At 19:43 02/07/97 GMT+200, Hennie Swanepoel wrote:
>Tjaart wrote:
>
>> You probably mean 4mA to 20mA, don't you? A 250 ohm resistor gives you
>> 1V to 5V.
>> Even though the 250 ohm is not standard E12, you can still get it in
>> very low
>> tolerances (0.1%).
>
>No, strangely enough 0 to 20mA is an old Siemens standard used for
>process control (on at least some of the South-African power
>stations).  I would pefer not to use only a simple resistor
>(preferably 50 ohm) since the accuracy of (subsequent) A/D conversion
>would be deteriorate due the quantization effect etc. 

I'm not sure whether I understand you here. How could a resistor based I-V
converter influence the subsequent ADC quantization? Just make sure that
the voltage at max current gives you something close to but lower than the
max voltage of your converter, and you are at the optimum you can get re
quantization.

Additionally, _every_ other solution (including opamps and, of course,
ready-made modules, which all are opamp based) depends on the same
principle (converting a current into a voltage by sending it trough a
resistor). No way around it, and rarely a need to use more than one
resistor in this case (which is also the optimum in both precision and
reliability).

If you have any special condition which makes it impossible to use it this
way, I'd really be interested to hear what it is.


>Also I need to
>do the opposite: convert a voltage to a current in the 0 to 20 mA
>range.  The usual solution to both these problems involve using opamp
>circuits.  I was hoping that someone had come around to designing
>IC's dedicated to current/voltage and voltage/current conversion.

An opamp is just this... you need only a few external resistors to do
either conversion (I->U 2-3, U->I 4 Rs), and all integrated modules I've
seen so far which do this are bigger than an opamp with some resistors.
(You can get down to a SOT-223 and 2-4 SMD resistors...) And they, of
course, are all based on opamps with a few resistors integrated in the same
package; you probably still need the one or other resistor to set the input
range. But really, I don't quite see how you can get it smaller and cheaper
and easier than with opamps. 


-------------------------------------
Gerhard Fiedler <gerhard@pobox.com>
S‹o Paulo - Brazil