Phaser suggestions

[Papa_lazerous]

Hi all a little advice if you will.........

Built the Phase 90 and I love it, a friend heard it and loved it too then he said he would love a phaser with the step mode that the PH-3 has.  I know you cant easily make the phase 90 step mode, then it got me thinking is there a phaser that has a step mode that is available for us cloners to build.  I know there are no PH-3 schems out there if there are I havent seen one yet.  Any suggestions on something I havent thought of yet?

Cheers and happy new year

[tskullt]

I really like the Ross version on Tonepad.  Nice clean layout, really smooth phaser.  Add'l stage board available!
m.

[Papa_lazerous]

I really like the Ross version on Tonepad.  Nice clean layout, really smooth phaser.  Add'l stage board available!
m.

I have the tonepad version up and running I am after a phaser that can do step mode though

[Mark Hammer]

Simple version is this.  Get or make a GGG Maestro FSH-1 Sample & Hold filter board.  Get or make a Tonepad Ropez board.  Where the control current feeds a 10k resistor that goes to pins 1 and 16 of the four phase shift stages on the Ropez, run an extra SPDT toggle or slide switch.  Now, take the point where Q3 and R20 meet on the FSH-1 board and connect that point, through the 10k resistor, to the phaser.

That's it.

Yeah, right...."that's it".

Seriously.  The PH-3 provides a digitally stepped mode that is functionally equivalent to an analog S&H mode.  The analog S&H mode simple samples a noise source at a predetermined rate, and uses whatever the noise value is at that instant to provide a voltage, which is converted into a control current.  Normally, the control current coming from the LFO on the Ropez smoothly glides up and down, while the S&H format converts it into a stair-step function.  All you need to do is switch between a stepped current source and a cyclical one.

Although it requires a bipolar voltage supply, the S&H portion of the FSH-1 is actually pretty simple, and could be built onto a small daughterboard that might easily fit in a 1590BB, along with a Ropez board.  The entire circuit is an 8-pin dual op-amp, four transistors, and a handful of passive components; easily feasible on a daughterboard the size of the Tonepad 4-stage expander board.

There are all manner of simple mods for the S&H function that you can find by perusing threads devoted to the FSH-1.  For instance, you can vary the degree of contrast between successive steps, the range of highest to lowest step, the initial starting point, and the "portamento" between successive steps.

[Papa_lazerous]

Thanks for the reply Mark, you may have hit the nail right on the head here.  Just want to get it striaght in my head first....

looking at the GGG Maestro FSH-1 schem, I build that complete.... Schem http://www.generalguitargadgets.com/diagrams/fsh1asc.gif

then I build the ropez complete..... schem http://www.tonepad.com/getFile.asp?id=99

then I just make the connections you speak of and I am rocking.  Man that sounds soooo simple!!  I see the Ropez can be modded for Trem and have 2 controls for "phase filter 1" and "phase filter 2" aswell.  I havent heard the Ropez.  I wonder why I didnt see it before building the phase 90 though.

I also see the FSH uses 3080's I am assuming thats just a bog standard CA3080 like in the Ross comp?  and the Ropex uses a dual OTA the LM13600.   I know Steve carries the 3080, but I cant see the LM13600 on his site, are they current and available?

I dont really understand much about the FSH-1 at present its a new one on me I will need to do some research, I assume I can just have the Ropez phasing and doing its own thing and stomp the FSH in and out to step? also do I need to build all of the FSH or just a part of it?  I think some of the circuit should be redundant?  as I wont be using the input and output on both circuits as one is just connected to the other. 

I am sorry for opening a can of worms after your reply Mark,  I am willing to do plenty of research and reading but I just dont know where to start on this so I figure if I ask you the obvious stuff and I can fill in all the blanks one I get a foundation of knowledge

[mdh]

I also see the FSH uses 3080's I am assuming thats just a bog standard CA3080 like in the Ross comp?  and the Ropex uses a dual OTA the LM13600.   I know Steve carries the 3080, but I cant see the LM13600 on his site, are they current and available?

That's because he carries the JRC part, NJM13600D.

[Ry]

Keep your wiring very neat around the FSH-1 portion.  Do a search and find some of the experiences of people building it.

[Papa_lazerous]

thanks mdh I didnt think of that 

I have been doing a little reading already on the FSH-1 one, and I can see it requiers an attention to detail.  still have the few questions from above though before I go building it

[Papa_lazerous]

Right I did a mass order from small bear last night, partly stocking up on a few parts and most importantly buying all the parts for the Ropez phaser and Maestro FSH-1 hybrid

Going to build both and try what Mark suggested in his post earlier in this thread.

I would still appreciate some input though as I am thinking some of the FSH-1 circuit will be redundant and not sure about the input output connections as there will be 4 in total as its really 2 effects joined at the hip.  Anyone willing to offer some assistance will be on my xmas card list for next year 

[Papa_lazerous]

Rather than put another post here I was going to modify the last but I cant 

Ok I put my thinking cap on and have gotten as far as I can.  I would really appreciate it if you could look at this for me, its a screen dump of the tonepad layout for the FSH-1 that I paint brushed.  the area inside the red box I made I consider redundant and not needed at all.

