Russell McMahon wrote:

>> Can anyone suggest a way of building a opto isolated input to feed a 
>> PIC.
>> I would like to be able to feed in say 5 to 250 volts ac or dc, 
>> frequency response is not important as the pins are only sampled 
>> every 10mS.
>>
>> I was considering feeding an opto coupler with a bridge rectifier and 
>> a small filter cap (i don't want 50Hz ac to appear as if its going 
>> off and on) but the wide voltage range has me wondering if there is a 
>> better way, maybe some sort of constant current.
>
>
> What are you trying to achieve.
> eg
>
> - Is the device simply meant to indicate absence/presence of signal

Yes, Just off or on

>
> - What bandwidth/response time do you need.


Bugger all, 200mS response would be more than enough, I don't want AC 
50Hz to appear to be 'toggling'

>
> - What power dissipation can you tolerate.


I can put this in a reasonable sized case but I may have 16 in one case 
and I don't want to add a fan, the device is a USB (based on the 
18f4550) input /output.
So I guess that 1 watt per channel would be in the ball park

>
>
> - Can you tolerate a power supply/battery/whatever on the input side 
> as well as the input signal.
>
I would prefer not to use a battery /fet, this needs to be cheap and robust.

> - What isolation voltage do you want?

Moderate, so what ever the spec of the opto will be fine.

>
> - Why, in each case :-)
>
>
> If you want say 5 mA optocoupler current you require 5 mA x 250V = 
> 1.25 watt dissipation.
> Adding an input  battery can help muchly in dissipation. depending on 
> usage duty cycle the battery could be very long lasting.
> A FET input stage and a battery would be about all that is needed. FET 
> can have a high value input resistor plus gate clamp zener and little 
> else is required. More demanding "real world' situations MAY require 
> some extra protection.

I would like to keep it simple as possible, I may be able to get around 
a constant current using a single transistor and a diode in the emitter 
but was looking for a sanity check, I thought there may be an opto that 
does this now days...




>
> A simple switching regulator could cut in at moderate voltages to 
> reduce the power dissipation at higher voltages. (My GSR buck 
> converter (in the archives) would work here - passes through input 
> voltage to opto at low voltages and turns into a switcher at higher 
> inputs.
>
Took me an hour to find in the archive, had a look at the buck 
converter, I expect I can do this a little more simple with the voltage 
range required if I can find a suitable transistor for a series reg, I 
may be able to drop the current to the diode a little ...

What do you think?

>
>
>        RM


-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist