Where i can find the "depth mod" for Ross Phaser (Ropez)?

[gigimarga]

Hi,

I've read that it's possible to add a depth pot to a Small Stone, but i can't find the same mod for a Ross Phaser.
It's possible to do that?

Thx a lot all!

[Mark Hammer]

The Color switch on the Small Stone adjusts resonance and sweep width at the same time.  There is some information about modding a Ropez to do this here:  http://www.diystompboxes.com/smfforum/index.php?topic=73208.0

You can also reduce how much sweep there is by increasing the 10k resistor on the output of the LFO.  I wouldn't go much higher than 22k.  A super wide-range Ropez would have a 9k1 fixed resistor in series with a 10k variable resistor in place of the 10k shown in the schematic.  Don't go below 9k1 or you risk burning out the LM13600.

[gigimarga]

Thx Mark...a 8.2K  instead of 9.1k could be enough safe?

[Mark Hammer]

It might be.  You may want to confirm with others though.

[foxfire]

i know i shouldn't say anything until i'm certain but, i'm pretty sure i have a 6.8k in my ross.

[Mark Hammer]

 I have an 8k2 in one of mine and the sweep goes up much higher than a stock unit.  We're talking "you mean there's MORE up there?" high.  Ultra-wide like an A/DA Flanger.  High enough that I have fear for the chip.  I don't mean to cast aspersions on you but if the change from 10k to 8k2 does that, then a 6k8 resistor strikes me as implausible.  A 9k1 unit is probably safe and wide enough for most folks.  8k2 may be okay but I'd hate to suggest something that results in someone's last LM13600 frying.

[oskar]

http://www.synthdiy.com/files/2002/LM13600.pdf

Actually you're rather on the safe side here. The rated Maximum Amplifier bias current/I_ABC is 2mA ( which is probably noisy ).
Say 1 mA per OP x 4 = 4mA maximum through the resistor in question. 10V/4mA=2k5. It would survive for sure!
The datasheet graphics stops at 1mA I_ABC so I think that is a bit passed the recommended span for linearity and noise and whatelse I don't know. You don't stop at a little noise now?   
I wouldn't go under 4k7.
Ps. Kirschoff say's I'm right under the condition that the current divides evenly in between the 4 OTA's ( lm13600 ): I don't know
that for a fact. I just assume that the filter behaviour is dependent on this... but someone else should fill in the blanks here and
yes... I do round off the numbers in the equation like someone sailing the oceans flagging jolly roger for a living.


I'm sort of staring at my russian smallstone with more or less unpure thoughts. It's allready opened... 

[MarcoMike]

I have a 5k resistor there and it works just fine... and, Mark, once you started suggesting like 2k... then increased every time till this 9.1.... how many chips did you burn??

and as usual I have to say... I regret selling my old Morley PFA... with its Centre and Travel controls... that was the real deal!!!
would it be so difficoult to make a Centre (of the sweep) control?

[Mark Hammer]

You can adjust where the sweep starts from by adjusting the voltage divider comprised of the 270k/100k pair (a [180k+100k]/100k trio if you follow my suggested sweep-width mod).  I experimented on one of mine by inserting a 50k pot with the wiper going to pin 3, a 75k fixed resistor on one side, and a 240k resistor on the other.  That gave approximately the same overall resistance, while maintaining the ability to shiftmore and less of the total resistance to the ground leg.  While it DID show some shift in the range of sweep, I didn't find it nearly as huge as simply twiddling the bias trimpot on a JFET-based phaser.

Perhaps the real thing to aim for is a final common pathway that pools/blends the LFO output and a DC offset source.  I wish I knew how to do that.

Oskar, if you have a SS you're willing to "defile", take a look at Charlie Barth's Frankenstone over at his website: http://moosapotamus.net/THINGS/frankenstone.htm  Although I don't think his clips show it off sufficiently, I strongly recommend the phase-filter mod.  Two stages of allpass followed by two stage of lowpass, with no clean signal mixed in, is a lovely thing to behold; the very sort of thing that producer Dan Lanois needs to have in his arsenal.

[oskar]

Frankenlink OK. I'll look into your mod too.
What did you mean with DC-Offset?
Something like this?
I just shot from the hip with values and I don't have any guitar, amp so I can try it and report back.
Diode is to cure the dead spot on the sweep from 13600 internal reference. Actually I think it's more like two 0.6V drops in there?!
Gotta check it up.
Edit! Just added a second diode on DC pot ground path. This mod makes for a total resistance of ~7k at extreme potentiometer settings

[Mark Hammer]

They don't let us see photobucket and similar files here at work, so I'll have to wait until I log in at home to respond.  Right now, I see "Something like this?" and no accompanying image.

