Ross phaser 4/8 build (no pics) observations, suggestions? (sound clip inside)

[jimmybjj]

sorry about the poor quality sound clip and the pops (from changing the power supply)

http://sites.google.com/site/foundher1234/phaser.mp3?attredirects=0&d=1

I have noticed significant between 18v and 9v power operation, has there been any discussion about how power effects how the phaser sounds. I am still a novice in my understanding of electronics and have nothing to offer technically but would love to ready more about it. Has anyone proposed mods to the effect of powering it with 18v and having it switchable to 9v, don't know how practical this would be.

In 9v operation (the first chord in the clip) I don't like how severe the end of the sweep is, is that wolfy sound normal? Is there a mod to smooth it?

[jmasciswannabe]

Dude, Mark Hammer has covered this pedal inside and out. Search for ross phaser as well as ropez. I recall a mod using an electrolytic to smooth out the sweep. You might also search my posts as I haven't posted nearly as much as him but know I implemented this mod at one time . Can't help you out on the voltage questions but surely Mark or Stephen will chime in. As far as I am concerned, they are the experts on this stuff. And while I am at it....thanks for the knowledge gentlemen!!!

[jimmybjj]

1) Stick a 22-47uf cap between the point where the 10k current-limiting resistor feeds the four LM13600 sections and ground.  It will serve to smooth out the sweep at fastest speeds but have little or no audible effect at slower speeds.  The net result is that the fast-Leslie settings will sound more pleasingly bubbly and not as jarring.  This was part of the very first Small Stone.  I'm not sure where it disappeared to over the years.

2) The Ross can be made to sweep a little wider (higher) by reducing the value of that 10k resistor a bit, and made to sweep more narrowly (i.e., turn back around at a lower point in the sweep) by increasing it.  DO NOT go below 8k2 or you risk frying the 13600 with too much current.  At 8k2 resistance, the sweep is tantamount to that of an A/DA flanger.  In other words, just when you think it is going to stop and make the turnaround, it keeps on going in very dramatic fashion.  Not everyone's taste, but it's impressive in comparison to what you're accustomed to from most phasers.  You can opt for something like an 8k2 - 9k1 fixed resistor in series with a 10k-25k variable resistor, or simple a wide/medium/narrow 3-position toggle that selects between, say, 8k2-15k-27k.  Of course, the stock 10k is fine too, but since you seem to want variations....  Keep in mind that it is a common principle in modulation effects that as speed goes up sweep width needs to go down.  The Ross, and a great many others, provide no means of adjusting sweep width so they provide a compromise width that will work reasonably well for all speeds attainable.  Doesn't mean it's the "best" width for you at that speed.

Is this resister referred to in the above post?

[frequencycentral]

The 10k you want to widen/narrow the sweep is between IC3 pin 9 and and IC2 and IC3 pins 1 and 16.

[ddpawel]

Is it really safe to run this pedal 18V?

[amptramp]

It sounds like your LFO frequency is proportional to your supply voltage.  There are circuit topologies (like the 555 oscillator) where this is nulled out to a first approximation.  I would be concerned about that - you never know what quality of power you are going to have.

[Mark Hammer]

I don't know what the voltage HAS to be, but the 270k/100k pair that feeds pin 3 on the LFO, divides the V+ down to an appropriate voltage.  Depending what that voltage is, you get different sweep speeds and starting point.  The "Color" switch on the first issue Small Stone would effectively "steal" 100k from the 270 resistance going to pin 1, and add it to the 270k feeding the equivalent of pin 3, such that the resistive divider feeding pin 3 was either 270k/100k (for a hypothetical bias voltage of 2.4V) or 370k/100k (a hypothetical bias voltage of 1.9V).

Using the same values for the resistive divider with an 18v supply gets you something different.

Realistically, I can't see any justification for going with 18v, unless one wishes to use it as the "default" supply voltage for a bunch of pedals.  Precious few phasers ever incorporate any gain, so running out of clean headroom is generally not an issue, and not a reason to migrate to higher supply voltages.

If you DO use 18v, I would be reticent to aim for dropping the value of that 10k current limiting resistor, unless someone here with strong EE chops gives the go ahead.

I've built a half dozen or so of these, and I have to say that nailing a pleasing sweep has been hit and miss.  Sometimes you get somethng nice, and other times it's an irritating and distracting sweep.  I wish I knew how to make it consistent.

[jimmybjj]

I've built a half dozen or so of these, and I have to say that nailing a pleasing sweep has been hit and miss.  Sometimes you get somethng nice, and other times it's an irritating and distracting sweep.  I wish I knew how to make it consistent.

I have built 2 so far and have not really been pleased with either one. they both have a sweep that seems to speed up and slow down rather than a constant speed, I know that's not really happening but that is how my ear hears it.

on a side note do you happen to know how to alter the sweep on mxr phase 90 I have built a couple of those as well with the same results, my factory unit is so much more pleasing to listen to, nice smooth sweep rather than a almost go and stop go type sweep.

[Mark Hammer]

One of the principle reasons for matching JFETs for phasing is precisely to achieve a nicer more musical sweep.  Ideally, all JFETs are changing resistance in unison.  Maybe not the exact same resistance in each JFET, but certainly getting higher at the same time as the others, and getting lower at the same time.  When unmatched, it can be the case that one or more reach their I-ain't-changing-one-little-bit-more point before the others, such that you may have 3 of 4 continuing to decrease (or increase) drain-source resistance as one remains at a fixed resistance.  This can make the "turnaround" at the extremes of sweep awkward and unmusical.

In theory, this should not be a problem with OTA-based phasers like the Ross or Small Stone.  Rather, I think the problem has to do with how well they pull off a hypertriangular sweep.

As you can see here, the sweep is sinusoidal at one extreme, and "pointy" (triangular....sort of) at the other.  The exaggeration at the "pointy end" comes partly from the value of the 270k resistor tied to pin 16 (the Tonepad layout flips the two sections and uses the functionally equivalent pin 1 instead).  On the first issue Small Stone, the identical LFO circuit is used, but with a pair of CA3094s instead of a 13600.  The 270k resistor on pin 16/1 is replaced with a 100k and 180k pair in series to produce 280k.  Both the 270k resistor shown here tied to pin 14, and the other 270k resistor are tied to V+.  What the Color switch in the old SS would do is move the V+ connection back and forth between either the junction of the 100k and 180k (imagine the leftmost 270k here is now 370k and the right one is dropped to 180k), or the junction of the 270k and 100k.  That would change the sweep width, the speed range, and the shape of the sweep, in one move.

The hypertriangular sweep is very useful for long slow-n-wide sweeps, but can be a little irritating for faster ones where you don't want quite so much width.  Tinker with the value of the 270k resistor going to pin 1 on the Ropez, and see if that produces a more satisfying sweep.  based on the Small Stone circuit, I will suggest using a....wait for it...180k fixed resistor and 100k trimpot.

This may well be one of those cases where the 13600s used back in the 80's, when this pedal was originally produced, provided the desired sweep with a 270 value, but chips available now, or from certain manufacturers, want a different value to behave in exactly the same way.  I also have to confess that I have not tinkered with the value of the 10k resistor at all....ever.

I'll keep tinkering too, and if I learn anything, I'll be sure to post it.