The arrow to the left leads onto the range control I see it gets a clean signal from the input buffer which I have removed so I thought I could now connect that point to pin 1 of IC1a ont he ropez layout which is just a buffer.

The arrow to the right is the point in the circuit that was described by Mark Hammer, which I will connect to one of the out poles of a SPDT switch and then on the other outer connected of the switch will be the output of the LFO on the Ropez and the centre connection on the switch will be where the LFO enters pin 1,16 of the IC's on the Ropez.

I understand what Mark was outlining before in his post and I like the idea, just dont want to build a whole FSH just to only be using part of it for what I need, it will make it too big and waste money on parts I dont need. 

ohhh the picture..........

[Mark Hammer]

You have correctly identified the parts of the FSH-1 that do the controlling.  Not all of those parts are needed, however, to accomplish the stairstep function.  You have also included the (modified) envelope follower from the FSH-1, in addition to the stairstep "generator" (S&H portion).  Envelope swept filtering is also interesting, so you are encouraged to include it as well, but it is not necessary for the stepped phasing in particular.  If all you want is that stepping, then you can eliminate everything that is brought into circuit when the slide switch you show in that screenshot is pushed to the left.  As an aside, the Line 6 Liqui-Flange pedal includes a stairstepped flange program in addition to an upwards and downwards envelope-swept flange.  All are quite nice, and I have no doubts that, although a 4-stage phaser can't be quite as sonically complex as a flanger, the outcome with a phase shifter will be musically useful.

Note that a Ropez, when modified for selecting between allpass and lowpass, can be converted into a 4-pole lowpass filter, simply by grounding all those 3300pf caps instead of tying them to the input of each stage.  I imagine that with some resonance added, a stair-stepped 4-pole lowpass would have some nice fat sounds.  Note that for the circuit to function well as a 4-pole filter, you would need to either cancel the dry signal as you would to get vibrato out of the basic Ropez circuit, or mix the dry signal well back using some sort of wet/dry blend control.  I'm also unsure of whether the resonance path needs to return to the 2nd stage or to the first stage, when in all lowpass mode.  Normally, in phaser mode, the resonance return comes back to the 2nd stage, but since we want to create peaks rather than notches, I'm thinking a switch needs to be in their to select return path.  Perhaps someone could confirm for me.

Whichever way you create it, the most robust effects using stairstep mode will be achieved when the resonance is fairly high.  Because the time spent on each step will be relatively short, one needs the extra resonance to provide the needed contrast for creating the "now I'm here...now I'm over here" effect.  In other words, fast abrupt changes in where the notches are being created requires a little extra push to focus those notches in the listener's mind.  There is a patch on my old Korg Pandora called "Aqua" that attempts to do a S&H effect, but the resonance is simply way too low.  It ends up sounding vaguely bubbly, but no more dramatic than that.

The simplest way to up the resonance is to replace the 27k resistor that goes from the resonance pot back to the 2nd phase shift stage with a 22k resistor and a 5k-10k trimpot in series to dial back from maximum resistance.  At 27k, assuming every part is on spec, the recirculated signal will be mixed in at unity gain.  With lower resistance values, there will be some gain added.  I've stuck a 24k fixed resistor on one of mine with no ill effects.  However, I note that when I switch from 4 to 8 stages of phase shift, the circuit is prone to oscillation when the resonance is too high (though not at all with 4 stages).  Happily, a) the oscillation takes a bit to kick in and it is highly likely the stairstep function will be moving too fast6 for that to occur, and b) all problems can be resolved by simply turning the resonance down.

Incidentally, there are many here whose skill and agility with the perfboard layout software is humbling.  If you have some time to spare, a layout for a daughter-board comprising the two op-amp circuit with the noise source and simple pulse generator would be deeply appreciated by many.

[niggez]

are there any audio samples of the tonepad ross phaser around?

[Papa_lazerous]

Thank you Mark I very much appreciate your indepth reply  

I was aware that I had included a few extra parts on the FSH-1 that weren't needed purely as I wasnt sure sonically what the outcome would be but figured it couldnt hurt and hell there'll be more knobs and switches!! haha

I could certainly tackle the layout you speak of, a daughter board with the noise source and simple pulse generator.   However I will be making my own layout for this project from scratch, I know there is a perfectly good layout at tonepad and GGG but where is the fun in that.  I like to make my own plus I can incorporate mods directly into the layout, I use Eagle and am actually getting quite good at designing pcb's on there.  I am hoping by morphing both projects together on one pcb and missing out the stuff I dont need I can cut down on size.  I will make the layouts publically available when finished though

[Mark Hammer]

I recommend for folks to take a peek at the L6 Liqui-Flange.  Though I harbour a soft spot in my heart for analog flanging, and I think the Flanger Hoax and FoxRox Paradox Flanger are excellent products, the Liqui-Flange gives you an idea of the many ways one can spruce up a basic effect simply by injecting other modes of modulation.