[MarcoMike]

well, just to let Mark know... the control voltage is tapped to a voltage divider from + to ground, with a couple of diodes on the ground side... more or less as it was done in the Morley PFA circuit... as I was studying that schematic I saw they used a pot to ground to set the LFO amplitude... will it work in our case as well... I'm tempted to open my RossPhaser and try this right now...

[oskar]

Haven't seen the ross Morley. However in this configuration the DC pot and LFO depth will interact. You can cure it with a couple of diodes but I don't think it will matter that much.

[MarcoMike]

yeah... just tried it... doesn't work 
but the morley works just fine that way... much higher resistors involved, but I guess it depends on the likely high impedance of the transistor which drives the light bulb...

[oskar]

  -But, it's just current... How can it not work? Is it just not in a usable range?
You don't need a lot of current to open the OTA. I'll try and do some creative abuse on my own here.
I'm tired so perhaps I've overseen something very basic here. The values were meant as a starting point...

but the morley works just fine that way... much higher resistors involved

Well it's a voltage divider so I reckoned the value of the pot really shouldn't matter that much or account for some "charming" unlinearity.

[MarcoMike]

the problem comes out on the + side of the pot.... there is one point at which the lfo just stops working and you need to turn back to more negative values and wait few seconds for the lfo to start again...
with a lower resistor from the wiper the pot act as amplitude control (shunting the lfo to ground) again, no lfo-centre-shift... oh... I didn't use diodes but resistors... were they mandatory?

[oskar]

Diodes are used because the controll pin of the OTA is referenced to the negative rail plus two forward diode jumps. So at under ~1.2V at the wiper the pot will be of no use. Not mandatory that is.
It's actually more delicate than I thought but it's definitely doable with a few components if you can stand the tradeoffs needed.

the problem comes out on the + side of the pot.... there is one point at which the lfo just stops working and you need to turn back to more negative values and wait few seconds for the lfo to start again...

But that's just a matter of tweek, tweek, tweek... right? Lower the range of the pot.

with a lower resistor from the wiper the pot act as amplitude control (shunting the lfo to ground)

That was the only thing I had forseen actually. Diodes is the cure...

again, no lfo-centre-shift...

There should be no resistor in front of the LFO so the LFO should be given priority. It should "swamp" the cv line.
both CV's should have diodes so they don't interfer with eachother.

[MarcoMike]

the point is: there is NO centre-shift at all!! what I listed are the only effects I could get tweaking the pot.
and I tried quite a few resistor combinations... always whit lower resistance on the LFO side in order to preserve the priority  I went from 0 to 20k here and tried 0 to 15 on the voltage divider, with the usual 5k from the meeting point to the 13600s

[oskar]

Like this!      ( Hm. I must remember to put a big "Not verified on scetchy stuff like this" )
R4? No idea if it's needed whith D4 in the circuit. Omit...
R2. Perhaps increase. Now the DC CV will lift up the minimum current for sure and the LFO will not be disturbed.
My guess is that the value of R2 is more important than any other values and R4 shouldn't be needed at all.

You just posted. Thank's for trying this out. I was really just thinking out loud with the first drawing and now apparently I'll be sleepless again. It's amazing how few components it takes to destroy a good nights sleep...     

[oskar]

OK. Last try. I extended the rev.3 mod to a full LTspicemodel and hit go!
both pots are 10k, D5/D6 is the OTA controlpin equivalent through which current is plotted below.
V1 represents the LFO but it only works to show how the LFO influences the OTA controlpin net. Not vice versa.

ps. I've incidentally written smallstone mod all over the place. The mod is for the Tonepad Ropez. Sorry!   
Caveats are as follows.
1. I know little about filters
2. I know nothing about all-pass filters.
3. I only homed in on the problem with controlling the current through the OTA control pins.

ps. For those unfamiliar with LTspice each of the two pot is in turn given three different wiper position which makes for a total of nine differen plots presented together. During the conversion to black and white some of the graphs got lost...  The general idea is still clear though.

SIMULATIION 1
ROSSMODrev.3
No surprise here. As the simulation shows, the diodes needed to prevent the DC shift pot to disturb the LFO also creates a floor where there will be no motion.


ROSSMOD2
In this simulation I have omitted the diodes so now V1 represents a lowZ source. An OP-amp.
And now were starting to get useful results.


Finally we take away as much resistance as possible leaving the pot to regulate current from buffered oscillator and only 500R to wiper on DC shift pot. And this looks to me as a usable result. Schematic will follow. The improvement over the previous simulation is that now the DC shift will have a more pronounced effect in the circuit when the depth pot is fully engaged.