0-10VDC isolation

[tallen]

I have a circuit that reads a pot (0-5VDC) into the A to D of the PIC. This works fine. In some applications I take a 0-10VDC signal for an external controller through voltage divider into the same A/D on the PIC. I have a switch that changes control from the POT to the 0-10V controller. All of this works most of the time. For some reason part of the time noise from the 0-10V controller reaks havock on my PIC. I have tried different values of caps on this line with no luck. The problem happens mostly when the site is being run on generator power. This does not seem to be a problem on line voltage.

Is there a way to isolate the 0-10V signal coming into my PIC? I was thinking of an Opto-coupler, but I'm not sure that will work. Any ideas would be greatly appreciated.

Thanks,

Terry

[HenrikOlsson]

Hi,
A "normal" opto-coupler (PC817 etc) are highly non-linear and won't work very well in your application. There are analog opto-isolators available such as the HCLP-7520 - read the datasheet - and others.

Another option is an isolation amplifer such as the ISO124P but it's a bit more expensive.

What's happening when it acts up?

/Henrik.

[Dave]

tallen , How about a ISO-100 from burr-brown?

Dave Purola,
N8NTA

[HankMcSpank]

You say the noise creates havoc within your PIC - are you certain it's noise & not the source voltage going over 10V? (& therefore your potentially divided 0-5V breaching 5V?).That said if the noise spikes are very high & prevelant, then in essence the voltage you're monitoring will spike up above 5V.

The reason I ask, is because just a couple of nights ago, I was getting one particular (static) AtoD reading all over the shop....& this for a simple pot arrangement with 5V across it (I spent ages going over all my register my PIC register settings with a fine tooth comb - ages) except ahem, the aforementioned Pot didn't have 5V across it...it was 6V!

The resulting PIC AtoD readings totally were crazy ....jumping about all over the show for a static pot wiper reading.....successive samples along the lines of 1024, 868, 976, 658, 950 etc etc

Moral of the story (for me!)....when you get crazy AtoD reading, double check to make sure the voltage you're AtoD'ing is never greater than the PIC's supply voltage!

[tallen]

When this happens, the pic usually just hangs up. It will not do anything until it is powered down and back up again. It happens so fast that I have not been able to catch it on the scope. It has blown up the 15V power supply before, which is even more strange. The 15 V supply is isolated via an opto to drive an IGBT only. Like I said this all works fine on line power, but when they use a generator things go bad. If they are using the pot instead of the external 0-10V control everything seems to be fine on line power or the generator.

Thanks for the input.

Terry

[Sneaky-geek]

GROUND LOOP?
Sneaky-geek