch frequency say 200Hz independent of where the LFO is at.    When the LFO makes a 0.1V change and a  notch is at low frequncies say 100Hz, a 200Hz change put the notch frequency at 300Hz; a 3 times the starting frequency.  When the notch is a higher frequency say 1000Hz, a 0.1V LFO change moves the notches to 1200Hz; only 1.2 times the starting frequency.  Your brain roughly interprets frequency changes on a log scale.   Therefore the 0.1V change at 1000Hz is perceived as a small change compared to that at 100Hz, you would need to move the notche from 1000Hz to  3000Hz to get the same perceived changed as 100Hz to 300Hz.

The parabolic output is like a set of hills next to each other, similar to like a full-wave rectified sine-wave.  In reality the hills are inverted.     What the parabolic LFO does is stretch out the amount of time spend in the low frequencies, ie. the rounder part of the hill, and less time in the higher frequency part, ie the spikey joins in the hills.

> what does the parabolic shape differ from a sine wave phase shift oscillator?

The parabolic shape is asymmetrical ie. has a narrow spikey part and a wider round part.    It spends more time with the notches at low frequency.  The sine is smooth on the upper and lower part of the hills but it doesn't balance out the times spent at high and low frequencies. The sine-wave makes smooth sweeps compared to the triangle but it still suffers the problem surging sound because the notches get moved quicker at low frequencies.  Mind you there's no "correct" LFO some people like the surging sound.

> The parabolic shape waveform would move/shift the notches/peaks up and down different than a sine wave (univibe) phase shift oscillator LFO?

Yes it's different, as explained in the first paragraph.

> Phase shifters move the notches/peaks up and down in frequency or ampitude?

Phaser change the frequency response over time.  The thing which is changing is frequency based not amplitude.  However, the frequency response is not flat, it has notches, and as the notches move you will hear some form of amplitude change if your signal is around the notch frequency.

>a flanger moves the notches/peaks in Time not up and down but left to right "shifting" the notches/peaks

The phaser does the same it moves the notches left/right on the frequency response as the LFO sweeps.

In terms of broard brush stokes there is a lot of similarity between Flangers and Phasers.

> http://crca.ucsd.edu/~msp/techniques/latest/book-html/node176.html

Yes!

[markphaser]

Thanks George Giblet for the help,your time and information about this

What causes that LFO surging sound? Throbbing? breathing LFO sound?

"To increase the LFO swing you widen the two threshold points of the Schmitt-Trigger which makes the LFO output go higher/lower"

Notch depth- means the mixture of the dry signal with the time delay frequencys notched out. The more the notches the more the notch
                   depth. Mostly this is the summing/mixing stage or the FETS or LDR used can change the Notch depth

Notch Width- means the frequency's thats are notched out, this is set by the phase shift filter

LFO offset- changes the range of the swing of the LFO. So example of the LFO output is 5voits p/p with zero DC offset the range of the
                 swing is 2.5volts Peak sweeping the FET or LDR.
                 If we add Positive DC offset of 1volt then the sweep range changes the "Range Width" ? because the FET or LDR are going to have a different Swing/sweep range
                 If we add Positive DC offset of 1.5volts then the sweep range changes ?
                 If we add Negative DC offset of -1.5 volts then the sweep range changes ?

Manual Parameter- this replaces the LFO with a static delay time, the notches/peaks are not moving.
                            If we add Positive DC offset to the Manual parameter then the notches/peaks change because the Range has
                            changed. If we add Negative DC offset to the Manual parameter then notches peaks changes the range.
                            The DC offset shifts the notches/peaks range up or down

If the LFO "sweeps" the Notches up and down the DC offset "shifts" the notches up and down also but whats different?
they both move the notches up and down but how they do it is different

LFO outputs a 5 volt p/p its going to swing/sweep the FET resistance
When adding positive or negative DC offset how does the FET or lamp or LDR change from having DC offset?

Thanks for the help and knowledge

[George Giblet]

> What causes that LFO surging sound? Throbbing? breathing LFO sound?
The rapid changes of the notch frequency (and associated pitch shifting) at low frequencies.

>  Notch depth- means the mixture of the dry signal with the time delay frequencys notched out.

Yes.  The notch depth is simply how deep the notches go on the frequency response plot.  You get deep notches when the "gain" of the all-pass filter bank is unity (=1) when the total phase shift is 180deg, 540deg etc.

> The more the notches the more the notch  depth.

No.  The number of notches is independent of the depth - different things.  The number of notches is determined by the number of stages.

> Notch Width- means the frequency's thats are notched out, this is set by the phase shift filter

Yes.

> LFO offset- changes the range of the swing of the LFO.

The primary intention of the LFO is that it changes the *position* of the notches.  In practice the range (or span) of the sweep is affected but that's not the main purpose (LFO offset is like the manual Knob on a flanger).

> Manual Parameter

> If the LFO "sweeps" the Notches up and down the DC offset "shifts" the notches up and down also but whats different?

There's two cases here.  The first is where the LFO is completely removed and the all-pass circuits are controlled by a manual control - this is the case where the notches don't move over time.  The second is a manual control which works with the LFO, this is basically an LFO offset which lets you move the notch positions but the sweep is still active.    As an idealized example, one notch might cover 300Hz to 1000Hz.  When the LFO is added the sweep will cover 600Hz to 2000Hz (double frequencies) for a logarithmic control but For the linear case the notch will instead cover 600Hz to 1300Hz (adding 300Hz).

> When adding positive or negative DC offset how does the FET or lamp or LDR change from having DC offset?

It behaves as descibed above.  However,  the amount of voltage offset you can add is limited by the device or circuit you are controlling.  A JFET works over a very small control voltage range say -3V to 0V.  Everything happens in that range.  If you want to add offsets that will need to be small like 0.5V to 1V for example. A +5V to -5V offset just isn't going to work here.   Each device has is own operating range, and each will require a *different offset voltage range* to achieve the *same result*.