A retraction concerning the Ross/Ropez phaser

[Mark Hammer]

Recently, I suggested a modification to the basic Ross circuit that would emulate what the first-issue Small Stone used to do with the Color switch, albeit using a pair of CA3094s rather than an LM13600.  Turns out I was wrong.  Or rather, I had "misremembered" details of the SS circuit and misdirected people.  I will try and post the original schematic for the 6-chip SS later this evening, but for now, here is what you need to know, using the Tonepad Ropez layout, and the original SS schem, as reference.

1) The shape of the LFO waveform is dictated by a 270k resistor at pin 1 of the LFO circuit, which connects to V+.  In the original SS, it was a series pair of fixed resistors of 100k+180k going to the equivalent pin.
2) The sweep range that the circuit aims for, and the rate, is dictated by the 270k/100k pair at pin 3.
3) The Color switch on the original SS would "borrow" 100k off the 100k+180k pair and "loan" it to the 270k resistor going to pin 1, making the wabveshaping resistance 370k, and creating a 180k/100k divider at the pin 3 equivalent, where there used to be a 280k/100k divider.

So, to replicate the wide/narrow aspect of the original Color switch on the SS, here's what you do:

a) replace the 270k resistor going to pin 1 with a 100k+180k pair, in series, with the 180k connected to pin 1 and the V+ end of the 100k left free/unconnected,
b) lift the V+ side of the 270k resistor that goes to pin 3,
c) connect the free end of the 270k to the free end of the 100k,
d) get thineself a SPDT toggle/slide, and connect the common to V+,
e) connect one outside lug of the switch to the new 270k/100k junction, and the other outside lug to the 100k/180k junction.

What you now have is essentially a 550k "chain" (270k+100k+180k) that extends from pin 3 to pin 1.  The toggle will connect that chain to V+ at two different points, providing a change in the nature of the sweep.  The part that remains true over both the correct and incorrect mods is that the sweep will be narrower in one setting that the other, and will be a little faster and higher up in the range for tha narrow setting, while somewhat slower and starting down lower in the range for the other.

What I had posted earlier had things ass backward.  Apologies for the confusion.

[frequencycentral]

Well, no harm done here - I must have missed your earlier posting. I just today printed off all the Tonepad Ross phaser schematics and the Tonepad SS schematic so I can compare/contrast for my next project, which will include elements of both, so this thread is very useful. Thanks Mark

[gigimarga]

Little offtopic, but which is the maximum useful stages for a Ross Phaser?
I will build one use 4 and 8 stages, so it's useful to have more than 8?

[Mark Hammer]

More than 8 stages probably makes little difference if you are using it for guitar only.  The extra stages start to be useful when the signal source has broader bandwidth than a normal electric guitar.  Perhaps acoustic guitar or a band mix.

Keep in mind that the phase-shift stages rely on the current from the LFO to sweep them.  When the same current is divided over more stages you may not get as much sweep unless you drop the value of the 10k current-limiting resistor a bit.  I'd be careful with that, however.  Too much current will fry the LM13600.

Also keep in mind that the phase shift stages can unintentionally ADD gain due to tolerances of all those resistors.  As a result, you may find that feedback settings which are perfectly acceptable for 4 stages will produce howling sounds with 8+ stages.  You may want to include a feedback trim control/trimpot as some phasers do, to set the maximum feedback.  This would probably be something like a 10k trimpot in series with the 27k resistor that goes to "Regen".

[gigimarga]

Thx a lot Mark (very nice, as usually)!

It seems to me that i was right when i thought that 8 stages are enough for my guitar

[Mark Hammer]

Here is the early (first?) issue "6 chip" Small Stone.

[notchboy]

I want to build a stereo 8-stage Ropez using the Tonepad PCBs.  After reading the various Ropez threads, I have a couple of questions...

Doesn't the 10k resistor on the output of the LFO need to be reduced if the "extra stages" board(s) is/are used?  For example, when you double the number of stages, the resistor needs to be halved in order to deliver the same current to Iabc on each OTA, right?  Mark H also asked about this but I never saw a definite answer.  There was also a discussion about replacing the 10k resistor with a pot and fixed resistance to get variable sweep depth.  Does anyone have favorite values that give a nice wide sweep without frying the OTAs?

Is it confirmed that the vibrato mod works?

[Mark Hammer]

Perhaps the thing to do is to simply have one 10k for one group of 4 stages, and another for the other group.

And yes, lifting the dry connection on ANY phaser, whether 4 or 8 stages, whether OTA-based, LDR-based, or FET-based, will get you vibrato....assuming that is your question.

[notchboy]

Yes, I know that you get vibrato by using only the wet (allpass) signal.  The modified schematic looks right, but someone mentioned having a problem getting vibrato to work, so I thought I'd ask before cutting traces...