which Ross Phaser sweep mod?

[Boogdish]

I'm repairing/modding a vintage Ross Phaser for a guy that I work with.  I've been doing searches on mods people up here have done (by "people" I mean Mark Hammer) and found two separate methods for controlling sweep range:

http://www.diystompboxes.com/smfforum/index.php?topic=48283.0

this one is a control of how much V+ goes into the beginning of the LFO circuit

and

http://www.diystompboxes.com/smfforum/index.php?topic=64973.0  (scroll to the very bottom, this was mentioned in other threads as well)

this one controlling how much current is going out of the LFO circuit.

So to anyone who has tried both of these, which did you prefer and why?  I'd like to try both of these on the unit I'm working on but I'm trying to do as little soldering on the old board, since I found out today while replacing the old LM16000s that those traces are just dying to come off.  Right now I'm sort of leaning towards modding how much current leaves the LFO because I could also easily add a bleeder cap there to get some of the spike out of the fast oscillations.

Thanks, and if anyone knows where to get replacement pots for these old things that have 2-inch D-shafts and inch and half bushings with PCB mounts I'm all ears.

[Mark Hammer]

They both have their charms and quirks.

The linked-to adaptation of the colour switch will have the same effect in both the Small Stone and Ross units; namely, reducing width while upping the speed a bit.  The idea behind the colour switch is that when you want slower sweeps, you want them wider and more resonant, and when you want faster sweeps you want them narrower and less resonant.  The Ross provides variable resonance so the adapted colour switch I described simply changes the width and rate but not the resonance.

For whatever reasons - and probably due to resistor tolerances - I find that the sweep is not always to my liking in the various Ross/Ropez units I've made. Some are bang on , and some just feel "wrong" and kind of "lumpy".  Recently, I've started tinkering with the 270k resistor that goes to pin 3 in the LFO, and replaced it with a 220k trimpot and 100k.  Doing so results in the voltage going to pin 3 being adjustable, without impacting on the value of the 270k resistor at pin 1.  Naturally, as one decreases the value of that 100k+220k combination, the voltage appearing at pin 3 goes up (the stock 270k/100k combo acts as a voltage divider), the speed tends to go up as well, and the range shifts slightly. 

Typically, when one aims for a narrower sweep width, you want the notes to "gurgle" somewhere in the middle of the sweep range rather than just at the top or bottom.  When one plays with the value of the 10k resistor from pin 9 of the LFO circuit going to the 4 allpass sections, that does not change the lowest point of the sweep.  It simply alters what the highest point will be.  So, while it will reduce sweep range in a very easy way, it will not necessarily situate that reduced sweep where you want it to be.  So, I'm think that maybe the ideal is to have some kind of 3-position toggle that does a couple of things simultaneously.  Essentially it would change the value of that normally 10k resistance on the output of the LFO, at the same time as it changes the value of the 270k resistance going to pin 3 of the LFO.

The trend you want to achieve is to increase the resistance of the 10k LFO output resistor at the same time as you decrease the value of the 270k resistance to pin 3.  That would result in moving the sweep range and rate up a bit, and reducing how far the LFO sweeps from its initial starting point.

The key question to be answered is what the ideal resistance changes ought to be.  For the LFO output resistor, you don't want the resistance to go much above 18k, and you don't want it to sink much below 9.1k  For the replacement of the 270k resistance, you probably don't want anything that goes outside the range of 150k to 300k.  However, I encourage folks to experiment with some trimpots and identify what they feel are suitable resistance changes.

Another approach to try is to replace the SS-style colour switch with a dual-ganged pot that will simultaneously vary the "V+ tie point" and the LFO output resistance.  What is this "tie point"?  You will note that there are two 270k resistors going from pin 3 and pin 1 to V+.  The colour switch on the original SS breaks up the 270k from pin 3 into a 180k and 100k in series.  That resistor pair goes to the equivalent of the 270k resistor on pin 1.  The switch then either "taps" the point where they connect to V+ from the 180/100k junction or the 100k/270k junction.  In other words, that 100k resistor is either part of the voltage divider feeding pin 3, or else it is "re-allocated" to the 270k resistance from pin 1, leaving the voltage divider on pin 3 consisting of a 180k/100k pair instead of a 270k/120k pair.

Okay, so now imagine that instead of a 180k/100k/270k resistor chain, as in the SS, we have a pot section where the wiper goes to V+, and the amount of resistance that can be allocated to either the pin 1 resistance or the voltage divider on pin 3 can be varied continuously.  Imagine as well, that the other half of this dual-ganged 100k pot sits in parallel with the LFO output resistor.  We'll change that resistor from 10k to 24k.  And we will stick a 15k fixed resistor in series with that 100k pot section such that the parallel resistance will range from 24k in parallel with 115k (=19.9k) down to 24k in parallel with 15k (=9.2k).

Making sure that the direction of the two pots is wired up correctly, what that dual resistance change will do is simultaneously shift the "starting point" of the sweep up a bit, and increase the speed a bit, as it reduces how much current is fed to the OTA-based allpass sections.  Note that the shape of the LFO waveform also varies with the pin 1 resistance.  In theory, that dual (actually triple) resistance change ought to get us a useful degree of control over sweep and "feel".