Russell McMahon ha scritto:
> MAY be a partial resend.
> VCO cct attached is from LM324 datasheet.
> Dual opamp, 1C, 1Q, 7R.
> 
> Probably more complex than you wanted.

A bit :)
since still it has to be added a voltage regulator (since max is 72V), 
and the opto. But what you state down is interesting...
BTW I used to use 324 with much success. Just I was not sure about 
accuracy in such a case, especially considering all of the passive involved.

>> In this schematic, I also need a LDO or alike to power the leftmost
>> OpAmp - same as with the VFC...
>>
>> That's why I wanted to stay away from this all , with something simpler.
> 
> Can you specify your requirement more tightly.
> eg explain the degree of complexity that would be acceptable at max.
> And how much isolation do you need or want.

Ok
Basically, there's not much room available on the board, but we'll try 
to arrange something.
Isolation is from a telephone line, so not so critical in the end. 
Accuracy +-2-3% can do, say even 5%.
What I feel like "bad" is the use of so many parts to just measure a 
voltage with such a low precision. So the single Opto approach looked 
very attractive.

BTW (again) has anybody taken a look at the datasheet that I posted some 
days ago, concerning optoisolator ageing? thanks! :)



> eg in many case an opamp differential amp on the OUTPUT side with
> largish input resistors will allow you do measure that voltage range
> IF the linbe floats NEAR ground. If CM voltage is large then maybe
> not. But for offsets of say  +/- 100V wrt ground you can probably do
> it all with LM358/2 and resistors.
> 
> If a dual  opamp is OK on input side (cheap, smallish) then a V2F and
> more is easy.
> 
> Powering say 1 mA from 70V = 70 mW = OK to use dropper resistor and
> shunt regulator.



This is something I actually did not consider.
I once tried (but I was a young boy :) ) to measure and detect on-hook 
and off-hook out of a telephone line without isolation, and it was not 
working...
In this case, we have a power-line running around telephone 
street-boxes, which powers monitoring stations. Isolation *might* not be 
a great issue, but in case of lightning or such, it would be just bad - 
as in any other case. Otherwise, the 100V (or the LM3900 unlimited 
voltage that Michael suggests) would work.

Still, the OpAmp has to be powered, referenced, converted into pulses 
etc. Not much of a saving, just the opto.

Oh, I probably forgot to say that it's not a money-issue: the whole 
board is going to cost from 150 to 250 EUR (due to Solid-State Relays, 
Triacs, Ethernet, DC/DC and AC/DC converters, pressure and humidity 
sensors...)

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