[sdiy] Unrolling the OBX voice card

[Steve Lenham]

----- Original Message ----- 
From: "Antonio Tuzzi" <antoniotuzzi at gmail.com>


> Hi Steve!

Hi Antonio,

> On the CEM3310 data sheet, the attack, decay and release input are
> connected with a 10k resistor,  and there is a suggestion to use a 0v
> to -5volts swing to go from 2ms to 20secs.
>
> In the OBX , a lot of continuous value are set using pots traveling
> between 0V and 5.6V, values are buffered with tl084. For the Attack
> Decay and Release values, there is another stage of inverting and
> scaling using LM324 (scale -(4.87k/100K)) and on the card there are 10
> Ohm resistor connected to the A D R inputs of the CEM3310.
> But I really was expectiong a 487 ohm resistor to have exactly the
> same range as the one indicated in the CEM datasheet.
> Any idea ? (i checked OBXA schematics and there is the same) Am I
> overlooking something ?

As per Tony's previous post, the less-than-unity gain of the inverting 
amplifiers gives the correct voltage span when directly connecting to the 
CEM3310 (an external 10K/470R attenuator is assumed by the datasheet when 
quoting the 0 to -5V span).

The values Oberheim used give a slightly higher maximum voltage than the CEM 
datasheet but all this means is that the maximum envelope times will be a 
bit longer than the 20 seconds that the datasheet quotes.

I would guess that the 10R resistors are there to balance out any slight 
variations in bias voltage/current between the CEM3310s which might cause 
problems if all the control inputs were directly strapped together.

Cheers,

Steve L.