Most folks are accustomed to the sweep-up-sweep-down phase shifter, and a smaller number are perhaps familiar with the EHX Polyphase and some of the more recent boutique phasers that respond to envelope strength or hand movements, but there is a lot more that is possible.  An 8-stage or 12-stage Ropez merits being built into a rack chassis with all the trimmings, like stage-selection, modulation type/source, stereo capability, sum/difference capability, phase-filter capability, and so on.  The "Frankenstone" that Charlie Barth has up at the Moosapotamus site is simply the beginning. 

[jmasciswannabe]

Hey papa.....you try this out yet? Mark, right on with the rack box. i will be reboxing this for the third time if I add the step function.

[Mark Hammer]

Another fairly simple way to inject a little variability and unpredictability into your phaser is to build yourself two Ropez phasers and treat the second one as if it were a 4-stage expansion board with its own LFO.  So, the pads on the main Ropez board that would normally connect to the in and out of the 4-stage expansion board would now go to the in and out of the 2nd board.  The regen function on the 2nd board would be disabled, and the input stage and mixer section of the 2nd board would be ignored as well.

What you end up with is an 8-stage phaser with two simultaneous and unsynchronized sweep patterns going on.  The point where the combined amounts of phase shift total something that will produce a cancellation/notch will be largely unpredictable.  Once you go aperiodic, its hard to go back, but if you want, you can install a switch that selects between the output of one LFO, to sync all stages, or directs LFO A to this group of 4 and LFO B to that group.  You can also use that same switch to a) switch between two speeds, and b) combine the two LFO outputs to produce one master aperiodic control current that will behave a bit like a slow-n-smooth S&H.  Note that if one LFO is driving all 8 stages, you may want to reduce the value of the 10k resistor.  Alternatively, leave it at 10k and make note that sweep will not be quite as wide.  Conversely, if the two control currents are being summed, you may need to insert some additional current limiting.  Now that I ponder it, an A/B/Both/Dual sweep-pattern switch would be kinda cool.  I think I'll look into that on mine.

Looking at Osamu Hoshuyama's brief analysis of the Ross LFO circuit (which is essentially the same as the first Small Stone, but adapted from two 3094s to a single 13600) -

- it would seem that you can also tap a square wave from it as well.  He flips the two OTAs around so the equivalent pin would appear to be pin 14, the input pin on the 2nd OTA.  The diagram he presents suggests that the first OTA outputs a squarewave which is then shaped into a hypertriangular waveform.  Note that you could not tap pin 14 directly and feed that to the phase-shift stages.  You would need to run that squarewave control current through the 10k current limiting resistor normally found off pin 9.  So, to switch between square and hypertriangular, you cut the link between the 10k resistor and the LFO, solder the sovered end of the 10k resistor to the common of a SPDT toggle, one outside lug of the toggle goes to pin 9 and the other to pin 14.

Now that I'm thinking about it more, it would seem that if you had two unsynchronized squarewave generators, and combined them (with appropriate current limiting) you would have something not too far off from a stepped S&H-type sweep.  If you haven't embarked too far on the FSH stairstep generator, maybe this would be a good time to carry out the dual Ropez experiment.  Do it in several steps.  First, verify that the hypertriangular/square switch works.  Then try combining the square of one and the hypertriangular of another, making sure to change the 10k output current-limiting resistor of the square source to a 15-18k resistor.  I say this because the square has nothing in between max and min, where the hypertriangular is graded.  You need the square output to be constrained so that when the momentary level of the hypertriangular is added to the square, that addition can be heard as being somewhere below the absolute maximum point of the sweep range.

I suppose another way to do it would be to use a 25k linear pot, and leave the 10k output resistors as is.  One outside lug of the pot goes to one 10k resistor and the other goes to the other 10k resistor.  The wiper is what goes to the control-current pins on the phase-shift stages.  That would both limit the current so you don't blow the chips by driving them with the current from two LFOs, as well as provide a means for mixing the balance of the two control current sources.  That would be useful whether using two hypertriangular, two square, or one of each, to sweep the phase-shift stages.

Oh man, its getting complicated!   The punchline, however, is that if you build yourself two Ropez boards, making 8 stages and 2 LFOs available for use, there is a wealth of possible sonic textures available for the asking, with the mere addition of a few extra switches and controls, and very little modding of the board other than trace cuts.  And one of them would appear to be a stairstep function, which is pretty much where we started out.

[A.S.P.]

http://www.diystompboxes.com/smfforum/index.php?topic=27129.msg181958#msg181958

[Papa_lazerous]

Hey papa.....you try this out yet? Mark, right on with the rack box. i will be reboxing this for the third time if I add the step function.

Still waiting on a few parts from smallbear, so not got it up and running yet

Just checked status with USPS tracking and got this "Your item left the United States from NJI at 1:21 PM on January 10, 2008"

So not bad at all.  should see it soon, although Uk postal service is extremely poor these days may be another 3-4days