How to read a voltage and keep it Isolated from PIC

jfk · Guest

Hi

I have to read the voltage of a high voltage battery (>300VDC) through an A/D input of a PIC and transmit the voltage over a CAN network, how is it best to do this and keep the PIC isolated from the high voltage?

The PIC I'm using is a 18F4580.

Thanks

JFK

SherpaDoug · Joined: 07 Sep 2003 Posts: 1640 Location: Cape Cod Mass USA

You have two choices of where to isolate the signal, analog or digital. For anaolg isolation look at "isolation amplifier" at www.analog.com. Other companies may have cheaper less accurate ones if you want. The other route is to run the A/D on the battery side and isolate the digital signals from it. You can use an all-in-one chip like the AD7400, or roll your own from your favorite A/D chip and and opto or magnetic isolator.
If you are used to working with PICs you might consider using a small PIC just for its A/D, and use a single channel optoisolator to link the small PIC to the main one.
_________________
The search for better is endless. Instead simply find very good and get the job done.

jds-pic · Joined: 17 Sep 2003 Posts: 205

see
http://www.ccsinfo.com/forum/viewtopic.php?t=19529

and refer to

jds-pic

SherpaDoug · Joined: 07 Sep 2003 Posts: 1640 Location: Cape Cod Mass USA

Yes jds-pic, that is an oldie but goodie that I had forgotten about! Fifty years ago they would have done that with a "resonant relay" that just vibrated very quickly and at low power. I don't think they make those any more, but any fast low power relay should work as long as the contacts are break-before-make which most are anyway.
_________________
The search for better is endless. Instead simply find very good and get the job done.

kender · Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley

jfk, recently there was a discussion about a problem similar to yours here
http://www.ccsinfo.com/forum/viewtopic.php?p=73441

Ttelmah · Guest

For DC, there are a whole 'raft' of solutions, depending on the accuracies required etc. etc..If there is significant power available on the 'high' side, then you can potentially use this to supply an op-amp, feeding an accurate opto-coupler (normal linear opto's, are very innacurate on the light source side, so may be adequate for a crude measurement, if individually set up, but for accuracy, the 'trick' to have an opto that has two light sensors, one on each side of the chip, and adjust the transmitted power using the reading from the sensor on the 'high' side, to give an accurate reading on the other side). Do a web search on the IL300 (probably the commonest such opto). This type of circuit can also be operated using a small dc-dc isolated inverter to run the op-amp on the high side.
Another solution, is the Texas (Burr Brown) ISO124, which behaves like an op-amp, 'split' by an isolated barrier. This actually sends the DC value from one side to the other as a frequency. You can do the same thing using a PIC as well. I have in the past used a cheap PIC like the 12F675, used it's ADC to read a value 'high side', and then sent this as simple serial, across a digital opto-coupler.

Best Wishes