4-20ma sensor - Overvoltage Protection

[jsmanson]

Just a quick question, I am building a picaxe project that will be ADC reading a series of 4-20ma sensors running off 24 volts.

I want to put in an overvoltage protection circuit so I don't 'blow' any more of my picaxes (lol).

I can't drop voltages with a resistance divider like most input schemes to drop voltage, as that will affect the operation of the 4-20ma (current output) device, so I thought I would insert a 5 volt zener and a series resister (R1) in there to limit possible overvoltages (I have already mistakenly bypassed the sensor once already and that was it for the picaxe, it didn't like 24 volts for some reason....).

Does this make sense, and would say a 5k or a 10k resister between the 4-20ma circuit and the pic input be about right?

Thanks!

John

 

[AllyCat]

Hi,

Yes, 10k is about right, that is the maximum source resistance recommended in the PIC data sheet for accurate ADC conversion.

However, a zener diode is probably not a good idea: A "nominal" 5v1 will probably be rated at 5 mA and may still conduct some current at (say) 4 volts, which will compromise the ADC measurement accuracy. Instead, connect a forward diode from the ADC input pin to the supply rail (Vdd). That will "catch" the voltage at about 5.6 V or, better use a Schottky diode which will drop less (~300 mV). Actually, there is already an internal "static (electricity)" protection diode in that position, but its current rating is not well defined.

Beware that the catching diode (internal or external) can feed current into the supply rail and pull that voltage up. Normally, just ensuring that there is some other load on the rail will be good enough, but if a high overload current is "expected" then put (say) a 5v6 zener across the Vdd rail.

Cheers, Alan.

 

[Reloadron]

I may have this all wrong but I do not see it working as drawn. The Maximum source resistance is 10K Ohm and there is no reason to go there. Yout sensor is a 4 to 20 mA device, if you run that through a 220 Ohm resistor as drawn the voltage drop will be 0.88V to 4.4V. Since the minimum is 4 mA you can never have 0V. Normally with a 4 - 20 mA loop a 250 Ohm resistor is used resulting in 4 - 20mA = 1 to 5 Volts. Then any code is written around that. The downside is obviously if the resistor opens you will exceed the PICAXE input voltage. The chip should have overvoltage clamp diodes on the inputs, not sure about that. The only way to scale the 4 to 20mA to 0 to 5 Volts is through the use of an op amp circuit, beyond that I can't see getting 0 Volts on the low end for your 4mA. Using 250 Ohm 1/2 watt resistors it is very unlikely one will ever open and allow an overvoltage to the ADC input.

Just My Take
Ron

 

[jsmanson]

If the 4-20ma device fails, and doesn't control the amerage properly through the device, you get 24 volts instead of 0-5 volts. I am using 24 volts to power not just the sensors, but also vairous float controls that return with 24 volts, going into a resistance bridge (DIG in's). I test the floats by sticking 24 volts on the return lines, but if I touch that 4-20ma return line by mistake...... the 10k resister in there would limit the draw without affecting the sensor operation. And a zener to VCC would also make sense to trap possible overvoltages as well.

Thanks for your help folks!

John