Interpolating Scanner (Tillman)

[Pehr]

I think I'm going to build the Tillman IS, but I have some questions:

Is a LM13700 a good replacement for the CA3280?

If so, is the 'Id' input of the CA3280 equivalent to the 'diode bias' input of the LM13700?

In J. Donald Tillmans I.S. from January 1999 the 'input sections' are 'mixed' into a voltage follower. Is it really possible to do that? I mean each 'pan' overlap each other a little.

I would prob have used a regular non-inverting mixer with a resistor in between each 'section'...

CCA:


Scan plot:


EDIT: Happy new DIY year everyone!

[yusynth]

Hi Pier

The inputs of the CA3280 are a bit different from these of the LM13700. I may be wrong cause I don't have the datasheets at hand but the CA3280 has current inputs while the LM13700 has voltage inputs. This means that for the LM13700 you must add some resistors (say 100/220 ohm or so) between the inputs and the ground. In your schemo, add a 220 ohm resistor between the non-inverting input and ground. I don't think it is necessary to add one for the inverting input because the offset trimmer is already providing a polarization.

The Id pin is for the linearizing diodes.

Yves

[Pehr]

Thanks for pointing that out! I had no idea of the input differences

However, one may use any VCA, right?

[yusynth]

[Pehr]

Great! Then I'll try the simple SL-VCA

[frijitz]

[Pehr]

[fonik]

say i'd liked to build a four channel version, could i just leave off four of the control current sources (edit: diodes/transistors) and double the value of the resistors R9-R20 or R21-28 (or what resistors would remain then)?

[fonik]

maybe my question was not understandable, because i simply don't understand how that circuit works!?
2nd try:
i'd thought the incoming voltage was somehow divided by the resistors R9-R20?
so for a four channel version, would it be appropriate to leave off just the diodes and transistors of the needless channels, while keeping all resistors R9-R20?

[yusynth]

I'd try this (and tweaking some resistor values, probably). Not garanteed though

---

schem.gif
Description:	Modified schematic (I removed R18 as well ;-) )
Filesize:	91.5 KB
Viewed:	86 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[fonik]

nice work, yves looks great.

thank you very much. so it's the ratio of these resistors (R9-R14 or R15-R20) to the sum of their values that creates the division!? so if i changed just one of these resistors i would widen/narrow the according window?

good breadboard application for me, i guess.

[fonik]

[yusynth]

[fonik]

thanks for clarification. it's good to get the feeling to be right, even sometimes

i was thinking building kind of morph controller

[Pehr]

The Morph Controller looks cool

I got another Idea: One could have more than one Scanner Controller, controlling a set of 4-8 VCAs, right? That means one will have more than one stage activated at once. That's possible, right?

[fonik]

...or even voltage controlled switching of scanner channels...

[Pehr]

[frijitz]

[Pehr]

Would the SL-VCA really work for this application? it is linear...

SL-VCA

EDIT: Btw, R7 and R8 is 300R, right? I got to breadboard and test this circuit.

[fonik]

[Pehr]

[fonik]

actually i had switches in mind, but your solution seems to be way more nifty (and more modular, since you could just use a DC coupled VCA that may is part of your system already).

[Pehr]

[fonik]

again actually not. the switches would deactivate certain channels of the driver circuitry (blocks of diodes and transistors). so one could choose from different counts of channels and the channels themselves would be selectable, thus even the VCA channels would be selectable (without repatching, just voltage controlled).

does that make any sense? i am just guessing here.

[Pehr]

Aha 0k Seems rather complicated though.